bmj open · for peer review only breast cancer table 8: breast cancer and pre-diagnosis diet...
TRANSCRIPT
BMJ Open is committed to open peer review. As part of this commitment we make the peer review
history of every article we publish publicly available.
When an article is published we post the peer reviewers’ comments and the authors’ responses
online. We also post the versions of the paper that were used during peer review. These are the
versions that the peer review comments apply to.
The versions of the paper that follow are the versions that were submitted during the peer review
process. They are not the versions of record or the final published versions. They should not be cited
or distributed as the published version of this manuscript.
BMJ Open is an open access journal and the full, final, typeset and author-corrected version of
record of the manuscript is available on our site with no access controls, subscription charges or pay-
per-view fees (http://bmjopen.bmj.com).
If you have any questions on BMJ Open’s open peer review process please email
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
The impact of diet on mortality in cancer survivors: A systematic review of current epidemiological literature
Journal: BMJ Open
Manuscript ID bmjopen-2016-014530
Article Type: Research
Date Submitted by the Author: 03-Oct-2016
Complete List of Authors: Jochems, Sylvia; Maastricht University, NUTRIM; University of Birmingham, Cancer and Genomic Sciences Van Osch, Frits; Maastricht University, NUTRIM; University of Birmingham, Cancer and Genomic Sciences Bryan, Richard; University of Birmingham Wesselius, Anke; Maastricht University, NUTRIM van Schooten, Frederik; Maastricht University, Toxicology Cheng, Kar Keung; University of Birmingham, Department of Public Health
and Epidemiology Zeegers, Maurice; University of Maastricht, NUTRIM School of Nutrition, Metabolism and Toxicology
<b>Primary Subject Heading</b>:
Epidemiology
Secondary Subject Heading: Oncology, Public health
Keywords: cancer survivors, mortality, diet
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open on D
ecember 7, 2020 by guest. P
rotected by copyright.http://bm
jopen.bmj.com
/B
MJ O
pen: first published as 10.1136/bmjopen-2016-014530 on 19 F
ebruary 2018. Dow
nloaded from
For peer review only
The impact of diet on mortality in cancer survivors:
A systematic review of current epidemiological literature
Sylvia H.J. Jochems (1, 2), Frits H.M. van Osch (1, 2), Richard T. Bryan (1), Anke Wesselius
(2), Frederik J. van Schooten (2), K.K. Cheng (3), Maurice P. Zeegers (2, 4)
(1) Institute of Cancer and Genomic Sciences, University of Birmingham, B15 2TT
Birmingham, United Kingdom
(2) NUTRIM School of Nutrition and Translational Research in Metabolism, Maastricht
University, 6200 MD Maastricht, The Netherlands
(3) Institute of Applied Health Research, Public Health, Epidemiology and Biostatistics,
University of Birmingham, B15 2TT Birmingham, United Kingdom
(4) CAPHRI School for Public Health and Primary Care, Maastricht University, 6200
MD Maastricht, The Netherlands
Corresponding author contact information:
Sylvia H.J. Jochems
Institute of Cancer and Genomic Sciences, University of Birmingham,
B15 2TT Birmingham, United Kingdom
Page 1 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
2
ABSTRACT
Background: It is plausible that dietary changes after a successful cancer treatment could
improve prognosis.
Objective: To determine whether there is an association between dietary intake prior to or
after cancer diagnosis and mortality in cancer survivors.
Design: Systematic review - Medline, EMBASE and the Cochrane Library were searched
from their inception to June 2016. Additional studies were identified by hand-searching and
citation tracking. Two authors independently assessed study quality and extracted the data.
Setting: A wide range of settings within primary and secondary care in Asia, Europe and
North America.
Participants: Cancer survivors of common cancers with a ten-year survival rate of 50% or
more including bladder, bowel, breast, cervical, kidney, laryngeal, prostate, testicular and
uterine cancer, malignant melanoma, and (non-)Hodgkin lymphoma.
Primary and secondary outcome measures: Mortality (overall, cancer-specific, death from
other causes) amongst cancer survivors.
Results: A total of 2 RCTs and 58 observational studies were identified. Results of this
systematic review suggest that adoption of a low-fat diet did not affect survival amongst
breast cancer survivors. Alcohol consumption is unlikely to be associated with survival
amongst breast and bowel cancer survivors. In contrast, pre-diagnosis alcohol consumption
amongst non-Hodgkin lymphoma survivors seems to negatively impact their overall life
expectancy. Multivitamin supplement use and fruit and vegetable consumption do not appear
to improve breast cancer survivors’ mortality. Other studies addressing diet and mortality
amongst groups of cancer survivors were too limited or could not be identified.
Conclusions: Evidence as of yet is still too limited to determine whether dietary behaviour
prior to or after diagnosis could influence mortality in cancer survivors. Nonetheless, the
Page 2 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
3
findings of this literature review together with results of well conducted reviews on dietary
constituents and micronutrients could guide future development of new guidelines for
nutrition in cancer survivorship.
Strengths and limitations of this study
- This systematic review provides an evaluation of studies addressing the relationship
between dietary intake and mortality amongst different groups of survivors
- Most included literature in this systematic review on diet and mortality in cancer
survivors has been obtained from results of observational studies
- By examining both food groups and dietary patterns, a translatable real-life condition
for clinical recommendations will be provided
Page 3 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4
INTRODUCTION
As cancer survival rates continue to improve, there is an increased need to identify
modifiable lifestyle factors amongst cancer survivors in order to improve long-term health.
Adherence to a diet rich in fruit and vegetables could decrease the risk of cancer 1 and
increase overall life expectancy 2-4
. The suggestion that epigenetic abbreviations occurring in
cancer could be altered by nutrients 5-7
makes it plausible that dietary changes after successful
cancer treatment could improve prognosis 8.
Although cancer survivors are susceptible to health promotion 9 10
, a recent study has
indicated that survivors had poorer diets than individuals without cancer 11
. A possible
explanation could be the difficulty for cancer survivors in adopting a healthier diet without
clear evidence that it will improve their survival 12-14
.
While dietary guidelines have been well documented for the prevention of cancer 15
,
many questions remain about nutrition after cancer treatment. In 2012, experts from the
American Institute for Cancer Research and World Cancer Research Fund concluded that
there was still a lack of sufficient studies evaluating the role of foods and drinks in cancer
survivorship to establish adequate evidence-based nutritional guidelines for survivors. Hence,
cancer survivors are advised to follow the guidelines on nutrition for primary cancer
prevention 15 16
. Healthcare providers could play a substantial role in promoting healthier
dietary choices 17
if there was clear evidence regarding the advice they could provide.
We have conducted a structured summary and evaluation of randomised controlled
trials and observational studies addressing the relationship between the highest and the lowest
intake of dietary patterns and food groups and mortality amongst different groups of
survivors of common cancers with a ten-year survival rate of 50% or more 18
to provide a
clear overview of the current status in nutritional research in this patient group.
Page 4 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
5
METHODS
From their inception up to June 2016, Medline, EMBASE and the Cochrane Library
were searched to find English language articles of randomized trials and observational studies
containing original and published research on the role of diet on mortality in cancer
survivors. Search strategies included search terms related to diet and outcomes of interest,
including overall mortality, cancer-specific mortality, and death from other causes. After the
database search was completed, articles were selected based on their title and abstract.
English language studies containing original research involving cancer survivors and
reporting on lifestyle-related diet and mortality were included. Additionally, hand searching
and reference tracking of included and related articles, literature reviews and meta-analyses
were performed. The data extraction was performed independently by two of the authors (SJ
and FvO) and any disagreements about study inclusion were resolved through consensus or a
third party.
Eligibility criteria included survivors of cancer (no sex or age restriction) who were
defined as individuals who had been diagnosed with a primary cancer, received cancer
therapy, and were in remission or had recovered completely from their cancer. Considered
cancer types were commonly-occurring cancers in the Western world with a ten-year net
survival of at least 50% (based on cancer diagnoses of men and women during 2010-2011 in
England and Wales). These include in decreasing order of net survival: testicular cancer
(98%), malignant melanoma (89%), prostate cancer (84%), Hodgkin lymphoma (80%), breast
cancer (78%), uterine cancer (77%), non-Hodgkin lymphoma (63%), cervical cancer (63%),
laryngeal cancer (62%), bowel cancer (57% including both colon and rectal cancer), bladder
cancer (50%) and kidney cancer (50%).
Dietary patterns that were considered for inclusion had to be assessed by data-driven
approaches as principal component analysis (factor analysis) and cluster analysis 19
- index-
Page 5 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
6
based methods were not considered. The following dietary patterns were included in the
search: a Western diet, prudent diet, Mediterranean diet, low fat diet, low calorie diet,
macrobiotic diet, ketogenic diet, diet low in carbohydrates, vegetarian diet, and a vegan diet.
Foods and beverages that to date appear to be promising in lowering or increasing the risk of
cancer were considered for this systematic review including fruit, vegetables, red (processed)
meat, poultry, fish, dairy, coffee, tea, alcohol and multivitamin supplements. The composition
of the investigated food and drink groups was as follows: (I) fruit and vegetables: citrus
fruits, green leafy vegetables, red vegetables, yellow vegetables, cruciferous vegetables, other
greens; (II) red (processed) meat: pork, veal, beef; (III) poultry: chicken, duck, goose, turkey,
quail, other poultry; (IV) fish: oily fish, white fish, shellfish; (V) dairy (total, high- and low-
fat items): milk, butter, cheese, casein, yogurt, gelato, ice cream; (VI) coffee and tea: coffee
("black" or with added milk products, sweeteners, or substitutes), tea (black, green, white,
oolong); (VII) alcohol: beers, wines, spirits (e.g. liquor); and (VIII) multivitamins: dietary
supplement with vitamins and minerals (and other nutritional elements). Information on
dietary intake of these foods and drinks was obtained before and/or after cancer diagnosis
with food frequency questionnaire (FFQ) (self-administered or via an interview), and
expressed in servings, grams or millilitres per day, week or month. No restrictions were made
for time of follow-up, and timing or frequency of dietary intake.
Considered endpoints were overall mortality, cancer-specific mortality, and death
from other causes. The cause of death was confirmed via death certificates or the National
Death Index in each of the studies. Recurrence or progression of cancer was not included as
endpoint of the review. Adjustments had to be made for at least pathological tumour
characteristics and disease stage at baseline and preferably for cancer treatment in the
statistical analyses. Exclusion criteria were articles not mentioning hazard ratios (HRs) or
relative risks (RRs) and 95% confidence intervals (95% CI), not providing information on
Page 6 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7
stage of disease and/or tumour grade and therapy, or where mortality was combined with
other outcomes.
The Cochrane Collaboration risk of bias assessment tools were used for appraisal of
RCTs 20
and observational studies 21
. Levels of quality were determined with the GRADE
system 22
; RCTs or multiple double-upgraded observational studies were considered as high
quality, downgraded RCTs or upgraded observational studies were considered as moderate
quality, double-downgraded RCTs or observational studies were considered as low quality,
and triple-downgraded RCTs, downgraded observational studies or case series/case reports
were considered as very low quality 22
. The principal summary measures included adjusted
hazard ratio’s (HR) or relative risks (RRs) along with their 95% confidence intervals (CI).
Finally, this review was written according to the PRISMA guidelines 23
and the review
protocol used for this systematic review can be obtained by request from the corresponding
author.
Page 7 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
8
RESULTS
The detailed search for bladder cancer survivors in MEDLINE is shown in Box 1.
This search strategy was adapted for other cancers and for use in the EMBASE database and
the Cochrane Library. Results of the literature search are shown in a PRISMA flowchart 23
and is available online as supplementary data (Figure S1). The great heterogeneity amongst
the included studies let to the decision of conducting a systematic review only and no meta-
analysis.
A total of 2 RCTs and 58 observational studies were identified. No studies could be
identified for testicular cancer, uterine cancer, Hodgkin lymphoma or malignant melanoma
survivors. Therefore, we report only on bladder, bowel, breast, cervical, kidney, laryngeal,
and prostate cancer, and non-Hodgkin lymphoma survivors. Some studies were drawn from
the same population; the number of studies per cancer type for both pre- and post-diagnosis
dietary intake was: three for bladder cancer, ten for bowel cancer, twenty-five for breast
cancer, one for cervical cancer, one for kidney cancer, one for laryngeal cancer, three for
prostate cancer, and six for non-Hodgkin lymphoma. Data from studies could not be pooled
due to the methodological diversity between the studies. A summary of the number of studies
for all cancer types on pre- and post-diagnosis dietary patterns and mortality is provided in
Tables 1 and 2. Tables 3 and 4 provide the number of studies for all cancers on pre- and post-
diagnosis food and beverages and mortality outcome. Dietary patterns that could be identified
included a diet low in fat (and high fibre), a Mediterranean diet 24
, a prudent diet (a
modification of the Mediterranean diet 25
), and a Western diet. No literature could be
identified for low calorie diets, macrobiotic diets, ketogenic diets, diets low in carbohydrates,
vegetarian diets, or vegan diets.
The RCTs investigating a low fat diet and mortality amongst breast cancer survivors
could be identified and were considered of high quality 20
. All 58 observational studies had an
Page 8 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
9
acceptable risk of bias according to the Cochrane Collaboration risk of bias assessment tool 21
for inclusion in this systematic review. The quality level of the body of evidence of the
included studies was rated ‘low’ to ‘moderate’ by two of the authors (SJ and FvO) when
applying the grading system developed by the GRADE collaboration 22
. Tables S1 – S9
present details of the included studies per dietary pattern and food or beverage and results are
presented in terms of adjusted hazard ratio’s (HR) or relative risks (RRs) along with their
95% confidence intervals (CI).
3.1 Low fat diet
Table S1 presents details of two RCTs aiming to reduce dietary fat intake to decrease the risk
of earlier mortality in breast cancer survivors.
3.1.1. Breast cancer
Two dietary intervention trials amongst breast cancer survivors met the eligibility
criteria of our literature review 26 27
. The study of Chlebowski et al. aimed to reduce post-
diagnosis dietary fat intake to almost one sixth of total energy intake in women participating
in the Women’s Intervention Nutrition Study (WINS). Breast cancer survivors in the
intervention group were informed extensively on maintaining their weight based on their
energy intake, whilst minimum dietary advice on nutrient intake was provided to breast
cancer survivors in the control group. Women in the intervention group had a statistically
significant lower dietary fat intake compared to those in the control group, whereas no
differences could be observed for a lower energy or higher dietary fibre intake. There was no
significant difference in overall mortality between women adhering to a low fat diet and
women given minimum dietary advice (HR=0.89; 95% CI 0.65-1.21) 26
.
In the Women’s Healthy Eating and Living (WHEL) study breast cancer survivors in
the intervention group received telephone counselling with additional cooking classes and
Page 9 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
10
brochures to support adherence to a post-diagnosis diet high in fruit (3 servings/day), high in
vegetables (5 servings/day and 16oz of vegetable juice), high in fibre (30g/day), and low in
fat (15-20% of energy intake from fat) 27
. In the control group, breast cancer survivors
received written advice to eat at least 5 portions of fruit and vegetables each day (5-a-day
advice). Statistically significant differences between the former and latter groups in mean
consumption of vegetables (+65%), fruit (+25%), fibre (+30%), and energy from fat (-13%)
were observed at 4 years. No statistically significant differences were observed for overall
survival comparing women in the intervention group with those in the control group
(HR=0.91; 95% CI 0.72-1.15) 27
.
3.2 Mediterranean and prudent diet
Table S2 presents details of six observational studies investigating adherence to a
Mediterranean or prudent diet and mortality in different groups of cancer survivors.
3.2.1 Bowel cancer
Three observational cohort studies could be identified investigating the role of a pre-
and post-diagnosis prudent diet on mortality outcome in bowel cancer survivors 28-30
. Results
of the Cancer and Leukemia Group B (CALGB) study indicated no associations between a
prudent diet after cancer diagnosis and lower mortality 28
. When comparing subjects recruited
from the Familial Bowel Cancer Registry with the highest and lowest intakes of a prudent
diet before cancer diagnosis, no associations were found with mortality 29
. In the Nurses'
Health Study (NHS), no associations were observed between a post-diagnosis prudent diet
and mortality when comparing the highest and lowest intakes in bowel cancer survivors 30
.
3.2.2. Breast cancer
Page 10 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
11
Pre- and post-diagnosis prudent diet intake was investigated in three observational studies on
mortality amongst breast cancer survivors 31-33
. Results of the NHS indicated that a post-
diagnosis prudent diet was not associated with overall mortality or breast cancer specific
mortality whilst death from other causes was associated with a prudent diet when comparing
breast cancer survivors of the highest and lowest intake group (HR=0.54; 95% CI 0.31-
0.95)31
. In the study of Kroenke et al., a prudent diet before and after diagnosis was not
associated with overall mortality or breast cancer specific mortality 31
. In the Life After
Cancer Epidemiology (LACE) study for a post-diagnosis prudent diet in women with early-
stage breast cancer 32
, breast cancer survivors with the highest adherence to a prudent diet had
a decreased risk of death from other causes (HR=0.35; 95% CI 0.17-0.73) and overall
mortality (HR=0.57; 95% CI 0.36-0.90) compared to women with the lowest adherence to
this diet 32
. The study of Vrieling et al. investigated associations between a ‘healthy’ and
‘unhealthy’ pre-diagnosis dietary pattern and mortality in German breast cancer survivors in
the Mammary carcinoma Risk factor Investigation (MARIE) study 33
. Because of the
characteristics of the defined healthy diet it is comparable with a prudent diet; nevertheless,
no associations between the highest and lowest intake of this defined ‘healthy’ diet before
cancer diagnosis and mortality in breast cancer survivors were observed 33
.
3.2.3. Prostate cancer
The Health Professionals Follow-up Study (HPFS) reported on a Mediterranean diet
and mortality in prostate cancer survivors after diagnosis 34
. Kenfield et al. compared subjects
with high adherence to a Mediterranean diet to those with low adherence and it appeared that
post-diagnosis adherence to a Mediterranean diet was associated with decreased overall
mortality (HR= 0.78; 95% CI 0.67-0.90); no association was observed for prostate cancer
specific mortality and adherence to the Mediterranean diet 34
. Data of the Physician’s Health
Page 11 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
12
Study (PHS) was used for investigating dietary pattern adherence after prostate cancer
diagnosis in non-metastatic prostate cancer survivors 35
. Here both a Western and prudent diet
were investigated, with the latter having overlapping characteristics with the Mediterranean
diet examined in the HPFS 34 35
. Adherence to a prudent diet after prostate cancer diagnosis
was inversely associated with overall mortality (HR=0.64; 95% CI 0.44–0.93) and appeared
to be driven by the use of oil and vinegar dressing in this group 35
.
3.2.4. Non-Hodgkin lymphoma
The only study identified for non-Hodgkin lymphoma survivors indicated that there
was no association between a fruit, vegetable and starch dietary pattern and survival 36
.
3.3 Western diet
Table S3 presents details of six observational studies investigating adherence to a Western
diet and mortality in different groups of cancer survivors.
3.3.1. Bowel cancer
Results of the CALGB study indicated significantly higher overall mortality in
women who had bowel cancer with the highest post-diagnosis intake of a Western diet in
comparison with female bowel cancer survivors in the lowest category (HR= 2.32; 95% CI
1.36-3.96) 28
. Zhu et al. identified two dietary patterns comparable with a Western diet: a
high processed meat pattern and a high sugar pattern diet for subjects recruited from the
Familial Bowel Cancer Registry 29
. No associations were reported for a dietary pattern high
in sugar and mortality when comparing the highest to the lowest intake group, whereas a high
processed meat diet was specifically related to increased colon cancer mortality (HR=2.13;
Page 12 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
13
95% CI 1.03-4.43) 29
. In the NHS, no associations were observed between a Western diet and
mortality when comparing the highest and lowest intakes in bowel cancer survivors.
3.3.2. Breast cancer
No associations were found between a Western diet and mortality in women after
breast cancer diagnosis in the LACE study 32
. However, results of the MARIE study in
Germany indicated that a higher intake of an ‘unhealthy’ diet could increase the risk of death
from other causes (HR=3.69; 95% CI 1.66-8.17) amongst breast cancer survivors compared
to those with the lowest intake of this diet 33
. Kroenke et al. found associations for both pre-
and post-diagnosis Western dietary intake and death from other causes in women of the
highest intake group compared to those in the lowest intake group (respectively RR=1.95;
95% CI 1.06-3.60 and RR=2.31; 95% CI 1.23-4.32) 31
. No associations were observed
between a pre- or post-diagnosis Western and overall or breast cancer specific mortality 31
.
3.3.3 Prostate cancer
Finally, adherence to a Western diet after prostate cancer diagnosis was associated
with earlier overall mortality (HR=1.67; 95% CI 1.16–2.42) and earlier prostate-cancer
mortality (HR= 2.53; 95% CI 1.00-6.42) amongst non-metastatic prostate cancer survivors in
the PHS 35
. The derived Western dietary patterns appeared to be driven by the consumption
of processed meat 35
.
3.4 Fruit and vegetables
Table S4 presents details of thirteen observational studies comparing highest versus
lowest or never fruit and/or vegetable intakes and mortality in different groups of cancer
survivors.
Page 13 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
14
3.4.1 Bladder cancer
Tang et al. investigated pre-diagnosis fruit and vegetable consumption with data from
bladder cancer patients from the Roswell Park Cancer Institute (RPCI) Tumor Registry, New
York 37
. No differences were observed between mortality in participants with the highest
intakes of total fruit, total vegetables or other cruciferous vegetables raw or cooked in
comparison with those in the lowest intake group 37
. Moreover, investigation of individual
cruciferous vegetable intake showed an inverse association for broccoli intake and overall
mortality (broccoli raw HR=0.57; 95% CI 0.39-0.83 and broccoli cooked HR=0.67; 95% CI
0.49-0.91) and bladder cancer-specific mortality (broccoli raw HR=0.43; 95% CI 0.25-0.74)
37. The intake of other raw and cooked vegetables including cabbage, cauliflower, Brussels
sprouts, kale, turnip, collard or mustard greens was not associated with overall mortality or
bladder cancer-specific mortality 37
.
3.4.2 Bowel cancer
Dray et al. indicated no prognostic significance for pre-diagnosis fruit and/or
vegetable intake and mortality in bowel cancer survivors 38
in cases from a case-control study
recruited from the Cote d'Or area of France and who had a histologically proven primary
bowel cancer.
3.4.3 Breast cancer
Two observational studies reported no associations between (total) pre-diagnosis fruit
and/or vegetable intake and mortality in breast cancer survivors 39 40
. However, one other
study found that, when comparing postmenopausal breast cancer survivors in the highest
tertile to the lowest tertile group, pre-diagnosis total vegetable intake improved overall
Page 14 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
15
survival (HR=0.57; 95% CI 0.35-0.94) 41
- no association was found for total fruit intake and
mortality in this cohort of breast cancer survivors. Also Dal Maso et al. found a borderline
statistical significance for total fruit and vegetable consumption and overall mortality (HR=
1.27; 95% CI 1.00-1.61) when comparing survivors of the lowest intake group to the highest
intake group 42
. Post-diagnosis total fruit and/or vegetables intake was not associated with
mortality in breast cancer survivors in three other studies 43-45
.
3.4.4 Laryngeal cancer
When comparing male laryngeal cancer survivors in the highest and lowest intakes of
vegetables pre-diagnosis, an increased survival was observed (HR=0.57; 95% CI 0.35-0.94)
46. However, no differences were found between the groups regarding the intake of fruit
46.
3.4.5 Non-Hodgkin lymphoma
The study of Han et al. indicated that pre-diagnosis high total fruit and vegetable
intake and vegetable intake only were associated with decreased overall mortality
(respectively HR= 0.68; 95% CI 0.49-0.95 and HR=0.58; 95% CI 0.38-0.89) amongst female
all types non-Hodgkin lymphoma 47
. In addition to total fruit and vegetables intake, citrus
fruits, yellow vegetables, red vegetables, green leafy vegetables and bean vegetables were
individually analysed. It appeared that high intakes of citrus fruits and green leafy vegetables
were both beneficially related to overall survival amongst survivors with all types of non-
Hodgkin lymphoma (respectively HR=0.73; 95% CI 0.54-0.99 and HR=0.71; 95% CI 0.51-
0.98). No associations were observed for total fruit intake, yellow vegetables, red vegetables
or bean vegetables and mortality 47
whilst sub-analysis investigating fruit and vegetables
separately for each non-Hodgkin lymphoma histological subtypes did 47
. A high consumption
of citrus fruits improved survival in diffuse large B-cell lymphoma survivors (overall
Page 15 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
16
mortality HR=0.40; 95% CI 0.22–0.72 and cancer-specific mortality HR=0.36; 95% CI 0.16–
0.80), a high consumption of green leafy vegetables favoured overall mortality in follicular
lymphoma survivors (HR= 0.27; 95% CI 0.10–0.76) 47
. Furthermore, a high consumption of
cruciferous vegetables and green leafy vegetables was beneficially associated with the overall
survival of women who had a T-cell lymphoma 47
. Ollberding et al. conducted a similar study
amongst both male and female non-Hodgkin lymphoma survivors 36
. No associations were
found for pre-diagnosis fruit and/or vegetables intake and non-Hodgkin lymphoma survivors’
survival. However, a high consumption of a-carotene was associated with increased survival
amongst smokers 36
.
3.5 Red meat, poultry and fish
Table S5 presents details of seven observational studies comparing highest versus lowest or
never red meat, poultry or fish consumption and mortality in different groups of cancer
survivors.
3.5.1 Bowel cancer
Dray et al. reported on pre-diagnosis meat and fish consumption and mortality
amongst bowel cancer survivors 38
. The study of Zell et al. indicated a detrimental association
for mortality when comparing highest versus lowest or never pre-diagnosis meat
consumption amongst familial bowel cancer survivors (HR=2.24; 95% CI 1.25-4.03) 48
.
Results from the Côte d’Or area case-control study reported no associations between meat or
fish consumption and bowel cancer survivors’ mortality 38
.
McCullough et al. reported on both pre- and post-diagnosis red and processed meat intake
amongst bowel cancer survivors 49
. A detrimental association was observed between pre-
diagnosis red and processed meat intake and overall mortality (HR=1.29; 95% CI 1.05-1.59)
Page 16 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
17
49. Moreover, combined data from the same study found that low pre-diagnosis / high post-
diagnosis red and processed meat intake was associated with death from other causes
(HR=1.62; 95% CI 1.06-2.48), as well as an association between high pre-diagnosis / low
post-diagnosis red and processed meat consumption and overall mortality (HR=1.37; 95% CI
1.02-1.85) 49
. Recently, Carr et al. reported that red and processed meat consumption was not
associated with a poorer survival amongst stage I–III bowel cancer survivors in a follow-up
study of the Darmkrebs: Chancen der Verhutung durch Screening (DACHS) study in the
Rhine-Neckar region in southwest Germany 50
. However, the authors do suggest that major
changes in the consumption of red and processed meat measured at 5 year follow-up could
influence survival estimates 50
.
3.5.2 Breast cancer
Holmes et al. reported a beneficial association between high post-diagnosis poultry
consumption and mortality in women once diagnosed with breast cancer (HR=0.70; 95% CI
0.50–0.97) 44
. No associations were found for high fish or red meat consumption and
mortality in this population 44
.
3.5.3 Laryngeal cancer
Crosignani et al. investigated the role of pre-diagnosis meat and fish consumption on
mortality in male laryngeal cancer survivors 46
. When comparing the highest with the lowest
intakes of meat, a beneficial association was found for the consumption of meat (veal/beef)
on mortality (HR=0.50; 95% CI 0.30-0.83). The authors speculate that the beneficial relation
between high meat consumption and survival could tentatively be interpreted as an indicator
of a good nutritional status of those participants. Furthermore, no associations were identified
for poultry or fish consumption in this study 46
.
Page 17 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
18
3.5.4 Prostate cancer
A pre-diagnosis high fish consumption in men who were diagnosed with prostate
cancer while participating in the PHS was related to prolonged survival (HR=0.52; 95% CI
0.30-0.91) according to Chavarro et al. 51
. Richman et al. observed no association between
post-diagnosis red meat or poultry consumption and mortality amongst prostate cancer
survivors derived from the HPFS 52
.
3.6 Dairy
Table S6 presents details of three observational studies comparing highest versus lowest or
never dairy consumption and mortality in different groups of cancer survivors.
3.6.1 Bowel cancer
Dray et al. found no association between survivors with the highest pre-diagnosis
intakes of dairy comparing to those with the lowest or never intakes in a case-control study in
the Côte d’Or area in France 38
.
3.6.2 Breast cancer
Kroenke et al. found that the highest compared to the lowest or no total dairy intake
post-diagnosis amongst women diagnosed with early-stage invasive breast cancer in the
LACE study was associated with an increased overall mortality (HR=1.39; 95% CI 1.02-
1.90) – this could probably be explained by the high fat dairy consumption and overall
mortality and breast cancer specific mortality in these women (respectively HR= 1.64; 95%
CI 1.24-2.17 (p trend = <0.001) and HR= 1.49; 95% CI 1.00 to 2.24 (p trend = 0.05)) as low
fat dairy consumption was not related with mortality 53
. Results of Holmes et al. indicated
Page 18 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
19
that high dairy intake before diagnosis amongst female registered nurses who participated in
the NHS, had a borderline beneficial statistical significance with their overall survival (HR=
0.72; 95% CI 0.52–1.00) 44
.
3.6.3 Prostate cancer
One study of Yang et al. investigated post-diagnosis dairy intake amongst prostate cancer
survivors. The consumption of total dairy was non-beneficially associated with overall
mortality (HR=1.76; 95% CI 1.21-2.55) 54
. Both high-fat and low-fat dairy consumption
contributed to this adverse association and overall mortality (respectively HR=1.22; 95% CI
1.08-1.38 and HR=1.17; 95% CI 1.05-1.29) 54
.
3.7 Coffee and tea
Table S7 presents details of three observational studies comparing highest versus lowest or
never coffee and/or tea consumption and mortality in different groups of cancer survivors.
3.7.1 Bladder cancer
Wakai et al. examined pre-diagnosis coffee and tea consumption in a study of bladder
cancer survivors who were originally recruited from a population-based case-control study of
patients with bladder cancer and of controls drawn randomly from the general population of
Metropolitan Nagoya in Japan and reported no association with mortality in bladder cancer
survivors 55
.
3.7.2 Bowel cancer
Dray et al. found no association between a high consumption of coffee and tea pre-
diagnosis and mortality in bowel cancer patients of the Côte d’Or area in France 38
.
Page 19 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
20
3.7.3 Breast cancer
Results from the Swedish Mammography Cohort (SMC) conducted by Harris et al.
indicated no association between the post-diagnosis consumption of coffee or tea and
mortality amongst women diagnosed with breast cancer 56
.
3.8 Alcohol
Table S8 presents details of twenty-six observational studies comparing highest versus lowest
or never alcohol consumption and mortality in different groups of cancer survivors.
3.8.1 Bladder cancer
Wakai et al. investigated pre-diagnosis alcohol consumption amongst bladder cancer
survivors who were originally recruited from a population-based case-control study of
patients with bladder cancer and controls drawn randomly from the general population of
Metropolitan Nagoya in Japan. Pre-diagnosis alcohol consumption appeared to be associated
with an increased survival when comparing (moderate) drinkers with non-drinkers (HR=0.46;
95% CI 0.26-0.79) 55
. The same study that investigated pre-diagnosis alcohol consumption
amongst bladder cancer survivors also examined post-diagnosis alcohol use in the same
population. A beneficial role for (moderate) alcohol consumption on survival was found
compared non-drinkers (HR=0.43; 95% CI 0.24-0.77) 55
.
3.8.2 Bowel cancer
Zell et al. indicated no association between pre-diagnosis consumption of beer and
liquor and survival in neither familial nor sporadic bowel cancer survivors 57
. However, a
beneficial role for the consumption of wine was observed in familial bowel cancer survivors’
Page 20 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
21
overall survival (HR=0.50; 95% CI 0.25- 0.99) whilst no association was found for wine
drinkers amongst sporadic bowel cancer survivors 57
. Pre-diagnosis alcohol use was not
related with mortality in bowel cancer survivors according to Park et al. who investigated
data from male Korean government employees and teachers who participated in the National
Health Insurance Corporation Study (NHICS) 58
. Furthermore, no associations were found
between pre-diagnosis alcohol consumption and overall mortality or cancer-specific mortality
amongst bowel cancer survivors living in the USA 59 60
. Moreover, Pelser et al. did find a
relation between the consumption of alcohol before diagnosis and death from cardiovascular
disease amongst these survivors (HR=0.44; 95% CI 0.24-0.81) 59
.
Lochhead et al. used data from two prospective cohort studies, the NHS and HPFS,
and concluded that higher intakes of alcohol post-diagnosis were not related to earlier
mortality amongst stage I–III bowel cancer survivors 61
.
3.8.3 Breast cancer
Several observational studies found no association between pre-diagnosis alcohol
consumption and mortality in breast cancer survivors 39 42 62-66
. Reding et al. reported that pre-
diagnosis alcohol consumption amongst young women diagnosed with invasive breast cancer
from two population-based case-control studies in the Seattle-Puget Sound region, could
possibly prolong life in survivors of breast cancer (HR=0.7; 95% CI 0.5-0.9) 67
. A beneficial
association with alcohol consumption prior to diagnosis was also observed in the study of
Lowry et al. HR=0.74; 95% CI 0.61-0.89 68
. Results of the study of McDonald et al.
suggested the opposite; alcohol consumption amongst postmenopausal African American
women diagnosed with an invasive breast carcinoma was associated with a poorer survival
compared to non-drinkers (HR=2.8; 95% CI 1.2-7.0) 69
.
Page 21 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
22
Post-diagnosis alcohol consumption appeared not to be related with breast cancer survivors’
mortality in the majority of the identified studies 43 44 63 68 70 71
. Tominaga et al. found a
beneficial impact of post-diagnosis alcohol consumption on survival HR=0.1; 95% CI 0.01-
0.72 in women with treated breast cancers 72
. Similar results were observed for alcohol intake
and mortality for female breast cancer survivors who participated in the WHEL study,
HR=0.69; 95% CI 0.49-0.97 73
and the CBCS, HR= 0.64; 95% CI 0.47-0.88 66
. A non-
beneficial association with alcohol consumption after diagnosis was found for women
participating in the Memorial Sloan-Kettering Cancer Center Follow-up Study and cancers-
specific mortality (RR= 1.58; 95% CI 1.00-2.78) 74
and in the study of Kwan et al. regarding
alcohol consumption and mortality from breast cancer (HR= 1.51; 95% CI 1.00-2.29) 75
.
3.8.4 Cervical cancer
Serur et al. reported that pre-diagnosis alcohol abuse amongst young women with
cervical cancer registered at the University Hospital of Brooklyn and Kings County Hospital
Centre and living in a low income community, was not associated with cervical cancer-
specific mortality 76
.
3.8.5 Non-Hodgkin lymphoma
Three studies investigating pre-diagnosis dietary habits amongst non-Hodgkin
lymphoma survivors reported a non-beneficial role for highest versus lowest or never alcohol
consumption and overall mortality (HR=1.41; 95% CI 1.10-1.81 77
, HR=1.69; 95% CI 1.04-
2.76 78
and HR=1.55; 95% CI 1.06-2.27 79
). One other study reported no association between
pre-diagnostic alcohol consumption and survival in non-Hodgkin lymphoma survivors 80
.
3.8.6 Laryngeal cancer
Page 22 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
23
No relationship was found between pre-diagnosis alcohol consumption and survival in
male laryngeal cancer survivors in the only study that could be identified 46
.
3.9 Multivitamin supplements
Table S9 presents details of five observational studies comparing highest versus lowest or
never multivitamin supplement use and mortality in different groups of cancer survivors.
3.9.1 Bowel cancer
In a follow-up study from CALGB, post-diagnosis multivitamin supplement use
amongst survivors of stage III colon cancer during chemotherapy and after treatment was
investigated – no associations were found with survival when comparing consistent users or
inconsistent users to never users 81
.
3.9.2 Breast cancer
Pre-diagnosis multivitamin supplement use appeared not to be related to breast cancer
survivors’ mortality in a follow-up study of the NHS 44
. One study investigated combined
pre- and post-diagnosis multivitamin supplement use amongst women with early stage breast
cancer and found no association with survival 82
. The authors of this study also investigated
post-diagnosis multivitamin intake only – however these results will not be presented as
Greenlee et al. 83
also reported on this in the same cohort.
Holmes et al. 44
, Greenlee et al. 83
and Nechuta et al. 84
all reported no relationship
between post-diagnosis multivitamin supplement intake and breast cancer survivors’
mortality despite corrections for pre-diagnosis supplementary intake. The After Breast
Cancer Pooling Project (ABCPP) pooled data from breast cancer patients participating in the
Shanghai Breast Cancer Survival Study (SBCSS), the LACE study, the WHEL Study and the
Page 23 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
24
NHS 85
. Results from this consortium (including the studies of Holmes et al., Greenlee et al.,
and Nechuta et al.) indicated no association between post-treatment multivitamin use and
breast cancer survivors’ overall mortality 85
.
Page 24 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
25
DISCUSSION
This systematic review included current scientific literature with regard to diet and
mortality in cancer survivors. No studies could be identified for testicular or uterine cancer,
Hodgkin lymphoma or malignant melanoma survivors and the role of foods and drinks with
mortality as an outcome. The available literature on specific dietary items impacting
mortality in bladder, bowel, breast, cervical, laryngeal and prostate cancer and non-Hodgkin
lymphoma survivors is limited, although compelling.
Pre- and post-diagnosis alcohol consumption appears not to be associated with
survival in bowel cancer survivors (pre-diagnosis 3 studies no association versus 1 study
decreased mortality risk; post-diagnosis 1 study decreased mortality). Amongst breast cancer
survivors, neither pre- nor post-diagnosis fruit and/or vegetable consumption appears to be
associated with survival in breast cancer survivors (pre-diagnosis 2 studies no association
versus 2 studies decreased mortality risk; post-diagnosis 4 studies no association).
Furthermore, the majority of studies indicate no relationship between pre- or post-diagnosis
alcohol consumption and mortality when comparing moderate/heavy drinkers to non-drinkers
amongst breast cancer survivors (pre-diagnosis 6 studies no association versus 1 study
increased mortality risk versus 2 studies decreased mortality risk; post-diagnosis 8 studies no
association versus 2 studies increased mortality risk versus 1 study decreased mortality risk).
No associations were found between pre- or post-diagnosis multivitamin supplement intake
and prolonged survival in breast cancer survivors (pre-diagnosis 1 study no association; post-
diagnosis 5 studies no association). Finally, reducing or stop drinking alcohol might be
beneficial for prolonging life amongst non-Hodgkin lymphoma survivors (pre-diagnosis 3
studies increased mortality risk versus 1 study no association).
Regarding dietary patterns, 2 RCTs reported on adherence to a low fat diet high in
fruit, vegetables and fibre after diagnosis and mortality in breast cancer survivors, whereas 3
Page 25 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
26
studies investigated a Western and prudent diet. Associations between a Western and prudent
diet and mortality were investigated in 3 studies in bowel cancer survivors. For prostate
cancer survivors, 2 studies could be identified that investigated the Mediterranean diet and
mortality and 1 study investigating the Western diet.
The results of this systematic review should be interpreted with caution; evidence
from observational studies generally provides a lower strength of evidence than RCTs even if
they were well conducted. Furthermore, cancer is not one disease; the development of cancer
involves many different mechanisms and therefore the impact of dietary intake can vary
between different types of cancers.
4.1. Dietary patterns
Results of the included RCTs concluded that adoption of a diet low in fat (with or
without an increased intake of fruits, vegetables and fibre) did not reduce overall mortality 26
27. It could be speculated that the lack of effect in the two identified RCTs investigating a low
fat diet is a consequence of the relatively short follow-up period when using mortality as the
primary outcome. Furthermore, the true beneficial effect of diet remains uncertain since
increased exercise and weight loss during the intervention may also have advantaged these
breast cancer survivors.
The limited amount of studies indicate that additional long-term prospective cohort
studies (for example the DietCompLyf study 86
) are urgently needed to improve the strength
of evidence in the literature on the role of dietary pattern adherence on cancer survival.
4.2 Fruit and vegetables
Epidemiological research on fruit and vegetable intake and cancer risk increased
rapidly over the last decades and it has been suggested that people consuming high portions
Page 26 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
27
of fruit and vegetables, compared to those with a low consumption, have a reduced risk of
developing cancer 87
.
The wide variety of nutrients including vitamins, minerals, phytochemicals and fibre
in fruit and vegetables could influence epigenetic processes and via this way improve cancer
progression 5 7
. Furthermore, DNA repair capacity can be modulated by fruits and vegetables
88 89 and it is reasonable to assume that fruit and vegetable consumption after DNA damaging
cancer treatment could potentially help slow down the progression of cancer. However, the
exact mechanism how diet could alter genetic and epigenetic changes in cancer cells has still
to be established. Even though the majority of the identified studies found no association
between fruit and/or vegetable consumption and survivors’ mortality, the consumption of a
wide variety of fruit and vegetables should be encouraged in cancer survivors as they are an
important part of a balanced and healthy diet to maintain general health 90
.
4.3 Red meat, poultry and fish
Experts of the World Cancer Research Fund and American Institute for Cancer
Research determined that the consumption of red and processed meat (containing haem,
nitrates and nitrites) is a risk factor for the development of bowel cancer 15
. Notwithstanding,
red meat provides a useful source of protein, iron and zinc and eating smaller portions of
meat or less often could fit into a healthy diet. With vegetarianism rates increasing, future
research could focus on comparing vegetarians with individuals who eat meat to elucidate the
relationship between meat consumption and prolonging cancer survivorship. A role for the
consumption of fish, rich in n−3 polyunsaturated fatty acids and vitamin D, is still unclear in
both cancer development and cancer survival and needs further research.
4.4 Dairy
Page 27 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
28
Dairy products contain micronutrients and bioactive constituents that may influence
cancer progression. It is proposed that the link between dairy products and cancer
development could be caused by the increase in insulin-like growth factor I levels that
stimulate cancer cell growth. Hence, more well-conducted studies are needed in diverse
populations (e.g. Asian population versus Western population) because of the differences in
dairy consumption, to gain further understanding of the role dairy in cancer survivorship.
4.5 Coffee and tea
A growing body of evidence suggests polyphenols, found mostly in tea however also
in coffee, could inhibit inflammation and have anti-carcinogenic properties that can be
archived through DNA methylation 91
. Future studies should focus on epigenetic mechanisms
in different populations of cancer survivors to see whether survival can be improved by
increasing the intake of tea or coffee.
4.6 Alcohol
Alcohol consumption is widely recognized to be associated with an increased risk of
the development of breast cancer 15 92
. The conversion of alcohol into acetaldehyde and the
increase of circulating oestrogen levels as well as issues with maintaining DNA integrity are
suggested to play a role in the development and recurrence of some cancers 93
.
While most studies find no association with mortality in cancer survivors, some
studies report a beneficial role for alcohol consumption with mortality outcome. A possible
explanation for this could be that drinkers stopped drinking alcohol before their cancer
diagnosis because of complaints that could already have been the first symptoms of their
cancer – therefore these individuals will be grouped into the non-drinkers category when
participating in a study and potentially bias the outcomes. Another reason could be that
Page 28 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
29
drinking moderate amounts of alcohol could work stress-reducing for some cancer survivors
and therefore have an impact on their general well-being. Nevertheless, it should never be
encouraged to start consuming alcohol after cancer recovery or to exceed the recommended
maximum amount of alcohol provided in evidence-based guidelines. Future studies should,
however, investigate whether cancer survivors drinking light to moderate amounts of alcohol
could continue their drinking habits without decreasing their survival after cancer treatment.
4.7 Multivitamin supplements
A balanced diet with plenty of fruit and vegetables should be enough to get the
recommended adequate intake of vitamins and minerals 94
. Nevertheless, many people decide
to take multivitamin supplements to maintain or improve their health. Although Ng et al.
found no association for pre- and post-diagnosis multivitamin supplement intake in bowel
cancer survival, vitamin D appears to be a promising candidate; high serum 25(OH)D
concentrations in bowel cancer patients were associated with a lower risk of mortality 95
and
should therefore be further investigated.
Page 29 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
30
4.8 The lack of literature on diet and mortality outcome in cancer survivors
No literature could be found for the role of diet and mortality after cancer of the
cervix, testis, uterus, multiple myeloma or Hodgkin’s lymphoma. When a particular diet or
food group is suspected to influence the risk of developing cancer, this does not necessarily
mean that this is true for all types of cancer or for people who already survived their first
cancer. Hence, more research on dietary intake should be conducted amongst cancer
survivors to provide better guidance on nutrition in the period after initial treatment 96
. Future
research should not only focus on the ‘big four’ (breast, bowel, prostate and lung cancer), but
also expand to other cancers with a relatively high survival rate. Understandably, it can be
challenging to recruit enough participants to take part in RCTs or cohort studies on long-term
survival for less common cancers. However, it could be considered to combine studies
involving cancer survivors from different countries to increase participant recruitment as well
as generalisation. Also the rapidly increasing populations of cancer survivors should make
conducting RCTs more feasible. However, caution must be taken since comparing outcomes
from multiple countries could introduce many confounding factors and lead to misleading
results.
The majority of studies included in this literature review investigated dietary intakes
before cancer diagnosis and only a few studies dietary intake after diagnosis. Therefore, the
answer to the question whether dietary changes after cancer diagnosis could improve survival
is still unclear. Information on dietary intake after diagnosis is very valuable in evaluating the
effect of dietary changes amongst cancer survivors in prolonging their life expectancy.
Further studies are needed to investigate whether changing a diet after diagnosis could indeed
improve survival. Fortunately, there has been a growing interest in investigating post-
diagnosis dietary changes in cancer survivors in the last few years 9.
Page 30 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
31
4.9 Dietary needs amongst cancer survivors
Even though healthcare providers could play a substantial role in promoting dietary
choices amongst cancer survivors 17
, this is not part of routine care and dietary advice is often
discussed on request of the patient only. This is unfortunate as it has been demonstrated that
the need for dietary guidance is often unmet 97 98
. Consequently, survivors search for
guidance elsewhere, often leaving them with conflicting information. It is indicated that
uncertainty about nutrition after treatment makes survivors stressed and vulnerable 99
. More
accessible evidence-based information on diet and cancer 100
and advice from an oncology
nutritionist after treatment on dietary intake could contribute to cancer survivors’ needs 99
.
4.10 Study strengths and limitations
Whilst most nutritional studies focus predominantly on individual nutrients, this
review summarizes evidence on dietary patterns and food groups. A food- and drink-based
approach was chosen based on promising foods and drinks that are likely to influence the risk
of developing cancer – identifying dietary patterns and groups of foods and drinks reflects a
more real-life condition as most people eat more than one single food item from a certain
food group. Hence, it is unlikely that one dietary item will impact survival after a successful
cancer treatment; therefore survivors should always be encouraged to adhere to an overall
healthy and balanced diet. By examining the whole diet, intake of nutrients in combination is
considered, which will provide a translatable real-life condition for clinical recommendations
25.
Most included literature in this systematic review on diet and mortality in cancer
survivors has been obtained from results of observational studies. Due to potential
confounding biases in observational studies, evidence from these studies is weaker than from
randomized controlled trials. Nevertheless, observational studies provide strong preliminary
Page 31 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
32
evidence that can be used for hypotheses in randomized controlled trials. Conducting RCTs
investigating diet in cancer survivors with mortality as an outcome can be challenging for
cancers with a relatively long survival and adherence to a diet long-term. Nevertheless, small
changes like increasing the amounts of fruit and vegetables and reducing the consumption of
meat and alcohol in the diet are feasible goals and should be highly considered based on the
results identified. All studies were conducted in cancer survivors that were treated for their
primary cancer. Investigating the role of nutrition during treatment would not be
representative for an individual’s usual dietary intake as side effects of treatments (e.g.
nausea, vomiting during chemotherapy) could influence this 101
. All studies used a FFQ as an
instrument to collect dietary information from the participants. The use of FFQs is an
inexpensive and convenient approach to collect data from hundreds of individuals.
Furthermore, there was a great heterogeneity between the studies in time of follow-up,
timing of dietary intake (pre-diagnosis versus post-diagnosis), frequency of dietary intake
(highest versus lowest or yes versus no comparisons), and age (varying from early adulthood
to late adulthood). Furthermore, studies from different continents (North America, Europe
and Asia) were included - developments in screening, diagnosis and treatment of cancers
differ greatly between countries and therefore could influence survival. Finally, often no
adjustments could be made for other influential lifestyle factors - including body weight,
physical activity and smoking. Residual confounding could also be induced by other
nutritional factors. It remains difficult to disentangle the impact of diet from other lifestyle
factors and should always be taken into consideration when interpreting study results.
Page 32 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
33
CONCLUSION
This systematic review accentuates the lack of evidence-based literature on the role of
diet in mortality outcome after a successful cancer treatment. The limited and conflicting
amount of evidence and limitations in the design or execution of the studies makes it difficult
to provide evidence-based recommendations for specific groups of cancer survivors.
Nonetheless, the findings of this literature review together with results of well conducted
reviews on dietary constituents and micronutrients, could guide future development of new
guidelines for nutrition in cancer survivorship.
Currently, the best advice for cancer survivors is to follow the guidelines for cancer
prevention unless other specific dietary advice has been provided by a physician or dietician.
It is important that results of well-conducted scientific studies, although preliminary, reach
health care providers so that cancer survivors can be informed by them on dietary factors that
could possibly influence their survival. This could motivate survivors to make dietary
changes and provide them with a greater sense of control over their own survival.
Conflict of interest
None of the authors have any conflict of interest in connection with this systematic review.
Funding
This research received no specific grant from any funding agency in the public, commercial
or not-for-profit sectors.
Acknowledgements
We wish to thank medical librarian Jon Andrews of the University of Birmingham for his
help with the literature search.
Page 33 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
34
Contributorship statement
SJ drafted the manuscript and worked on the conception, design and interpretation of data. SJ
and FvO selected articles, assessed study quality and extracted data. SJ, FvO, RB, AW, FJvS,
KKC and MZ were involved in the interpretation and discussion of the results and critically
revised the systematic review for important intellectual content. And all authors, SJ, FvO,
RB, AW, FJvS, KKC and MZ, approved the final version of the systematic review. SJ is the
guarantor.
Data sharing statement
No additional data available.
Page 34 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
35
LITERATURE
1. Parkin DM. 1. The fraction of cancer attributable to lifestyle and environmental factors in
the UK in 2010. British journal of cancer 2011;105 Suppl 2:S2-5.
2. Leenders M, Sluijs I, Ros MM, et al. Fruit and vegetable consumption and mortality:
European prospective investigation into cancer and nutrition. American journal of
epidemiology 2013;178(4):590-602.
3. Oyebode O, Gordon-Dseagu V, Walker A, et al. Fruit and vegetable consumption and all-
cause, cancer and CVD mortality: analysis of Health Survey for England data. Journal
of epidemiology and community health 2014;68(9):856-62.
4. Nguyen B, Bauman A, Gale J, et al. Fruit and vegetable consumption and all-cause
mortality: evidence from a large Australian cohort study. The international journal of
behavioral nutrition and physical activity 2016;13(1):9.
5. Hardy TM, Tollefsbol TO. Epigenetic diet: impact on the epigenome and cancer.
Epigenomics 2011;3(4):503-18.
6. Tollefsbol TO. Dietary epigenetics in cancer and aging. Cancer treatment and research
2014;159:257-67.
7. Daniel M, Tollefsbol TO. Epigenetic linkage of aging, cancer and nutrition. The Journal of
experimental biology 2015;218(Pt 1):59-70.
8. Pekmezi DW, Demark-Wahnefried W. Updated evidence in support of diet and exercise
interventions in cancer survivors. Acta oncologica (Stockholm, Sweden)
2011;50(2):167-78.
9. Bours MJ, Beijer S, Winkels RM, et al. Dietary changes and dietary supplement use, and
underlying motives for these habits reported by colorectal cancer survivors of the
Patient Reported Outcomes Following Initial Treatment and Long-Term Evaluation of
Survivorship (PROFILES) registry. The British journal of nutrition 2015;114(2):286-
96.
10. Humpel N, Magee C, Jones SC. The impact of a cancer diagnosis on the health behaviors
of cancer survivors and their family and friends. Supportive care in cancer : official
journal of the Multinational Association of Supportive Care in Cancer
2007;15(6):621-30.
11. Zhang FF, Liu S, John EM, et al. Diet quality of cancer survivors and noncancer
individuals: Results from a national survey. Cancer 2015;121(23):4212-21.
12. Costanzo ES, Lutgendorf SK, Roeder SL. Common-sense beliefs about cancer and health
practices among women completing treatment for breast cancer. Psycho-oncology
2011;20(1):53-61.
13. Rabin C, Pinto B. Cancer-related beliefs and health behavior change among breast cancer
survivors and their first-degree relatives. Psycho-oncology 2006;15(8):701-12.
14. Rabin CS, Pinto BM, Trunzo JJ, et al. Physical activity among breast cancer survivors:
regular exercisers vs participants in a physical activity intervention. Psycho-oncology
2006;15(4):344-54.
15. WCRF/AICR. World Cancer Research Fund/American Institute for Cancer Research
(2007) Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global
Perspective. Washington, DC: AICR. 2007.
16. Rock CL, Doyle C, Demark-Wahnefried W, et al. Nutrition and physical activity
guidelines for cancer survivors. CA: a cancer journal for clinicians 2012;62(4):243-
74.
17. Coa KI, Smith KC, Klassen AC, et al. Capitalizing on the "teachable moment" to promote
healthy dietary changes among cancer survivors: the perspectives of health care
Page 35 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
36
providers. Supportive care in cancer : official journal of the Multinational Association
of Supportive Care in Cancer 2015;23(3):679-86.
18. Quaresma M, Coleman MP, Rachet B. 40-year trends in an index of survival for all
cancers combined and survival adjusted for age and sex for each cancer in England
and Wales, 1971-2011: a population-based study. Lancet (London, England)
2015;385(9974):1206-18.
19. Reedy J, Wirfalt E, Flood A, et al. Comparing 3 dietary pattern methods--cluster analysis,
factor analysis, and index analysis--With colorectal cancer risk: The NIH-AARP Diet
and Health Study. American journal of epidemiology 2010;171(4):479-87.
20. Higgins JP, Altman DG, Gotzsche PC, et al. The Cochrane Collaboration's tool for
assessing risk of bias in randomised trials. BMJ (Clinical research ed)
2011;343:d5928.
21. Sterne JAC HJ, Reeves BC on behalf of the development group for ACROBAT-NRSI. A
Cochrane Risk Of Bias Assessment Tool: for Non-Randomized Studies of
Interventions (ACROBAT-NRSI), Version 1.0.0, 24 September 2014. Available from
http://www.riskofbias.info [accessed September 2015]. 2014.
22. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality
of evidence and strength of recommendations. BMJ (Clinical research ed)
2008;336(7650):924-6.
23. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews
and meta-analyses: the PRISMA statement. PLoS medicine 2009;6(7):e1000097.
24. Bach-Faig A, Berry EM, Lairon D, et al. Mediterranean diet pyramid today. Science and
cultural updates. Public health nutrition 2011;14(12a):2274-84.
25. Hu FB, Rimm EB, Stampfer MJ, et al. Prospective study of major dietary patterns and
risk of coronary heart disease in men. The American journal of clinical nutrition
2000;72(4):912-21.
26. Chlebowski RT, Blackburn GL, Thomson CA, et al. Dietary fat reduction and breast
cancer outcome: interim efficacy results from the Women's Intervention Nutrition
Study. Journal of the National Cancer Institute 2006;98(24):1767-76.
27. Pierce JP, Natarajan L, Caan BJ, et al. Influence of a diet very high in vegetables, fruit,
and fiber and low in fat on prognosis following treatment for breast cancer: the
Women's Healthy Eating and Living (WHEL) randomized trial. Jama
2007;298(3):289-98.
28. Meyerhardt JA, Niedzwiecki D, Hollis D, et al. Association of dietary patterns with
cancer recurrence and survival in patients with stage III colon cancer. Jama
2007;298(7):754-64.
29. Zhu Y, Wu H, Wang PP, et al. Dietary patterns and colorectal cancer recurrence and
survival: A cohort study. BMJ Open 2013;3 (2) (no pagination)(002270).
30. Fung TT, Kashambwa R, Sato K, et al. Post diagnosis diet quality and colorectal cancer
survival in women. PloS one 2014;9 (12) (no pagination)(e115377).
31. Kroenke CH, Fung TT, Hu FB, et al. Dietary patterns and survival after breast cancer
diagnosis. Journal of Clinical Oncology 2005;23(36):9295-303.
32. Kwan ML, Weltzien E, Kushi LH, et al. Dietary patterns and breast cancer recurrence and
survival among women with early-stage breast cancer. Journal of Clinical Oncology
2009;27(6):919-26.
33. Vrieling A, Buck K, Seibold P, et al. Dietary patterns and survival in German
postmenopausal breast cancer survivors. British journal of cancer 2013;108(1):188-
92.
Page 36 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
37
34. Kenfield SA, DuPre N, Richman EL, et al. Mediterranean diet and prostate cancer risk
and mortality in the Health Professionals Follow-up Study. European urology
2014;65(5):887-94.
35. Yang M, Kenfield SA, Van Blarigan EL, et al. Dietary patterns after prostate cancer
diagnosis in relation to disease-specific and total mortality. Cancer prevention
research (Philadelphia, Pa) 2015;8(6):545-51.
36. Ollberding NJ, Aschebrook-Kilfoy B, Caces DB, et al. Dietary intake of fruits and
vegetables and overall survival in non-Hodgkin lymphoma. Leukemia & lymphoma
2013;54(12):2613-9.
37. Tang L, Zirpoli GR, Guru K, et al. Intake of cruciferous vegetables modifies bladder
cancer survival. Cancer epidemiology, biomarkers & prevention : a publication of the
American Association for Cancer Research, cosponsored by the American Society of
Preventive Oncology 2010;19(7):1806-11.
38. Dray X, Boutron-Ruault MC, Bertrais S, et al. Influence of dietary factors on colorectal
cancer survival. Gut 2003;52(6):868-73.
39. Saxe GA, Rock CL, Wicha MS, et al. Diet and risk for breast cancer recurrence and
survival. Breast cancer research and treatment 1999;53(3):241-53.
40. Buck K, Zaineddin AK, Vrieling A, et al. Estimated enterolignans, lignan-rich foods, and
fibre in relation to survival after postmenopausal breast cancer. British journal of
cancer 2011;105(8):1151-7.
41. McEligot AJ, Largent J, Ziogas A, et al. Dietary fat, fiber, vegetable, and micronutrients
are associated with overall survival in postmenopausal women diagnosed with breast
cancer. Nutrition and cancer 2006;55(2):132-40.
42. Dal Maso L, Zucchetto A, Talamini R, et al. Effect of obesity and other lifestyle factors
on mortality in women with breast cancer. International journal of cancer Journal
international du cancer 2008;123(9):2188-94.
43. Beasley JM, Newcomb PA, Trentham-Dietz A, et al. Post-diagnosis dietary factors and
survival after invasive breast cancer. Breast cancer research and treatment
2011;128(1):229-36.
44. Holmes MD, Stampfer MJ, Colditz GA, et al. Dietary factors and the survival of women
with breast carcinoma. Cancer 1999;86(5):826-35.
45. Nechuta S, Caan BJ, Chen WY, et al. Postdiagnosis cruciferous vegetable consumption
and breast cancer outcomes: a report from the After Breast Cancer Pooling Project.
Cancer epidemiology, biomarkers & prevention : a publication of the American
Association for Cancer Research, cosponsored by the American Society of Preventive
Oncology 2013;22(8):1451-6.
46. Crosignani P, Russo A, Tagliabue G, et al. Tobacco and diet as determinants of survival
in male laryngeal cancer patients. International journal of cancer Journal international
du cancer 1996;65(3):308-13.
47. Han X, Zheng T, Foss F, et al. Vegetable and fruit intake and non-Hodgkin lymphoma
survival in Connecticut women. Leukemia & lymphoma 2010;51(6):1047-54.
48. Zell JA, Ziogas A, Bernstein L, et al. Meat consumption, nonsteroidal anti-inflammatory
drug use, and mortality among colorectal cancer patients in the California Teachers
Study. Cancer prevention research (Philadelphia, Pa) 2010;3(7):865-75.
49. McCullough ML, Gapstur SM, Shah R, et al. Association between red and processed
meat intake and mortality among colorectal cancer survivors. Journal of clinical
oncology : official journal of the American Society of Clinical Oncology
2013;31(22):2773-82.
Page 37 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
38
50. Carr PR, Jansen L, Walter V, et al. Associations of red and processed meat with survival
after colorectal cancer and differences according to timing of dietary assessment. The
American journal of clinical nutrition 2016;103(1):192-200.
51. Chavarro JE, Stampfer MJ, Hall MN, et al. A 22-y prospective study of fish intake in
relation to prostate cancer incidence and mortality. American Journal of Clinical
Nutrition 2008;88(5):1297-303.
52. Richman EL, Kenfield SA, Stampfer MJ, et al. Egg, red meat, and poultry intake and risk
of lethal prostate cancer in the prostate-specific antigen-era: incidence and survival.
Cancer prevention research (Philadelphia, Pa) 2011;4(12):2110-21.
53. Kroenke CH, Kwan ML, Sweeney C, et al. High-and low-fat dairy intake, recurrence, and
mortality after breast cancer diagnosis. Journal of the National Cancer Institute
2013;105(9):616-23.
54. Yang M, Kenfield SA, Van Blarigan EL, et al. Dairy intake after prostate cancer
diagnosis in relation to disease-specific and total mortality. International journal of
cancer Journal international du cancer 2015;137(10):2462-9.
55. Wakai K, Ohno Y, Obata K, et al. Prognostic significance of selected lifestyle factors in
urinary bladder cancer. Japanese Journal of Cancer Research 1993;84(12):1223-29.
56. Harris HR, Bergkvist L, Wolk A. Coffee and black tea consumption and breast cancer
mortality in a cohort of Swedish women. British journal of cancer 2012;107(5):874-
78.
57. Zell JA, McEligot AJ, Ziogas A, et al. Differential effects of wine consumption on
colorectal cancer outcomes based on family history of the disease. Nutrition and
cancer 2007;59(1):36-45.
58. Park SM, Lim MK, Shin SA, et al. Impact of prediagnosis smoking, alcohol, obesity, and
insulin resistance on survival in male cancer patients: National Health Insurance
Corporation Study. J Clin Oncol 2006;24(31):5017-24.
59. Pelser C, Arem H, Pfeiffer RM, et al. Prediagnostic lifestyle factors and survival after
colon and rectal cancer diagnosis in the National Institutes of Health (NIH)-AARP
Diet and Health Study. Cancer 2014;120(10):1540-7.
60. Phipps AI, Baron J, Newcomb PA. Prediagnostic smoking history, alcohol consumption,
and colorectal cancer survival: The Seattle Colon Cancer Family Registry. Cancer
2011;117(21):4948-57.
61. Lochhead P, Nishihara R, Qian ZR, et al. Postdiagnostic intake of one-carbon nutrients
and alcohol in relation to colorectal cancer survival. The American journal of clinical
nutrition 2015;102(5):1134-41.
62. Holm M, Olsen A, Christensen J, et al. Pre-diagnostic alcohol consumption and breast
cancer recurrence and mortality: results from a prospective cohort with a wide range
of variation in alcohol intake. International journal of cancer Journal international du
cancer 2013;132(3):686-94.
63. Barnett GC, Shah M, Redman K, et al. Risk factors for the incidence of breast cancer: do
they affect survival from the disease? J Clin Oncol 2008;26(20):3310-6.
64. Hellmann SS, Thygesen LC, Tolstrup JS, et al. Modifiable risk factors and survival in
women diagnosed with primary breast cancer: Results from a prospective cohort
study. European Journal of Cancer Prevention 2010;19(5):366-73.
65. Harris HR, Bergkvist L, Wolk A. Alcohol intake and mortality among women with
invasive breast cancer. British journal of cancer 2012;106(3):592-5.
66. Newcomb PA, Kampman E, Trentham-Dietz A, et al. Alcohol consumption before and
after breast cancer diagnosis: associations with survival from breast cancer,
cardiovascular disease, and other causes. Journal of clinical oncology : official journal
of the American Society of Clinical Oncology 2013;31(16):1939-46.
Page 38 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
39
67. Reding KW, Daling JR, Doody DR, et al. Effect of prediagnostic alcohol consumption on
survival after breast cancer in young women. Cancer epidemiology, biomarkers &
prevention : a publication of the American Association for Cancer Research,
cosponsored by the American Society of Preventive Oncology 2008;17(8):1988-96.
68. Lowry SJ, Kapphahn K, Chlebowski RT, et al. Alcohol use and breast cancer survival
among participants in the Women's Health Initiative. Cancer epidemiology,
biomarkers & prevention : a publication of the American Association for Cancer
Research, cosponsored by the American Society of Preventive Oncology 2016.
69. McDonald PA, Williams R, Dawkins F, et al. Breast cancer survival in African American
women: is alcohol consumption a prognostic indicator? Cancer causes & control :
CCC 2002;13(6):543-9.
70. Ewertz M, Gillanders S, Meyer L, et al. Survival of breast cancer patients in relation to
factors which affect the risk of developing breast cancer. International Journal of
Cancer 1991;49(4):526-30.
71. Borugian MJ, Sheps SB, Kim-Sing C, et al. Insulin, macronutrient intake, and physical
activity: are potential indicators of insulin resistance associated with mortality from
breast cancer? Cancer epidemiology, biomarkers & prevention : a publication of the
American Association for Cancer Research, cosponsored by the American Society of
Preventive Oncology 2004;13(7):1163-72.
72. Tominaga K, Andow J, Koyama Y, et al. Family environment, hobbies and habits as
psychosocial predictors of survival for surgically treated patients with breast cancer.
Jpn J Clin Oncol 1998;28(1):36-41.
73. Flatt SW, Thomson CA, Gold EB, et al. Low to moderate alcohol intake is not associated
with increased mortality after breast cancer. Cancer Epidemiol Biomarkers Prev
2010;19(3):681-8.
74. Hebert JR, Hurley TG, Ma Y. The effect of dietary exposures on recurrence and mortality
in early stage breast cancer. Breast cancer research and treatment 1998;51(1):17-28.
75. Kwan ML, Kushi LH, Weltzien E, et al. Alcohol consumption and breast cancer
recurrence and survival among women with early-stage breast cancer: The life after
cancer epidemiology study. Journal of Clinical Oncology 2010;28(29):4410-16.
76. Serur E, Fruchter RG, Maiman M, et al. Age, substance abuse, and survival of patients
with cervical carcinoma. Cancer 1995;75(10):2530-8.
77. Battaglioli T, Gorini G, Costantini AS, et al. Cigarette smoking and alcohol consumption
as determinants of survival in non-Hodgkin's lymphoma: A population-based study.
Annals of Oncology 2006;17(8):1283-89.
78. Talamini R, Polesel J, Spina M, et al. The impact of tobacco smoking and alcohol
drinking on survival of patients with non-Hodgkin lymphoma. Int J Cancer
2008;122(7):1624-9.
79. Geyer SM, Morton LM, Habermann TM, et al. Smoking, alcohol use, obesity, and overall
survival from non-Hodgkin lymphoma: A population-based study. Cancer
2010;116(12):2993-3000.
80. Han X, Zheng T, Foss FM, et al. Alcohol consumption and non-Hodgkin lymphoma
survival. Journal of cancer survivorship : research and practice 2010;4(2):101-09.
81. Ng K, Meyerhardt JA, Chan JA, et al. Multivitamin use is not associated with cancer
recurrence or survival in patients with stage III colon cancer: findings from CALGB
89803. J Clin Oncol 2010;28(28):4354-63.
82. Kwan ML, Greenlee H, Lee VS, et al. Multivitamin use and breast cancer outcomes in
women with early-stage breast cancer: The life after cancer epidemiology study.
Breast cancer research and treatment 2011;130(1):195-205.
Page 39 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
40
83. Greenlee H, Kwan ML, Kushi LH, et al. Antioxidant supplement use after breast cancer
diagnosis and mortality in the Life After Cancer Epidemiology (LACE) cohort.
Cancer 2012;118(8):2048-58.
84. Nechuta S, Lu W, Chen Z, et al. Vitamin supplement use during breast cancer treatment
and survival: a prospective cohort study. Cancer Epidemiol Biomarkers Prev
2011;20(2):262-71.
85. Poole EM, Shu X, Caan BJ, et al. Postdiagnosis supplement use and breast cancer
prognosis in the After Breast Cancer Pooling Project. Breast cancer research and
treatment 2013;139(2):529-37.
86. Swann R, Perkins KA, Velentzis LS, et al. The DietCompLyf study: a prospective cohort
study of breast cancer survival and phytoestrogen consumption. Maturitas
2013;75(3):232-40.
87. Wiseman M. The second World Cancer Research Fund/American Institute for Cancer
Research expert report. Food, nutrition, physical activity, and the prevention of
cancer: a global perspective. Proc Nutr Soc 2008;67(3):253-6.
88. Chang JL, Chen G, Ulrich CM, et al. DNA damage and repair: fruit and vegetable effects
in a feeding trial. Nutr Cancer 2010;62(3):329-35.
89. Langie SA, Wilms LC, Hamalainen S, et al. Modulation of nucleotide excision repair in
human lymphocytes by genetic and dietary factors. The British journal of nutrition
2010;103(4):490-501.
90. Wang X, Ouyang Y, Liu J, et al. Fruit and vegetable consumption and mortality from all
causes, cardiovascular disease, and cancer: systematic review and dose-response
meta-analysis of prospective cohort studies. BMJ (Clinical research ed)
2014;349:g4490.
91. Pandey KB, Rizvi SI. Plant polyphenols as dietary antioxidants in human health and
disease. Oxidative medicine and cellular longevity 2009;2(5):270-8.
92. Kushi LH, Byers T, Doyle C, et al. American Cancer Society Guidelines on Nutrition and
Physical Activity for cancer prevention: reducing the risk of cancer with healthy food
choices and physical activity. CA: a cancer journal for clinicians 2006;56(5):254-81;
quiz 313-4.
93. Al-Sader H, Abdul-Jabar H, Allawi Z, et al. Alcohol and breast cancer: the mechanisms
explained. Journal of clinical medicine research 2009;1(3):125-31.
94. Willett WC. Diet, nutrition, and avoidable cancer. Environ Health Perspect 1995;103
Suppl 8:165-70.
95. Mohr SB, Gorham ED, Kim J, et al. Could vitamin D sufficiency improve the survival of
colorectal cancer patients? The Journal of steroid biochemistry and molecular biology
2015;148:239-44.
96. Anderson AS, Steele R, Coyle J. Lifestyle issues for colorectal cancer survivors--
perceived needs, beliefs and opportunities. Supportive care in cancer : official journal
of the Multinational Association of Supportive Care in Cancer 2013;21(1):35-42.
97. Zebrack B. Information and service needs for young adult cancer patients. Supportive
care in cancer : official journal of the Multinational Association of Supportive Care in
Cancer 2008;16(12):1353-60.
98. Isenring E, Cross G, Kellett E, et al. Nutritional status and information needs of medical
oncology patients receiving treatment at an Australian public hospital. Nutrition and
cancer 2010;62(2):220-8.
99. Maley M, Warren BS, Devine CM. A second chance: meanings of body weight, diet, and
physical activity to women who have experienced cancer. Journal of nutrition
education and behavior 2013;45(3):232-9.
Page 40 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
41
100. van Veen MR, Beijer S, Adriaans AM, et al. Development of a Website Providing
Evidence-Based Information About Nutrition and Cancer: Fighting Fiction and
Supporting Facts Online. JMIR research protocols 2015;4(3):e110.
101. Farrell C, Brearley SG, Pilling M, et al. The impact of chemotherapy-related nausea on
patients' nutritional status, psychological distress and quality of life. Supportive care
in cancer : official journal of the Multinational Association of Supportive Care in
Cancer 2013;21(1):59-66.
Page 41 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
42
Table 1: Number of studies investigating the association between pre-diagnosis dietary pattern adherence and mortality in different populations of cancer survivors
Cancer type Low fat diet Western diet Prudent / Mediterranean diet
RCT Obs RCT Obs RCT Obs
Bladder cancer 0 0 0 0 0 0
Bowel cancer 0 0 0 1���� 0 1����
Breast cancer 0 0 0 2���� 0 2����
Cervical cancer 0 0 0 0 0 0
Hodgkin lymphoma 0 0 0 0 0 0
Kidney cancer 0 0 0 0 0 0
Laryngeal cancer 0 0 0 0 0 0
Malignant melanoma 0 0 0 0 0 0
Non-Hodgkin lymphoma 0 0 0 0 0 1����
Prostate cancer 0 0 0 0 0 0
Testicular cancer 0 0 0 0 0 0
Uterine cancer 0 0 0 0 0 0
RCT= randomised controlled trials; Obs= observational studies; �= increased risk of earlier mortality, �= decreased risk of earlier mortality, � =
no association with mortality
Page 42 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
43
Table 2: Number of studies investigating the association between post-diagnosis dietary pattern adherence and mortality in different populations of cancer survivors
Cancer type Low fat diet Western diet Prudent / Mediterranean diet
RCT Obs RCT Obs RCT Obs
Bladder cancer 0 0 0 0 0 0
Bowel cancer 0 0 0 2 (1���� , 1����) 0 2����
Breast cancer 0 2���� 0 2 (1���� , 1����) 0 2 (1���� , 1����)
Cervical cancer 0 0 0 0 0 0
Hodgkin lymphoma 0 0 0 0 0 0
Kidney cancer 0 0 0 0 0 0
Laryngeal cancer 0 0 0 0 0 0
Malignant melanoma 0 0 0 0 0 0
Non-Hodgkin lymphoma 0 0 0 0 0 0
Prostate cancer 0 0 0 1���� 0 2����
Testicular cancer 0 0 0 0 0 0
Uterine cancer 0 0 0 0 0 0
RCT= randomised controlled trials; Obs= observational studies; �= increased risk of earlier mortality, �= decreased risk of earlier mortality, � =
no association with mortality
Page 43 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
44
Table 3: Summary of studies investigating the association between pre-diagnosis dietary intake and mortality in different populations of cancer survivors
Cancer type
Fruit and
vegetables
Meat (red) Poultry Fish Dairy Coffee and tea Alcohol
Multivitamin
supplements
RCT Obs RCT Obs RCT Obs RCT Obs RCT Obs RCT Obs RCT Obs RCT Obs
Bladder cancer 0 1���� 0 0 0 0 0 0 0 0 0 1���� 0 2 (1����, 1���� ) 0 0
Bowel cancer 0 1���� 0 3 (1����,
2���� )
0 0 0 1���� 0 1���� 0 1���� 0 4 (1����, 3���� ) 0 1����
Breast cancer 0 4 (2����,
2���� )
0 0 0 0 0 0 0 0 0 0 0 10 (2����, 1����,
7���� )
0 2����
Cervical cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 1���� 0 0
Hodgkin lymphoma 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Kidney cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 1���� 0 0
Laryngeal cancer 0 1���� 0 1���� 0 1���� 0 1���� 0 0 0 0 0 1���� 0 0
Malignant melanoma 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Non-Hodgkin lymphoma 0 2 (1����,
1���� )
0 0 0 0 0 0 0 0 0 0 0 4 (3����, 1���� ) 0 0
Prostate cancer 0 0 0 0 0 0 0 1���� 0 0 0 0 0 1���� 0 0
Testicular cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Uterine cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Page 44 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
45
RCT= randomised controlled trials; Obs= observational studies; �= increased risk of earlier mortality, �= decreased risk of earlier mortality, � =
no association with mortality
Page 45 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
46
Table 4: Summary of studies investigating the association between post-diagnosis dietary intake and overall mortality / cancer-specific mortality in different populations of
cancer survivors
Cancer type
Fruit and
vegetables
Meat (red) Poultry Fish Dairy Coffee and tea Alcohol
Multivitamin
supplements
RCT Obs RCT Obs RCT Obs RCT Obs RCT Obs RCT Obs RCT Obs RCT Obs
Bladder cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 1���� 0 0
Bowel cancer 0 0 0 2���� 0 0 0 0 0 0 0 0 0 1���� 0 1����
Breast cancer 0 4���� 0 2���� 0 1���� 0 1���� 0 2 (1����,
1����)
0 1���� 0 11 (2����,
3����, 6���� )
0 5����
Cervical cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Hodgkin lymphoma 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Kidney cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Laryngeal cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Malignant melanoma 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Non-Hodgkin lymphoma 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Prostate cancer 0 0 0 1���� 0 1���� 0 0 0 1���� 0 0 0 0 0 0
Testicular cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Uterine cancer 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Page 46 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
47
RCT= randomised controlled trials; Obs= observational studies; �= increased risk of earlier mortality, �= decreased risk of earlier mortality, � =
no association with mortality
Page 47 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
48
Box 1: literature search for the pubmed database addressing the relationship between diet and mortality among bladder cancer survivors
(“urinary bladder neoplasms”[Mesh] OR “urinary bladder neoplasm*” OR “bladder cancer*” OR “bladder tumor*” OR “bladder tumour*”)
AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR
“survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier estimate”[Mesh] OR “kaplan-
meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet therapy”[Mesh] OR “food
habits”[Mesh] OR “diet fads”[Mesh] OR “diet, atherogenic”[Mesh] OR “diet, carbohydrate-restricted”[Mesh] OR “diet, cariogenic”[Mesh] OR
“diet, diabetic”[Mesh] OR “diet, fat-restricted”[Mesh] OR “diet, gluten-free”[Mesh] OR “diet, high-fat”[Mesh] OR “diet, Mediterranean”[Mesh]
OR “diet, paleolithic”[Mesh] OR “diet, protein-restricted”[Mesh] OR “diet, reducing”[Mesh] OR “diet, sodium-restricted”[Mesh] OR “diet,
vegetarian”[Mesh] OR “diet, macrobiotic”[Mesh] OR “diet, vegan”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric
restriction”[Mesh] OR “fasting”[Mesh] OR “ketogenic diet”[Mesh] OR “low calorie diet” OR “low carbohydrate diet” OR “low fat diet” OR
“low sodium diet” OR “low salt diet” OR “low protein diet” OR “south beach” OR “atkins” OR “fruitarian” OR “pescovegetarian” OR
“lactovegetarian” OR “lactoovovegetarian” OR “raw food” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary
protein*” OR “nutrition*” OR “food habits”[Mesh] OR “food and beverages”[Mesh] OR “food*” OR “beverage*” OR “dietary
supplements”[Mesh] OR “fluid intake” OR “drinking water”[Mesh] OR “alcoholic beverages*” OR “fruit”[Mesh] OR “fruit*” OR “citrus
fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “tea” OR “mate” OR “matcha” OR “green
tea” OR “teas, herbal” OR “black tea” OR “coffee” OR “caffeine” OR “beer” OR “alcohol” OR “wine” OR “liquor” OR “spirits” OR
“meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR
“fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR
“nut*” OR “vitamins”[Mesh] OR “vitamin*” OR “micronutrients”[Mesh] OR “multivitamin*” OR “minerals”[Mesh] OR “functional
food”[Mesh]) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR
"Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh] AND ("humans"[MeSH
Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 48 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Online supplement for manuscript ‘the impact of diet on mortality in cancer survivors: A systematic review of current
epidemiological literature’
SHJ Jochems et al., 29-09-2016
Page 49 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Table S1: Summary of findings on low-fat diets (with high intakes of fruits, vegetables and fibre) adherence and mortality in cancer survivors
Author (year) Study / country Timeframe /
exposure
assessment
Number
of cases /
sex (m/w)
Follow-up
period
(median
or
average)
Outcome
measures
Results
(HR or RR and 95% CI)
Adjustments
Breast cancer
Chlebowski (2006)
(Chlebowski et al.,
2006)
RCT - Women’s
Intervention
Nutrition Study
(WINS) / United
States
Post-
diagnosis /
interview
2437 w 5.0 years Overall
mortality
Women receiving the dietary
intervention versus control
group women
HR1= 0.89; 95% CI 0.65-1.21
nodal status, systemic
adjuvant therapy, ER status,
tumour size, mastectomy
Pierce (2007) (Pierce
et al., 2007)
RCT - Women's
Healthy Eating
and Living
(WHEL) study /
United States
Post-
diagnosis /
interview
3088 w 7.3 years Overall
mortality
Intervention group that was
intensively counselled to adopt
a dietary pattern very high in
vegetables, fruit, and fibre and
low in fat versus a comparison
group advised to follow the 5-
A-Day diet
anti-oestrogen use, bilateral
oophorectomy, age, BMI,
physical activity, energy
intake, tumour characteristics
(including hormone receptor
status), years from diagnosis
to study entry
Page 50 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Total vegetables and fruit:
HR1= 0.76; 95% CI 0.47-1.23
Total vegetables:
HR1= 1.19; 95% CI 0.74-1.90
Total fruit:
HR1= 0.76; 95% CI 0.48-1.19
First author’s name, year of publication, study name, country, timing dietary intake (pre- or post-diagnosis), number of cases, sex (m/w), follow-up period (median or
average), outcome measures, results; HR1= overall mortality, HR2= cancer-specific mortality, and adjustment factors
Page 51 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Table S2: Summary of findings on a prudent / Mediterranean diet adherence and mortality in cancer survivors
Author (year) Study / country Timeframe /
exposure
assessment
Number
of cases /
sex (m/w)
Follow-up
period
(median
or
average)
Outcome
measures
Results
(HR or RR and 95% CI)
Adjustments
Bowel cancer
Meyerhardt (2007)
(Meyerhardt et al.,
2007)
Cancer and
Leukemia Group
B (CALGB)
study / United
States
Post-
diagnosis /
FFQ
1009 m/w 5.3-5.6
years
Overall
mortality
HR1= 1.32; 95% CI 0.86-2.04 sex, age, depth of invasion
through bowel wall (T1-2 vs
T3-4), number of positive
lymph nodes (1-3 vs 4),
presence of clinical
perforation at time of surgery,
presence of bowel obstruction
at time of surgery, baseline
performance status (0 vs 1-2),
treatment group, weight
change between first and
second questionnaire, time-
Page 52 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
varying BMI, time-varying
physical activity level, time-
varying total calories
Zhu (2013) (Zhu et al.,
2013)
Familial CRC
registry in
Newfoundland /
Canada
Pre-diagnosis
/ FFQ
529 m/w 6.4 years Overall
mortality, colon
cancer-specific
mortality, rectal
cancer specific
mortality
HR1= 1.03; 95% CI 0.61-1.75
HR2a= 0.96; 95% CI 0.47-
1.96
HR2b= 1.00; 95% CI 0.42-
2.40
total energy intake, sex, age at
diagnosis, stage at diagnosis,
marital status, family history,
reported screening procedure,
reported chemo-radiotherapy
and microsatellite instability
status
Fung (2014) (Fung et
al., 2014)
Nurses' Health
Study (NHS) /
United States
Post-
diagnosis /
FFQ
1201 w 11.2 years Overall
mortality,
bowel cancer
mortality
HR1= 0.93; 95% CI 0.65-1.34
HR2= 0.67; 95% CI 0.37-1.22
age, physical activity, BMI,
weight change, cancer grade,
chemotherapy, smoking
status, energy intake, colon or
rectal cancer, stage of disease,
and date of colorectal cancer
diagnosis
Breast cancer
Kroenke (2005) Nurses' Health Pre-diagnosis 2619 w 9.0 years Overall Intake pre-diagnosis: age, time since diagnosis,
Page 53 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
(Kroenke et al., 2005) Study (NHS) /
United States
and Post-
diagnosis /
FFQ
mortality,
breast cancer
mortality
RR1= staticstically
nonsignificant (data not
shown)
RR2= staticstically
nonsignificant (data not
shown)
Intake post-diagnosis:
RR1= 0.78; 95% CI 0.54-1.12
RR2= 1.07; 95% CI 0.66-1.73
BMI, energy intake, smoking,
physical activity, diet missing,
age at menarche, oral
contraceptive use, menopausal
status and use of
postmenopausal hormone
therapy, age at menopause,
tamoxifen, chemotherapy,
stage at diagnosis
Kwan (2009) (Kwan et
al., 2009)
Life After Cancer
Epidemiology
(LACE) study /
United States
Post-
diagnosis /
FFQ
1901 w
4.2 years Overall
mortality,
breast cancer
mortality
HR1= 0.57; 95% CI 0.36-0.90
(p for trend = 0.02)
HR2= 0.79; 95% CI 0.43-1.43
age at diagnosis, total energy
intake, ethnicity, BMI, weight
change before diagnosis to
baseline, smoking status,
menopausal status at
diagnosis, stage, hormone
receptor status, treatment
Vrieling (2013)
(Vrieling et al., 2013)
Mammary
carcinoma Risk
Pre-diagnosis
/ FFQ
2522 w 5.5 years Overall
mortality,
HR1= 0.87; 95% CI 0.61-1.23
HR2= 0.89; 95% CI 0.59-1.35
age at diagnosis and study
centre, tumour size, nodal
Page 54 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
factor
Investigation
(MARIE) study /
Germany
breast cancer
mortality
status, metastases, tumour
grade, ERPR status,
radiotherapy, HRT use
diagnosis, mode of detection,
total energy intake
Non-Hodgkin lymphoma
Ollberding (2013)
(Ollberding et al.,
2013)
Nebraska
Lymphoma
Study Group /
United States
Pre-diagnosis
/ FFQ
301 m/w 8.2 years Overall
mortality
HR1= 1.0; 95% CI 0.5-1.8 age, sex, education, smoking
status, total energy intake
Prostate cancer
Kenfield (2014)
(Kenfield et al., 2014)
Health
Professionals
Follow-up Study
(HPFS) / United
States
Post-
diagnosis /
FFQ
4538 m in
case only
analysis
23.2 years Overall
mortality and
prostate cancer
mortality
HR1= 0.78; 95% CI 0.67-0.90
(p for trend = 0.0007)
HR2= 1.01; 95% CI 0.75-1.38
age at diagnosis, time period,
time diagnosis to FFQ,
energy, BMI, physical
activity, smoking status,
clinical stage, Gleason score,
treatment
Yang (2015) (Yang et
al., 2015a)
Physicians'
Health Study
Post-
diagnosis /
926 m 9.9 years Overall
mortality and
RR1= 0.64; 95% CI 0.44-0.93
(p for trend = 0.02)
age at diagnosis, total energy
intake, BMI, smoking status,
Page 55 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
(PHS) / United
States
FFQ prostate cancer
mortality
RR2= 0.46; 95% CI 0.17-1.24
exercise, Gleason score,
clinical stage, PSA level, time
between diagnosis and FFQ,
initial treatment, family
history prostate cancer
First author’s name, year of publication, study name, country, timing dietary intake (pre- or post-diagnosis), number of cases, sex (m/w), follow-up period (median or
average), outcome measures, results; HR1= overall mortality, HR2= cancer-specific mortality, HR2a= colon cancer specific mortality, HR2b= rectal cancer specific
mortality, and adjustment factors
Page 56 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Table S3: Summary of findings on a Western diet adherence and mortality in cancer survivors
Author (year) Study / country Timeframe /
exposure
assessment
Number
of cases /
sex (m/w)
Follow-up
(median
or
average)
Primary
outcome
Results
(HR or RR and 95% CI)
Adjustments
Bowel cancer
Meyerhardt (2007)
(Meyerhardt et al.,
2007)
Cancer and
Leukemia Group
B (CALGB)
study / United
States
Post-
diagnosis /
FFQ
1009 m/w 5.6 years Overall
mortality
HR1= 2.32; 95% CI 1.36-3.96
(p for trend = <0.001)
sex, age, depth of invasion
through bowel wall (T1-2 vs
T3-4), number of positive
lymph nodes (1-3 vs 4),
presence of clinical
perforation at time of surgery,
presence of bowel obstruction
at time of surgery, baseline
performance status (0 vs 1-2),
treatment group, weight
change between first and
second questionnaire, time-
varying BMI, time-varying
Page 57 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
physical activity level, and
time-varying total calories.
Zhu (2013) (Zhu et al.,
2013)
Familial CRC
registry in
Newfoundland /
Canada
Pre-diagnosis
/ FFQ
529 m/w 6.4 years Overall
mortality, colon
cancer
mortality, rectal
cancer
mortality
High processed meat pattern:
HR1= 1.53; 95% CI 0.85-2.74
HR2a= 2.13; 95% CI 1.03-
4.43 (p for trend = 0.40)
HR2b= 1.17; 95% CI 0.41-
3.36
High sugar pattern:
HR1= 1.27; 95% CI 0.72-2.25
HR2a= 1.16; 95% CI 0.54-
2.47
HR2b= 1.68; 95% CI 0.55-
5.08
total energy intake, sex, age at
diagnosis, stage at diagnosis,
marital status, family history,
reported screening procedure,
reported chemo- radiotherapy
and microsatellite instability
status
Fung (2014) (Fung et
al., 2014)
Nurses' Health
Study (NHS) /
United States
Post-
diagnosis /
FFQ
1201 w 11.2 years Overall
mortality,
bowel cancer
mortality
HR1= 1.32; 95% CI 0.89-1.97
HR2= 1.66; 95% CI 0.85-3.23
age, physical activity, BMI,
weight change, cancer grade,
chemotherapy, smoking
status, energy intake, colon or
rectal cancer, stage of disease,
Page 58 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
and date of colorectal cancer
diagnosis
Breast cancer
Kroenke (2005)
(Kroenke et al., 2005)
Nurses' Health
Study (NHS) /
United States
Pre-diagnosis
and Post-
diagnosis /
FFQ
2619 w 9.0 years Overall
mortality,
breast cancer
mortality
Intake pre-diagnosis:
RR1= 1.40; 95% CI 0.93-2.09
RR2= 1.01; 95% CI 0.59-1.72
Intake post-diagnosis:
RR1= 1.53; 95% CI 1.03-2.29
(p for trend = 0.08)
RR2= 1.01; 95% CI 0.60-1.70
Age, time since diagnosis,
BMI, energy intake, smoking,
physical activity, diet missing
in 1986/1990/1994/1998, age
at menarche, oral
contraceptive use, menopausal
status and use of
postmenopausal hormone
therapy, age at menopause,
tamoxifen, chemotherapy,
stage at diagnosis
Kwan (2009) (Kwan et
al., 2009)
Life After Cancer
Epidemiology
(LACE) study /
United States
Post-
diagnosis /
FFQ
1901 w
4.2 years Overall
mortality,
breast cancer
mortality
HR1= 1.53; 95% CI 0.93-2.54
HR2= 1.20; 95% CI 0.62-2.32
age at diagnosis, total energy
intake, ethnicity, BMI at
enrolment, weight change
from before diagnosis to
baseline, smoking,
Page 59 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
menopausal status at
diagnosis, stage, hormone
receptor status, treatment
Vrieling (2013)
(Vrieling et al., 2013)
Mammary
carcinoma Risk
factor
Investigation
(MARIE) study /
Germany
Pre-diagnosis
/ FFQ
2522 w 5.5 years Overall
mortality,
breast cancer
mortality
HR1= 1.34; 95% CI 0.93-1.94
HR2= 0.99; 95% CI 0.64-1.52
age at diagnosis and study
centre, tumour size, nodal
status, metastases, tumour
grade, ERPR status,
radiotherapy, HRT use at
diagnosis, mode of detection,
total energy intake
Prostate cancer
Yang (2015) (Yang et
al., 2015a)
Physicians'
Health Study
(PHS) / United
States
Post-
diagnosis /
FFQ
926 m 9.9 years Overall
mortality and
prostate cancer
mortality
RR1= 1.67; 95% CI 1.16–2.42
(p for trend = 0.01)
RR2= 2.53; 95% CI 1.00-6.42
(p for trend = 0.02)
age at diagnosis, total energy
intake, BMI, smoking status,
exercise, Gleason score,
clinical stage, PSA level, time
between diagnosis and FFQ,
initial treatment, family
history of prostate cancer
Page 60 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
First author’s name, year of publication, study name, country, timing dietary intake (pre- or post-diagnosis), number of cases, sex (m/w), follow-up period (median or
average), outcome measures, results; HR1= overall mortality, HR2= cancer-specific mortality, HR2a= colon cancer specific mortality, HR2b= rectal cancer specific
mortality, and adjustment factors
Page 61 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Table S4: Summary of findings on fruit and vegetables consumption and mortality in cancer survivors
Author (year) Study /
country
Timeframe /
exposure
assessment
Number of
cases / sex
(m/w)
Follow-up period
(median or
average)
Outcome
measures
Results
(HR or RR and 95% CI)
Adjustments
Bladder cancer
Tang (2010)
(Tang et al.,
2010)
Roswell Park
Cancer
Institute
(RPCI) /
United States
Pre-diagnosis
/ FFQ
239 m/w 8.0 years Overall
mortality,
cancer-specific
mortality
Total fruit:
HR1= 0.91; 95% CI 0.62-
1.33
HR2= 1.09; 95% CI 0.66-
1.81
Total vegetables:
HR1= 0.91; 95% CI 0.62-
1.36
HR2= 1.06; 95% CI 0.63-
1.78
Cruciferous vegetables:
HR1= 0.87; 95% CI 0.60-
1.26
HR2= 0.89; 95% CI 0.53-
age at diagnosis, total meat
intake, pack-years of
smoking, tumour stage,
radiation therapy
Page 62 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
1.48
Raw cruciferous
vegetables:
HR1= 0.73; 95% CI 0.50-
1.06
HR2= 0.73; 95% CI 0.44-
1.21
Bowel cancer
Dray (2003)
(Dray et al.,
2003)
Study in the
Cote d'Or area
/ France
Pre-diagnosis /
FFQ
148 m/w
5-year survival Overall
mortality
Total fruit:
RR1= 0.84; 95% CI 0.37-
1.88
Total vegetables:
RR1= 1.09; 95% CI 0.49-
2.45
age, sex, tumour stage,
tumour location, energy
intake
Breast cancer
Dal Maso
(2008) (Dal
Maso et al.,
2008)
Six Italian
Regions
Follow-up
Study / Italy
Pre-diagnosis /
FFQ
1453 w 12.6 years Overall
mortality,
cancer-specific
mortality
Total fruit and vegetables:
HR1= 1.27; 95% CI 1.00–
1.61 (p for trend = 0.04)
HR2= 1.26; 95% CI 0.96–
region, age at diagnosis,
year of diagnosis, TNM
stage, receptor status
Page 63 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
1.64
!NOTE high versus low
intake)
Buck (2011)
(Buck et al.,
2011)
Study in
Karlsruhe
/ Germany
Pre-diagnosis /
FFQ
2653 w
6.1 years Overall
mortality,
cancer-specific
mortality
Total fruit:
HR1= 0.84; 95% CI 0.61-
1.16
HR2= 0.86; 95% CI 0.59-
1.25
Total vegetables:
HR1= 1.09; 95% CI 0.80-
1.48
HR2= 1.01; 95% CI 0.70-
1.46
tumour size, nodal
status, metastasis, grade,
oestrogen and
progesterone receptor
status, breast cancer
detection type, diabetes,
HRT use at diagnosis,
study centre, energy
intake, age at diagnosis
McEligot (2006)
(McEligot et al.,
2006)
Orange County
California
Study / United
States
Pre-diagnosis /
FFQ
516 w 6.7 years Overall
mortality
Total fruit:
HR1= 0.63; 95% CI 0.38-
1.05
Total vegetables:
HR1= 0.57; 95% CI 0.35–
0.94 (p for trend = 0.02)
tumour stage, age at
diagnosis, BMI, parity,
HRT, alcohol
intake, multivitamins,
energy intake
Page 64 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Saxe (1999)
(Saxe et al.,
1999)
Michigan
University
Follow-up
Study / United
States
Pre- diagnosis /
FFQ
149 w 5.0 years Overall
mortality
Total fruit:
HR1= 1.06; 95% CI 0.69-
1.63
Total vegetables:
HR2= 0.97; 95% CI 0.70-
1.35
tumour stage, energy
intake
Beasley (2011)
(Beasley et al.,
2011)
Collaborative
Women’s
Longevity
Study / United
States
Post-diagnosis /
FFQ
4441 w
5.5 years Overall
mortality,
cancer-specific
mortality
Total fruit:
HR1= 1.38; 95% CI 0.88-
2.17
HR2= 1.39; 95% CI 0.64-
2.99
Total vegetables:
HR1= 1.44; 95% CI 0.91-
2.27
HR2= 0.96; 95% CI 0.38-
2.45
age, state of residence,
menopausal status,
smoking, breast cancer
stage, alcohol, history of
hormone replacement
therapy at diagnosis,
interval between diagnosis
and diet assessment, and
energy intake, breast
cancer treatment, body
mass at follow-up
Hebert (1998)
(Hebert et al.,
Memorial
Sloan-
Post-diagnosis /
FFQ
472 w 8 – 10 years Cancer-
specific
Total vegetables:
RR2= 0.47; p=0.09 (95%
tumour stage, oestrogen
receptor status, age, BMI,
Page 65 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
1998) Kettering
Cancer
Center Follow-
up
Study /
United States
mortality CI not shown)
menopausal status, meat,
beer, butter/ margarine/
lard
Holmes (1999)
(Holmes et al.,
1999)
Nurses' Health
Study (NHS) /
United States
Post-diagnosis /
FFQ
1982 w
13.1 years Overall
mortality
Total vegetables:
RR1= 0.81; 95% CI 0.59–
1.11
Total fruit:
RR1= 1.07; 95% CI 0.77–
1.49
age, diet interval, calendar
year of diagnosis, body
mass index, oral
contraceptive use,
menopausal status,
postmenopausal hormone
use, smoking, age at first
birth and parity, number of
metastatic lymph nodes,
tumor size, calories
Nechuta (2013)
(Nechuta et al.,
2013)
After Breast
Cancer
Pooling
Post-diagnosis /
FFQ
11390 w 9.0 years Overall
mortality
Cruciferous vegetable
intake:
HR1= 0.99; 95% CI 0.86-
age at diagnosis, oestrogen
and progesterone receptor
status, TNM stage,
Page 66 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Project / China
and United
States
1.13 surgery, chemotherapy,
radiotherapy, hormonal
therapy, smoking, BMI,
exercise, menopausal
status, ethnicity, education
Non-Hodgkin lymphoma
Han (2010)
(Han et al.,
2010a)
Yale
Connecticut
Tumor
Registry New
York (CTR) /
United States
Pre-diagnosis /
FFQ
568 w 7.7 years Overall
mortality,
cancer-specific
mortality
Total fruit and vegetables:
HR1= 0.68; 95% CI 0.49-
0.95 (only two categories)
HR2= 0.70; 95% CI 0.45-
1.10
Total fruit:
HR1= 0.91; 95% CI 0.70-
1.18
HR2= 1.04; 95% CI 0.74-
1.45
Total vegetables:
HR1= 0.58; 95% CI 0.38-
0.89 (only two categories)
age, education, stage, B-
symptom, initial treatment,
total energy intake
Page 67 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
HR2= 0.58; 95% CI 0.33-
1.03
Total cruciferous
vegetables:
HR1= 0.58; 95% CI 0.33–
1.03
HR2= 0.75; 95% CI 0.49–
1.14
Ollberding
(2013)
(Ollberding et
al., 2013)
Nebraska
Lymphoma
Study Group /
United States
Pre-diagnosis /
FFQ
301 m/w 8.2 years Overall
mortality
Total fruit and vegetables:
HR1= 1.1; 95% CI 0.6-2.1
Total vegetables:
HR1= 0.9; 95% CI 0.5-1.5
Total fruit:
HR1= 0.9; 95% CI 0.5-1.6
Citrus fruit:
HR1= 1.2; 95% CI 0.7-2.1
age, sex, education,
smoking status, total
energy intake
Laryngeal cancer
Crosignani
(1996)
Lombardy
Cancer
Pre-diagnosis /
FFQ
218 m 8.0 years Overall
mortality
Citrus fruit:
HR1= 0.76; 95% CI 0.49-
age at diagnosis, clinical
stage, occurrence of new
Page 68 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
(Crosignani et
al., 1996)
Registry
(LCR) / Italy
1.19
Other fruit:
HR1= 0.65; 95% CI 0.39-
1.07
Total vegetables:
HR1= 0.57; 95% CI 0.35-
0.94 (x2 for trend = 4.79)
primaries
First author’s name, year of publication, study name, country, timing dietary intake (pre- or post-diagnosis), number of cases, sex (m/w), follow-up period (median or
average), outcome measures, results; HR1= overall mortality, HR2= cancer-specific mortality, and adjustment factors
Page 69 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Table S5: Summary of findings on meat and fish consumption and mortality in cancer survivors
Author (year) Study /
country
Timeframe /
exposure
assessment
Number of
cases / sex
(m/w)
Follow-up period
(median or
average)
Outcome
measures
Results
(HR or RR and 95% CI)
Adjustments
Bowel cancer
McCullough
(2013)
(McCullough et
al., 2013)
Cancer
Prevention
Study II
Nutrition
Cohort / United
States
Pre- and post-
diagnosis /
FFQ
2315 m/w
7.5 years Overall
mortality,
cancer-specific
mortality
Red and processed meat
pre-diagnosis:
RR1= 1.29; 95% CI 1.05-
1.59 (p for trend = 0.03)
RR2= 1.09; 95% CI 0.79-
1.51
Red and processed meat
post-diagnosis:
RR1= 1.02; 95% CI 0.76-
1.38
RR2= 1.28; 95% CI 0.74-
2.21
Red and processed meat
pre- and post-diagnosis
pre-diagnosis: age at
diagnosis, sex, tumour
stage at diagnosis, 1992
pre-diagnostic energy
intake, BMI in 1992,
history of diabetes, and
history of myocardial
infarction.
post-diagnosis: age at
diagnosis, sex, tumour
stage at diagnosis, and
post-diagnostic energy
intake (sex-specific
quartiles), weight change
Page 70 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
combined �
Low Pre-diagnosis /
High Post-diagnosis:
RR1= 1.25; 95% CI 0.93-
1.67
RR2= 0.96; 95% CI 0.55-
1.66
High Pre-diagnosis /
Low Post-diagnosis:
RR1= 1.37; 95% CI 1.02-
1.85 (only two categories)
RR2= 1.43; 95% CI 0.80-
2.57
High Pre-diagnosis /
High Post-diagnosis:
RR1= 1.28; 95% CI 0.98-
1.67
RR2= 1.79; 95% CI 1.11-
2.89 (only two categories)
between 1992 pre-
diagnostic, post-
diagnostic questionnaires,
and 1992 pre-diagnostic
meat intake.
Combined pre- and post-
diagnosis: age at
diagnosis, sex, tumour
stage at diagnosis, 1992
pre-diagnostic energy
intake, and post-
diagnostic energy intake
Page 71 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Dray (2003)
(Dray et al.,
2003)
Study in the
Côte d’Or area
/ France
Pre-diagnosis /
FFQ
148 m/w 5- and 10-years
survival
Overall mortality Total meat:
RR1= 1.64; 95% CI 0.75-
3.58
Total fish:
RR1= 1.24; 95% CI 0.58-
2.65
age, sex, tumour stage,
tumour location, and
energy intake
Zell (2010)
(Zell et al.,
2010)
California
Teachers Study
(CTS) / United
States
Pre-diagnosis /
FFQ
499 w 10 years survival Overall
mortality,
cancer-specific
mortality
Total meat (non-NSAID
use versus NSAID use):
HR1= 1.03; 95% CI 0.43–
2.45
HR2= 1.08; 95% CI 0.36–
3.24
age at baseline
questionnaire and stage,
with adjustment for family
history of CRC in a 1st
degree relative, site (colon
or rectum), treatment with
surgery, and total daily
energy intake
Carr (2016)
(Carr et al.,
2016)
DACHS
(Darmkrebs:
Chancen der
Verhutung
durch
Post-diagnosis /
FFQ
3122 m/w 4.8 years Overall
mortality,
cancer-specific
mortality
Total red and processed
meat intake:
HR1= 0.85; 95% CI 0.67-
1.09
HR2= 0.83; 95% CI 0.61-
age at diagnosis, sex,
cancer stage,
chemotherapy, surgery,
BMI, physical activity,
diabetes, stroke, heart
Page 72 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Screening)
study /
Germany
1.14 failure, myocardial
infarction, dairy intake,
wholegrain intake, time
between diagnosis and
interview, and a time-
dependent effect of
chemotherapy
Breast cancer
Holmes (1999)
(Holmes et al.,
1999)
Nurses' Health
Study (NHS) /
United States
Post-diagnosis /
FFQ
1982 w
13.1 years Overall mortality Total red meat:
RR1= 1.06; 95% CI 0.76–
1.49
Total poultry:
RR1= 0.70; 95% CI 0.50–
0.97 (p for trend = 0.02)
Total fish:
RR1= 0.80; 95% CI 0.60–
1.07
quantiles of nutrient or
food intake prior to
diagnosis, previous diet
interval, age, diet interval,
calendar year of
diagnosis, body mass
index, oral contraceptive
use, menopausal status,
postmenopausal hormone
use, smoking, age at first
birth and parity, number
Page 73 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
of metastatic lymph
nodes, tumor size, calories
Hebert (1998)
(Hebert et al.,
1998)
Memorial
Sloan-
Kettering
Cancer Center
Follow-up
Study /
United States
Post-diagnosis /
FFQ
472 w 8 – 10 years Cancer-specific
mortality
Red meat:
RR2= 1.43; 95% CI 0.74-
2.79
tumour stage, oestrogen
receptor status, age, BMI,
menopausal status, meat,
beer, butter/ margarine/
lard
Laryngeal cancer
Crosignani
(1996)
(Crosignani et
al., 1996)
Lombardy
Cancer
Registry (LCR)
/ Italy
Pre-diagnosis /
FFQ
218 m 8.0 years Overall mortality Total meat (beef/veal):
HR1= 0.50; 95% CI 0.30-
0.83 (x2 for trend = 7.39)
Total poultry:
HR1= 0.90; 95% CI 0.55-
1.46
Total fish:
HR1= 0.91; 95% CI 0.59-
1.39
age at diagnosis, clinical
stage and occurrence of
new primaries
Page 74 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Prostate cancer
Richman
(2011)
(Richman et
al., 2011)
Health
Professionals
Follow-up
Study (HPFS) /
United States
Post-diagnosis /
FFQ
3127 m 8.4 years Cancer-specific
mortality
Total red meat:
HR2= 1.13; 95% CI 0.60 -
2.10
Total unprocessed red
meat:
HR2= 0.94; 95% CI 0.52-
1.70
Total processed red meat:
HR2= 1.45; 95% CI 0.73-
2.87
Total poultry:
HR2= 1.69; 95% CI 0.96-
2.99
age at diagnosis, time
since diagnosis, energy
intake, Gleason sum,
clinical T-stage, primary
treatment, BMI, smoking,
vigorous activity, and pre-
diagnostic intake of the
exposure of interest.
Unprocessed red meat was
adjusted for processed red
meat and all red meat and
poultry categories
were adjusted for eggs
Chavarro
(2008)
(Chavarro et
al., 2008)
Physician’s
Health Study
(PHS) / United
States
Pre-diagnosis /
FFQ
2161 m
19.0 years Cancer-specific
mortality
Total fish:
HR2= 0.52; 95% CI 0.30-
0.91 (p for trend = 0.05)
age at prostate cancer
diagnosis, BMI, physical
activity, alcohol use,
tomato and dairy
products, smoking,
Page 75 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
ethnicity, multivitamin
and vitamin E
supplements, random
assignment to aspirin or
beta-carotene, tumour
stage, grade at diagnosis,
clinical presentation of
case
First author’s name, year of publication, study name, country, timing dietary intake (pre- or post-diagnosis), number of cases, sex (m/w), follow-up period (median or
average), outcome measures, results; HR1= overall mortality, HR2= cancer-specific mortality, and adjustment factors
Page 76 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Table S6: Summary of findings on dairy consumption and mortality in cancer survivors
Author (year) Study /
country
Timeframe /
exposure
assessment
Number of
cases / sex
(m/w)
Follow-up period
(median or
average)
Outcome
measures
Results
(HR or RR and 95% CI)
Adjustments
Bowel cancer
Dray (2003)
(Dray et al.,
2003)
Study in the
Côte d’Or area
/ France
Pre-diagnosis /
FFQ
148 m/w 5- or 10-y survival Overall
mortality
Total dairy:
RR1= 0.63; 95% CI 0.30-
1.33
age, sex, tumour stage,
tumour location, energy
intake
Breast cancer
Holmes (1999)
(Holmes et al.,
1999)
Nurses' Health
Study (NHS) /
United States
Post-diagnosis /
FFQ
1982 w
13.1 years Overall
mortality
Total dairy:
RR1= 0.72; 95% CI 0.52–
1.00 (p for trend = 0.04)
quantiles of nutrient or
food intake prior to
diagnosis, previous diet
interval, age, diet interval,
calendar year of diagnosis,
body mass index, oral
contraceptive use,
menopausal status,
postmenopausal hormone
use, smoking, age at first
Page 77 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
birth and parity, number of
metastatic lymph nodes,
tumor size, calories
Kroenke (2013)
(Kroenke et al.,
2013)
Life After
Cancer
Epidemiology
(LACE) study
/ United States
Post-diagnosis /
FFQ
1893 w 11.8 years Overall
mortality,
cancer-specific
mortality
Total dairy:
HR1= 1.39; 95% CI 1.02-
1.90 (p for trend = 0.05)
HR2= 1.26; 95% CI 0.81-
1.95
age, time between
diagnosis and dietary
assessment, high- and low-
fat dairy intake, race,
education, cancer stage at
diagnosis, tumour size,
human epidermal growth
receptor 2, nodal and
oestrogen receptor status,
chemotherapy, radiation,
tamoxifen, comorbidity,
menopausal status, BMI,
physical activity, energy
intake, alcohol intake, red
meat intake, fiber intake,
fruit intake
Page 78 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Prostate cancer
Yang (2015)
(Yang et al.,
2015b)
Physician’s
Health Study
(PHS) / United
States
Post-diagnosis /
FFQ
926 m 10 years Overall
mortality and
cancer-specific
mortality
Total dairy:
HR1 = 1.76; 95% CI 1.21-
2.55 (p for trend <0.001)
HR2 = 2.41; 95% CI 0.96-
6.02
High-fat dairy:
HR1= 1.22; 95% CI 1.08-
1.38 (p = 0.002)
HR2= 1.30; 95% CI 0.97-
1.73
Low-fat dairy:
HR1= 1.17; 95% CI 1.05-
1.29 (p = 0.003)
HR2= 1.16; 95% CI 0.88-
1.53
age at diagnosis, total
energy intake, BMI,
smoking status, exercise,
Gleason score, clinical
stage, prostate-specific
antigen level, time interval
between diagnosis and
FFQ completion,
initial treatment after
diagnosis, family history
of prostate cancer, and
indicators for prudent
dietary pattern and western
dietary pattern after
excluding dairy products
First author’s name, year of publication, study name, country, timing dietary intake (pre- or post-diagnosis), number of cases, sex (m/w), follow-up period (median or
average), outcome measures, results; HR1= overall mortality, HR2= cancer-specific mortality, and adjustment factors
Page 79 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Table S7: Summary of findings on coffee and tea consumption and mortality in cancer survivors
Author (year) Study /
country
Timeframe /
exposure
assessment
Number of
cases / sex
(m/w)
Follow-up period
(median or
average)
Outcome
measures
Results
(HR or RR and 95% CI)
Adjustments
Bladder cancer
Wakai (1993)
(Wakai et al.,
1993)
Study in
metropolitan
Nagoya / Japan
Pre-diagnosis /
FFQ
203 m 2.5 years Overall
mortality
Coffee:
HR1= 0.88; 95% CI 0.49-
1.59
Black tea:
HR1= 0.77; 95% CI 0.44-
1.33
Green tea:
HR1= 0.62; 95% CI 0.22-
1.74
Matcha:
HR1= 1.36; 95% CI 0.75-
2.44
age, histological type and
grade, stage, distant
metastasis
Bowel cancer
Dray (2003) Study in the Pre-diagnosis / 148 m/w 5- or 10-y survival Overall Coffee and tea: age, sex, tumour stage,
Page 80 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
(Dray et al.,
2003)
Côte d’Or area
/ France
FFQ mortality RR1= 1.46; 95% CI 0.64-
3.22
tumour location, energy
intake
Breast cancer
Harris (2012)
(Harris et al.,
2012b)
Swedish
Mammography
Cohort
(SMC) /
Sweden
Post-diagnosis /
FFQ
3243 w 8.8 years Overall
mortality,
cancer-
specific
mortality
Coffee:
HR1= 1.12; 95% CI 0.84–
1.51
HR2= 1.14; 95% CI 0.71–
1.83
Tea:
HR1= 0.94; 95% CI 0.72–
1.23
HR2= 1.02; 95% CI 0.67–
1.55
age, energy intake,
education, marital status,
menopausal status, BMI,
year of diagnosis, stage of
disease, radiotherapy,
chemotherapy, hormonal
therapy
First author’s name, year of publication, study name, country, timing dietary intake (pre- or post-diagnosis), number of cases, sex (m/w), follow-up period (median or
average), outcome measures, results; HR1= overall mortality, HR2= cancer-specific mortality, and adjustment factors
Page 81 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Table S8: Summary of findings on alcohol consumption and mortality in cancer survivors
Author (year) Study /
country
Timeframe /
exposure
assessment
Number of
cases / sex
(m/w)
Follow-up period
(median or
average)
Outcome
measures
Results
(HR or RR and 95% CI)
Adjustments
Bladder cancer
Wakai (1993)
(Wakai et al.,
1993)
Study in
metropolitan
Nagoya / Japan
Pre- and post-
diagnosis / FFQ
203 m 2.5 years Overall
mortality
Alcohol pre-diagnosis:
HR1= 0.46; 95% CI 0.26-
0.79
Alcohol post-diagnosis:
HR1= 0.43, 95% CI 0.24-
0.77
age, histological type and
grade, stage, distant
metastasis
Park (2006)
(Park et al.,
2006)
National
Health
Insurance
Corporation
Study
(NHICS) /
Korea
Pre-diagnosis /
FFQ
432 m 4.69 years Overall
mortality
HR1= 0.46; 95% CI 0.20-
1.02
age, alcohol consumption,
BMI, fasting serum
glucose level,
cholesterol level, physical
activity, food preference,
blood pressure, other
comorbidities
Bowel cancer
Page 82 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Park (2006)
(Park et al.,
2006)
National
Health
Insurance
Corporation
Study
(NHICS) /
Korea
Pre-diagnosis /
FFQ
1882 m 3.8 years Overall
mortality
HR1= 0.92; 95% CI 0.72-
1.19
age, alcohol consumption,
BMI, fasting serum
glucose level,
cholesterol level, physical
activity, food preference,
blood pressure, other
comorbidities
Zell (2007) (Zell
et al., 2007)
University of
California
Irvine CRC
gene
environment
study (UCI) /
United States
Pre-diagnosis /
FFQ
499 m/w
7 – 9 years Overall
mortality
Wine:
HR1= 0.50; 95% CI 0.25-
0.99 (only two categories)
for familial CRC
survivors, and HR1= 0.89;
95% CI 0.59–1.33 for
sporadic CRC survivors
Beer:
HR1= 1.07; 95% CI 0.50-
2.29 for familial CRC
survivors, and HR1= 1.01;
95% CI 0.65–1.57 for
age, stage, treatment,
clinically relevant factors
Page 83 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
sporadic CRC survivors
Liquor:
HR1= 1.13; 95% CI 0.56–
2.29 for familial CRC
survivors, and HR1= 0.79;
95% CI 0.51–1.24 for
sporadic CRC survivors
Phipps (2011)
(Phipps et al.,
2011)
North Central
Cancer
Treatment
Group / United
States
Pre-diagnosis /
structured
telephone
interview
2264 m/w Overall
mortality,
cancer-specific
mortality
Total alcohol:
HR1= 1.02; 95% CI 0.82-
1.27
HR2= 1.02; 95% CI 0.78-
1.32
Wine:
HR1= 0.97; 95% CI 0.69-
1.36
HR2= 1.08; 95% CI 0.73-
1.60
Beer:
HR1= 0.95; 95% CI 0.72-
age at diagnosis, time from
diagnosis to interview,
history of preventive
colorectal cancer
screening, sex, education
Page 84 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
1.25
HR2= 0.96; 95% CI 0.69-
1.34
Liquor/spirits:
HR1= 1.18; 95% CI 0.93-
1.50
HR2= 1.20; 95% CI 0.89-
1.62
Pelser (2014)
(Pelser et al.,
2014)
NIH-AARP
diet and health
Study / United
States
Pre-diagnosis /
FFQ
4213 m/w
colon cancer
cases and 1514
m/w rectal
cancer cases
5 years (maximum) Overall
mortality,
cancer-specific
mortality
Colon cancer cases:
HR1= 0.92; 95% CI 0.77–
1.11
HR2= 1.06; 95% CI 0.84–
1.33
Rectal cancer cases:
HR1= 1.03; 95% CI 0.76–
1.41
HR2= 0.97; 95% CI 0.66–
1.41
lag time, sex, education,
family history of colon
cancer, cancer stage, first
course of treatment,
quintiles of HEI-2005
scores, BMI, physical
activity, alcohol and
smoking history
Lochhead Nurses’ Health Post-diagnosis / 1550 m/w 14.9 y Overall HR1= 0.91; 95% CI 0.72- post-diagnostic alcohol
Page 85 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
(2015)
(Lochhead et
al., 2015)
Study (NHS)
and the Health
Professionals
Follow-Up
Study (HPFS) /
United States
FFQ (1063 women
from NHS and
487 men from
HPFS)
mortality,
cancer-specific
mortality
1.16
HR2= 0.53; 95% CI 0.28-
0.98 (p for trend = 0.33)
consumption, pre-
diagnostic alcohol
consumption,
age at diagnosis; year of
diagnosis, BMI, family
history of colorectal
cancer in any first-degree
relative, post-diagnostic
aspirin use, post-
diagnostic multivitamin
use, post-diagnostic
smoking status, post-
diagnostic physical
activity, post-diagnostic
folate, vitamin B-12,
methionine, and vitamin
B-6 intakes, tumour
location, tumour
differentiation, time from
Page 86 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
diagnosis to FFQ
(backward elimination p=
0.1 was used to select
variables in the final
model)
Breast cancer
Kwan (2010)
(Kwan et al.,
2010)
Life After
Cancer
Epidemiology
(LACE) study
/ United States
Post-diagnosis /
FFQ
1897 w 7.4 years Overall
mortality,
cancer-specific
mortality
HR1= 1.19; 95% CI 0.87-
1.62
HR2= 1.51; 95% CI 1.00-
2.29
(p for trend = 0.05)
age at diagnosis, BMI,
folate intake, tumour
stage, receptor status,
tamoxifen use, treatment,
nodal status
Holm (2013)
(Holm et al.,
2013)
Diet, Cancer
and Health
(DCH) study /
Denmark
Pre-diagnosis /
FFQ
1052 w 6.0 years Cancer-
specific
mortality
HR2= 1.10; 95% CI 0.67–
1.82
tumour size, lymph node
status, receptor status and
grade. BMI, smoking,
menopausal status, HRT
use, education level,
physical activity, total
folate intake
Page 87 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
McDonald
(2002)
(McDonald et
al., 2002)
Grampian
University
Hospitals
Follow-up
Study / United
States
Pre- diagnosis /
FFQ
125 w 5.4 years Cancer-
specific
mortality
HR2= 2.7, 95% CI 1.3–5.8
(only two categories)
tumour stage,
radiotherapy, smoking,
alcohol intake
Saxe (1999)
(Saxe et al.,
1999)
Michigan
University
Follow-up
Study / United
States
Pre-diagnosis /
FFQ
149 w 5 years Overall
mortality
HR1= 0.97; 95% CI 0.70-
1.35
tumour stage, energy
intake
Reding (2008)
(Reding et al.,
2008)
Fred
Hutchinson
Cancer
Research
Center / United
States
Pre-diagnosis /
FFQ
1286 w 9.0 years Overall
mortality
HR1= 0.7; 95% CI 0.5-
0.9 (p not shown)
age, diagnosis year, and
mammography
Dal Maso
(2008) (Dal
Six Italian
Regions
Pre-diagnosis /
FFQ
1453 w 12.6 years Overall
mortality,
HR1= 0.99; 95% CI 0.80-
1.22
age at diagnosis, year of
diagnosis, TNM stage,
Page 88 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Maso et al.,
2008)
Follow-Up
Study / Italy
cancer-specific
mortality
HR2= 1.03; 95% CI 0.81-
1.31
receptor status, region in
Italy
Barnett (2008)
(Barnett et al.,
2008)
Studies of
Epidemiology
and Risk
Factors in
Cancer
Heredity
Breast Cancer /
United
Kingdom
Pre-diagnosis
and Post-
diagnosis / FFQ
4560 w
6.82 years Overall
mortality
Pre-diagnosis (at age 30):
HR1= 0.96; 95% CI 0.79 -
1.17
Post-diagnosis:
HR1= 0.84; 95% CI 0.63-
1.12
age at diagnosis, tumor
stage, tumor grade
Hellmann
(2010)
(Hellmann et
al., 2010)
Copenhagen
City
Heart Study /
Denmark
Pre-diagnosis /
FFQ
528 w
7.8 years Overall
mortality
HR1= 1.06; 95% CI 0.68–
1.66
age, smoking, physical
activity , alcohol intake,
hormonal therapy, tumour
stage, menopausal status,
parity, education,
treatment
Harris (2012)
(Harris et al.,
Swedish
Mammography
Pre-diagnosis /
FFQ
3146 w
20 – 22 years Overall
mortality,
HR1= 1.03; 95% CI 0.71 –
1.51
HR2= 1.36; 95% CI 0.82 –
age, energy intake,
education, marital status,
Page 89 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
2012a) Cohort /
Sweden
cancer-specific
mortality
2.26 menopausal status, BMI,
year of diagnosis, stage of
disease, radiotherapy,
chemotherapy, hormonal
therapy
Ewertz (1991)
(Ewertz et al.,
1991)
Danish Breast
Cancer
Cooperative
Group /
Denmark
Post-diagnosis /
FFQ
2445 w 7.0 years Overall
mortality
HR1= 1.26; 95% CI 0.9-
1.74
age, tumour size, nodal
status, tumour grade, skin
invasion, area of residence
Tominaga
(1998)
(Tominaga et
al., 1998)
Tochigi
Cancer
Center
Hospital /
Japan
Post-diagnosis /
FFQ
398 w 9 years (maximum) Overall
mortality
HR1= 0.1; 95% CI 0.01-
0.72 (only two categories) age at diagnosis, TNM
stage, curability
Holmes (1999)
(Holmes et al.,
1999)
Nurses' Health
Study (NHS) /
United States
Post-diagnosis /
FFQ
1982 w
13.1 years Overall
mortality
RR1= 0.92; 95% CI 0.66–
1.27
age, diet interval, calendar
year of diagnosis, body
mass index, oral
contraceptive use,
menopausal status,
postmenopausal hormone
Page 90 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
use, smoking, age at first
birth and parity, number of
metastatic lymph nodes,
and tumor size, caloric
intake
Flatt S
(2010) (Flatt et
al., 2010)
Women’s
Healthy
Eating and
Living
Study /
United States
Post-diagnosis /
FFQ
3088 w
7.3
years
Overall
mortality,
cancer specific
mortality
HR1= 0.69; 95% CI 0.49-
0.97 (p trend not shown)
HR2= 0.70, 95% C.I. 0.48-
1.02
tumour grade, tumour
stage, years between
diagnosis and study entry,
alcohol intake, education,
ethnicity, smoking, parity,
physical activity
Beasley JM
(2011) (Beasley
et al., 2011)
Collaborative
Women’s
Longevity
Study / United
States
Post-diagnosis /
FFQ
4441 w 5.5
years
Overall
mortality,
cancer specific
mortality
HR1= 0.78; 95% CI 0.60-
1.01
HR2= 1.27; 95% CI 0.76–
2.14
age, residence,
menopausal status,
smoking, stage, alcohol
intake, hormonal therapy,
interval between diagnosis
and baseline interview,
BMI, physical activity,
breast cancer treatment,
energy intake
Page 91 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Borugian M
(2004)
(Borugian et
al., 2004)
Vancouver
Cancer
Centre of the
British
Columbia
Cancer
Agency /
Canada
Post-diagnosis /
FFQ
603 w 10.0 years
Cancer-
specific
mortality
HR2= 0.99; 95% CI 0.94-
1.04
age, total caloric intake,
stage at diagnosis
Hebert (1998)
(Hebert et al.,
1998)
Memorial
Sloan-
Kettering
Cancer
Center Follow-
up
Study /
United States
Post-diagnosis /
FFQ
472 w 8 – 10 years Cancer-
specific
mortality
Beer:
RR2= 1.58; 95% CI 1.00-
2.78 (continuous variable
p = 0.05)
tumour stage, oestrogen
receptor status, age, BMI,
menopausal status, meat,
beer, butter/ margarine/
lard
Newcomb
(2013)
(Newcomb et
The
Collaborative
Breast Cancer
Pre- and post-
diagnosis
22890 pre-
diagnosis and
4881 post-
11.3 years Overall and
cancer-specific
Pre-diagnosis:
HR1= 0.96; 95% CI 0.88-
1.05
Page 92 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
al., 2013) Study (CBCS)
/ United States
diagnosis HR2= 0.89; 95% CI 0.77-
1.04
Post-diagnosis
total alcohol:
HR1= 0.64; 95% CI 0.47-
0.88 (p for trend = 0.001)
HR2= 0.83; 95% CI 0.45-
1.54
Lowry (2016)
(Lowry et al.,
2016)
Women's
Health
Initiative
(WHI)
observational
study / United
States
Pre- and post-
diagnosis
7835 7.9 years Overall and
cancer-specific
mortality
Pre-diagnosis:
HR1= 0.74; 95% CI 0.61–
0.89 (only two categories)
HR2= 0.76; 95% CI 0.56–
1.04
Post-diagnosis:
HR1= 0.76; 95% CI 0.51–
1.12
HR2= 1.21; 95% CI 0.62–
2.34
age, income, study
(clinical trial vs.
observational study),
family history of breast
cancer, smoking history,
BMI, and history of
menopausal hormone
therapy
Cervical cancer
Page 93 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Serur (1995)
(Serur et al.,
1995)
Brooklyn and
Kings
County study /
United States
Pre- diagnosis /
FFQ
332 w 5.0 years Cancer-
specific
mortality
HR2= 1.4; 95% CI 0.7-1.9 age <50 yr, advanced
stage, smoking, heroin and
cocaine use, public
hospital admission
Non-Hodgkin lymphoma
Battaglioli
(2006)
(Battaglioli et
al., 2006)
Follow-up of
patients who
were originally
recruited in a
case-control
study / Italy
Pre-diagnosis /
FFQ
1138 m/w 6.6 years Overall
mortality
HR1= 1.41; 95% CI 1.10–
1.81 (only two categories)
sex, age, education, type
of interview, smoking,
other variables
Talamini (2008)
(Talamini et al.,
2008)
Follow-up of
patients who
were originally
recruited in a
case-control
study / Italy
Pre-diagnosis /
FFQ
268 m/w 5.0 years Overall
mortality
HR1= 1.69; 95% CI 1.04-
2.76 (p for trend = 0.02)
sex, age, B-symptoms and
International Prognostic
Index
Han (2010)
(Han et al.,
Follow-up of
patients who
Pre-diagnosis /
FFQ
575 w 7.8 years Overall
mortality
HR1= 0.90; 95% CI 0.70–
1.17
age, education, smoking,
disease stage, B-symptom
Page 94 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
2010b) were originally
recruited in a
case-control
study / United
States
presence, initial treatment,
and consumption of other
alcohol types for
each type of alcohol
Geyer (2010)
(Geyer et al.,
2010)
Follow-up of
patients who
were originally
recruited in a
case-control
study / United
States
Pre-diagnosis /
FFQ
458 m/w 7.7 years Overall
mortality
HR1= 1.55; 95% CI 1.06-
2.27 (p for trend = 0.03)
adjusted for age at
diagnosis, gender, race,
education, study site,
stage, chemotherapy,
radiation, B symptoms,
histology (all NHL
groups)
Kidney cancer
Park (2006)
(Park et al.,
2006)
National
Health
Insurance
Corporation
Study
(NHICS) /
Pre-diagnosis /
FFQ
421 m 4.1 years Overall
mortality
HR1= 0.86; 95% CI 0.42-
1.76
age, alcohol consumption,
BMI, fasting serum
glucose level,
cholesterol level, physical
activity, food preference,
blood pressure, other
Page 95 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Korea comorbidities
Laryngeal cancer
Crosignani
(1996)
(Crosignani et
al., 1996)
Lombardy
Cancer
Registry
(LCR) / Italy
Pre-diagnosis /
FFQ
218 m 8.0 years Overall
mortality
HR1= 1.12; 95% CI 0.70-
1.80
age at diagnosis, clinical
stage, occurrence of new
primaries
Prostate cancer
Park (2006)
(Park et al.,
2006)
National
Health
Insurance
Corporation
Study
(NHICS) /
Korea
Pre-diagnosis /
FFQ
256 m 3.9 years Overall
mortality
HR1= 1.85; 95% CI 0.79-
4.34
age, alcohol consumption,
BMI, fasting serum
glucose level,
cholesterol level, physical
activity, food preference,
blood pressure, other
comorbidities
First author’s name, year of publication, study name, country, timing dietary intake (pre- or post-diagnosis), number of cases, sex (m/w), follow-up period (median or
average), outcome measures, results; HR1= overall mortality, HR2= cancer-specific mortality, and adjustment factors
Table S9: Summary of findings on multivitamins supplement intake and mortality in cancer survivors
Author (year) Study / Timeframe / Number of Follow-up period Outcome Results Adjustments
Page 96 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
country exposure
assessment
cases / sex
(m/w)
(median or
average)
measures (HR or RR and 95% CI)
Bowel cancer
Ng (2010) (Ng
et al., 2010)
Cancer and
Leukemia
Group B
(CALGB) /
United States
Post-diagnosis
(during and after
treatment) / FFQ
1038 m/w 7.3 years Overall
mortality
During treatment:
HR1= 0.92; 95% CI 0.74-
1.16
After treatment:
HR1= 1.11; 95% CI 0.82-
1.52
age, sex, family history of
colorectal cancer, baseline
performance status, depth
invasion through bowel
wall, positive lymph
nodes, grade of tumour
differentiation, presence
bowel obstruction at
surgery, perineural
invasion, postoperative
CEA, treatment arm, BMI,
physical activity, aspirin
use, Western pattern diet
Breast cancer
Greenlee (2012)
(Greenlee et al.,
Life After
Cancer
Post-diagnosis /
FFQ
2264 w 58.3 years Overall
mortality,
HR1= 0.84; 95% CI 0.65-
1.08
age at diagnosis,
race/ethnicity, education,
Page 97 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
2012) Epidemiology
(LACE) study
/ United States
cancer-specific
mortality
HR2= 0.79; 95% CI 0.56-
1.12
stage positive lymph
nodes, hormone receptor
status, treatment BMI, 1
year before diagnosis,
smoking, alcohol, physical
activity, fruits and
vegetables, comorbidity
score at enrolment. Mutual
adjustment for the other
antioxidants
Nechuta (2011)
(Nechuta et al.,
2011)
Shanghai
Breast Cancer
Survival Study
(SBCSS) /
China
Post-diagnosis /
FFQ
4877 w 4.1 years Overall
mortality
HR1= 0.82; 95% CI 0.57–
1.17
TNM stage,
chemotherapy,
radiotherapy, tamoxifen
use, oestrogen and
progesterone receptor
status, education, income,
BMI, regular tea
consumption, regular
exercise participation,
Page 98 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
daily cruciferous vegetable
intake, daily soy protein
intake, and other vitamin
variables
Holmes (1999)
(Holmes et al.,
1999)
Nurses' Health
Study (NHS) /
United States
Pre- and post-
diagnosis / FFQ
1982 w 13.1 years Overall
mortality
Pre-diagnosis:
HR1= statistically non-
significant (not shown)
Post-diagnosis:
HR1= 1.07; 95% CI 0.80 –
1.43
quantiles of nutrient or
food intake prior to
diagnosis, previous diet
interval, age, diet interval,
calendar year of diagnosis,
body mass index, oral
contraceptive use,
menopausal status,
postmenopausal hormone
use, smoking, age at first
birth and parity, number of
metastatic lymph nodes,
tumor size, calories
Kwan (2011)
(Kwan et al.,
Life After
Cancer
Post-diagnosis
and combined
2264 w 10.0 years Overall
mortality,
Post-diagnosis:
HR1= 0.92; 95% CI 0.70-
age at diagnosis, ethnicity,
education, stage positive
Page 99 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
2011) Epidemiology
(LACE) study
/ United States
pre- and post-
diagnosis / FFQ
cancer-specific
mortality
1.20
Combined pre- and post-
diagnosis:
HR1= 0.79; 95% CI 0.56-
1.12
Post-diagnosis:
HR2= 0.88; 95% CI 0.61-
1.28
Combined pre- and post-
diagnosis:
HR2= 0.70; 95% CI 0.44-
1.11
lymph nodes, hormone
receptor status, treatment,
BMI 1 year before
diagnosis, smoking,
alcohol, physical activity,
fruits and vegetables,
comorbidity score at
enrolment
Poole (2013)
(Poole et al.,
2013)
The After
Breast Cancer
Pooling Project
/ China and
United States
Post-diagnosis /
FFQ
12019 w 8.4 years (mean) Overall
mortality,
cancer-specific
mortality
HR1= 0.94; 95% CI 0.83–
1.07
HR2= 0.95; 95% CI 0.82–
1.11
age at diagnosis, exercise,
stage, treatment, BMI,
menopausal status,
smoking status, use of
supplements
(multivitamins, vitamin A,
B, C, D, E, anti-oxidants)
Page 100 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
First author’s name, year of publication, study name, country, timing dietary intake (pre- or post-diagnosis), number of cases, sex (m/w), follow-up period (median or
average), outcome measures, results; HR1= overall mortality, HR2= cancer-specific mortality, and adjustment factors
Page 101 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
LITERATURE
BARNETT, G. C., SHAH, M., REDMAN, K., EASTON, D. F., PONDER, B. A. & PHAROAH, P. D. 2008. Risk factors for the incidence of breast cancer: do they affect
survival from the disease? J Clin Oncol, 26, 3310-6.
BATTAGLIOLI, T., GORINI, G., COSTANTINI, A. S., CROSIGNANI, P., MILIGI, L., NANNI, O., STAGNARO, E., TUMINO, R. & VINEIS, P. 2006. Cigarette smoking and
alcohol consumption as determinants of survival in non-Hodgkin's lymphoma: A population-based study. Annals of Oncology, 17, 1283-1289.
BEASLEY, J. M., NEWCOMB, P. A., TRENTHAM-DIETZ, A., HAMPTON, J. M., BERSCH, A. J., PASSARELLI, M. N., HOLICK, C. N., TITUS-ERNSTOFF, L., EGAN, K. M.,
HOLMES, M. D. & WILLETT, W. C. 2011. Post-diagnosis dietary factors and survival after invasive breast cancer. Breast Cancer Res Treat, 128, 229-
36.
BORUGIAN, M. J., SHEPS, S. B., KIM-SING, C., VAN PATTEN, C., POTTER, J. D., DUNN, B., GALLAGHER, R. P. & HISLOP, T. G. 2004. Insulin, macronutrient
intake, and physical activity: are potential indicators of insulin resistance associated with mortality from breast cancer? Cancer Epidemiol
Biomarkers Prev, 13, 1163-72.
BUCK, K., ZAINEDDIN, A. K., VRIELING, A., HEINZ, J., LINSEISEN, J., FLESCH-JANYS, D. & CHANG-CLAUDE, J. 2011. Estimated enterolignans, lignan-rich foods,
and fibre in relation to survival after postmenopausal breast cancer. Br J Cancer, 105, 1151-7.
CARR, P. R., JANSEN, L., WALTER, V., KLOOR, M., ROTH, W., BLAKER, H., CHANG-CLAUDE, J., BRENNER, H. & HOFFMEISTER, M. 2016. Associations of red and
processed meat with survival after colorectal cancer and differences according to timing of dietary assessment. Am J Clin Nutr, 103, 192-200.
CHAVARRO, J. E., STAMPFER, M. J., HALL, M. N., SESSO, H. D. & MA, J. 2008. A 22-y prospective study of fish intake in relation to prostate cancer incidence
and mortality. American Journal of Clinical Nutrition, 88, 1297-1303.
CHLEBOWSKI, R. T., BLACKBURN, G. L., THOMSON, C. A., NIXON, D. W., SHAPIRO, A., HOY, M. K., GOODMAN, M. T., GIULIANO, A. E., KARANJA, N.,
MCANDREW, P., HUDIS, C., BUTLER, J., MERKEL, D., KRISTAL, A., CAAN, B., MICHAELSON, R., VINCIGUERRA, V., DEL PRETE, S., WINKLER, M., HALL, R.,
SIMON, M., WINTERS, B. L. & ELASHOFF, R. M. 2006. Dietary fat reduction and breast cancer outcome: interim efficacy results from the Women's
Intervention Nutrition Study. J Natl Cancer Inst, 98, 1767-76.
CROSIGNANI, P., RUSSO, A., TAGLIABUE, G. & BERRINO, F. 1996. Tobacco and diet as determinants of survival in male laryngeal cancer patients. Int J Cancer,
65, 308-13.
DAL MASO, L., ZUCCHETTO, A., TALAMINI, R., SERRAINO, D., STOCCO, C. F., VERCELLI, M., FALCINI, F. & FRANCESCHI, S. 2008. Effect of obesity and other
lifestyle factors on mortality in women with breast cancer. Int J Cancer, 123, 2188-94.
DRAY, X., BOUTRON-RUAULT, M. C., BERTRAIS, S., SAPINHO, D., BENHAMICHE-BOUVIER, A. M. & FAIVRE, J. 2003. Influence of dietary factors on colorectal
cancer survival. Gut, 52, 868-73.
EWERTZ, M., GILLANDERS, S., MEYER, L. & ZEDELER, K. 1991. Survival of breast cancer patients in relation to factors which affect the risk of developing
breast cancer. International Journal of Cancer, 49, 526-530.
FLATT, S. W., THOMSON, C. A., GOLD, E. B., NATARAJAN, L., ROCK, C. L., AL-DELAIMY, W. K., PATTERSON, R. E., SAQUIB, N., CAAN, B. J. & PIERCE, J. P. 2010.
Low to moderate alcohol intake is not associated with increased mortality after breast cancer. Cancer Epidemiol Biomarkers Prev, 19, 681-8.
Page 102 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
FUNG, T. T., KASHAMBWA, R., SATO, K., CHIUVE, S. E., FUCHS, C. S., WU, K., GIOVANNUCCI, E., OGINO, S., HU, F. B. & MEYERHARDT, J. A. 2014. Post
diagnosis diet quality and colorectal cancer survival in women. PLoS ONE, 9 (12) (no pagination).
GEYER, S. M., MORTON, L. M., HABERMANN, T. M., ALLMER, C., DAVIS, S., COZEN, W., SEVERSON, R. K., LYNCH, C. F., WANG, S. S., MAURER, M. J., HARTGE,
P. & CERHAN, J. R. 2010. Smoking, alcohol use, obesity, and overall survival from non-Hodgkin lymphoma: A population-based study. Cancer, 116,
2993-3000.
GREENLEE, H., KWAN, M. L., KUSHI, L. H., SONG, J., CASTILLO, A., WELTZIEN, E., QUESENBERRY, C. P., JR. & CAAN, B. J. 2012. Antioxidant supplement use
after breast cancer diagnosis and mortality in the Life After Cancer Epidemiology (LACE) cohort. Cancer, 118, 2048-58.
HAN, X., ZHENG, T., FOSS, F., HOLFORD, T. R., MA, S., ZHAO, P., DAI, M., KIM, C., ZHANG, Y., BAI, Y. & ZHANG, Y. 2010a. Vegetable and fruit intake and non-
Hodgkin lymphoma survival in Connecticut women. Leuk Lymphoma, 51, 1047-54.
HAN, X., ZHENG, T., FOSS, F. M., MA, S., HOLFORD, T. R., BOYLE, P., LEADERER, B., ZHAO, P., DAI, M. & ZHANG, Y. 2010b. Alcohol consumption and non-
Hodgkin lymphoma survival. Journal of cancer survivorship : research and practice, 4, 101-109.
HARRIS, H. R., BERGKVIST, L. & WOLK, A. 2012a. Alcohol intake and mortality among women with invasive breast cancer. Br J Cancer, 106, 592-5.
HARRIS, H. R., BERGKVIST, L. & WOLK, A. 2012b. Coffee and black tea consumption and breast cancer mortality in a cohort of Swedish women. British
Journal of Cancer, 107, 874-878.
HEBERT, J. R., HURLEY, T. G. & MA, Y. 1998. The effect of dietary exposures on recurrence and mortality in early stage breast cancer. Breast Cancer Res
Treat, 51, 17-28.
HELLMANN, S. S., THYGESEN, L. C., TOLSTRUP, J. S. & GRONBAEK, M. 2010. Modifiable risk factors and survival in women diagnosed with primary breast
cancer: Results from a prospective cohort study. European Journal of Cancer Prevention, 19, 366-373.
HOLM, M., OLSEN, A., CHRISTENSEN, J., KROMAN, N. T., BIDSTRUP, P. E., JOHANSEN, C., OVERVAD, K. & TJONNELAND, A. 2013. Pre-diagnostic alcohol
consumption and breast cancer recurrence and mortality: results from a prospective cohort with a wide range of variation in alcohol intake. Int J
Cancer, 132, 686-94.
HOLMES, M. D., STAMPFER, M. J., COLDITZ, G. A., ROSNER, B., HUNTER, D. J. & WILLETT, W. C. 1999. Dietary factors and the survival of women with breast
carcinoma. Cancer, 86, 826-35.
KENFIELD, S. A., DUPRE, N., RICHMAN, E. L., STAMPFER, M. J., CHAN, J. M. & GIOVANNUCCI, E. L. 2014. Mediterranean diet and prostate cancer risk and
mortality in the Health Professionals Follow-up Study. Eur Urol, 65, 887-94.
KROENKE, C. H., FUNG, T. T., HU, F. B. & HOLMES, M. D. 2005. Dietary patterns and survival after breast cancer diagnosis. Journal of Clinical Oncology, 23,
9295-9303.
KROENKE, C. H., KWAN, M. L., SWEENEY, C., CASTILLO, A. & CAAN, B. J. 2013. High-and low-fat dairy intake, recurrence, and mortality after breast cancer
diagnosis. Journal of the National Cancer Institute, 105, 616-623.
KWAN, M. L., GREENLEE, H., LEE, V. S., CASTILLO, A., GUNDERSON, E. P., HABEL, L. A., KUSHI, L. H., SWEENEY, C., TAM, E. K. & CAAN, B. J. 2011. Multivitamin
use and breast cancer outcomes in women with early-stage breast cancer: The life after cancer epidemiology study. Breast Cancer Research and
Treatment, 130, 195-205.
Page 103 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
KWAN, M. L., KUSHI, L. H., WELTZIEN, E., TAM, E. K., CASTILLO, A., SWEENEY, C. & CAAN, B. J. 2010. Alcohol consumption and breast cancer recurrence and
survival among women with early-stage breast cancer: The life after cancer epidemiology study. Journal of Clinical Oncology, 28, 4410-4416.
KWAN, M. L., WELTZIEN, E., KUSHI, L. H., CASTILLO, A., SLATTERY, M. L. & CAAN, B. J. 2009. Dietary patterns and breast cancer recurrence and survival
among women with early-stage breast cancer. Journal of Clinical Oncology, 27, 919-926.
LOCHHEAD, P., NISHIHARA, R., QIAN, Z. R., MIMA, K., CAO, Y., SUKAWA, Y., KIM, S. A., INAMURA, K., ZHANG, X., WU, K., GIOVANNUCCI, E., MEYERHARDT, J.
A., CHAN, A. T., FUCHS, C. S. & OGINO, S. 2015. Postdiagnostic intake of one-carbon nutrients and alcohol in relation to colorectal cancer survival.
Am J Clin Nutr, 102, 1134-41.
LOWRY, S. J., KAPPHAHN, K., CHLEBOWSKI, R. T. & LI, C. I. 2016. Alcohol use and breast cancer survival among participants in the Women's Health Initiative.
Cancer Epidemiol Biomarkers Prev.
MCCULLOUGH, M. L., GAPSTUR, S. M., SHAH, R., JACOBS, E. J. & CAMPBELL, P. T. 2013. Association between red and processed meat intake and mortality
among colorectal cancer survivors. J Clin Oncol, 31, 2773-82.
MCDONALD, P. A., WILLIAMS, R., DAWKINS, F. & ADAMS-CAMPBELL, L. L. 2002. Breast cancer survival in African American women: is alcohol consumption a
prognostic indicator? Cancer Causes Control, 13, 543-9.
MCELIGOT, A. J., LARGENT, J., ZIOGAS, A., PEEL, D. & ANTON-CULVER, H. 2006. Dietary fat, fiber, vegetable, and micronutrients are associated with overall
survival in postmenopausal women diagnosed with breast cancer. Nutr Cancer, 55, 132-40.
MEYERHARDT, J. A., NIEDZWIECKI, D., HOLLIS, D., SALTZ, L. B., HU, F. B., MAYER, R. J., NELSON, H., WHITTOM, R., HANTEL, A., THOMAS, J. & FUCHS, C. S.
2007. Association of dietary patterns with cancer recurrence and survival in patients with stage III colon cancer. Jama, 298, 754-64.
NECHUTA, S., CAAN, B. J., CHEN, W. Y., KWAN, M. L., LU, W., CAI, H., POOLE, E. M., FLATT, S. W., ZHENG, W., PIERCE, J. P. & SHU, X. O. 2013. Postdiagnosis
cruciferous vegetable consumption and breast cancer outcomes: a report from the After Breast Cancer Pooling Project. Cancer Epidemiol
Biomarkers Prev, 22, 1451-6.
NECHUTA, S., LU, W., CHEN, Z., ZHENG, Y., GU, K., CAI, H., ZHENG, W. & SHU, X. O. 2011. Vitamin supplement use during breast cancer treatment and
survival: a prospective cohort study. Cancer Epidemiol Biomarkers Prev, 20, 262-71.
NEWCOMB, P. A., KAMPMAN, E., TRENTHAM-DIETZ, A., EGAN, K. M., TITUS, L. J., BARON, J. A., HAMPTON, J. M., PASSARELLI, M. N. & WILLETT, W. C. 2013.
Alcohol consumption before and after breast cancer diagnosis: associations with survival from breast cancer, cardiovascular disease, and other
causes. J Clin Oncol, 31, 1939-46.
NG, K., MEYERHARDT, J. A., CHAN, J. A., NIEDZWIECKI, D., HOLLIS, D. R., SALTZ, L. B., MAYER, R. J., BENSON, A. B., 3RD, SCHAEFER, P. L., WHITTOM, R.,
HANTEL, A., GOLDBERG, R. M. & FUCHS, C. S. 2010. Multivitamin use is not associated with cancer recurrence or survival in patients with stage III
colon cancer: findings from CALGB 89803. J Clin Oncol, 28, 4354-63.
OLLBERDING, N. J., ASCHEBROOK-KILFOY, B., CACES, D. B., SMITH, S. M., WEISENBURGER, D. D. & CHIU, B. C. 2013. Dietary intake of fruits and vegetables
and overall survival in non-Hodgkin lymphoma. Leuk Lymphoma, 54, 2613-9.
PARK, S. M., LIM, M. K., SHIN, S. A. & YUN, Y. H. 2006. Impact of prediagnosis smoking, alcohol, obesity, and insulin resistance on survival in male cancer
patients: National Health Insurance Corporation Study. J Clin Oncol, 24, 5017-24.
Page 104 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
PELSER, C., AREM, H., PFEIFFER, R. M., ELENA, J. W., ALFANO, C. M., HOLLENBECK, A. R. & PARK, Y. 2014. Prediagnostic lifestyle factors and survival after
colon and rectal cancer diagnosis in the National Institutes of Health (NIH)-AARP Diet and Health Study. Cancer, 120, 1540-7.
PHIPPS, A. I., BARON, J. & NEWCOMB, P. A. 2011. Prediagnostic smoking history, alcohol consumption, and colorectal cancer survival: The Seattle Colon
Cancer Family Registry. Cancer, 117, 4948-4957.
PIERCE, J. P., NATARAJAN, L., CAAN, B. J., PARKER, B. A., GREENBERG, E. R., FLATT, S. W., ROCK, C. L., KEALEY, S., AL-DELAIMY, W. K., BARDWELL, W. A.,
CARLSON, R. W., EMOND, J. A., FAERBER, S., GOLD, E. B., HAJEK, R. A., HOLLENBACH, K., JONES, L. A., KARANJA, N., MADLENSKY, L., MARSHALL, J.,
NEWMAN, V. A., RITENBAUGH, C., THOMSON, C. A., WASSERMAN, L. & STEFANICK, M. L. 2007. Influence of a diet very high in vegetables, fruit, and
fiber and low in fat on prognosis following treatment for breast cancer: the Women's Healthy Eating and Living (WHEL) randomized trial. Jama, 298,
289-98.
POOLE, E. M., SHU, X., CAAN, B. J., FLATT, S. W., HOLMES, M. D., LU, W., KWAN, M. L., NECHUTA, S. J., PIERCE, J. P. & CHEN, W. Y. 2013. Postdiagnosis
supplement use and breast cancer prognosis in the After Breast Cancer Pooling Project. Breast Cancer Res Treat, 139, 529-37.
REDING, K. W., DALING, J. R., DOODY, D. R., O'BRIEN, C. A., PORTER, P. L. & MALONE, K. E. 2008. Effect of prediagnostic alcohol consumption on survival
after breast cancer in young women. Cancer Epidemiol Biomarkers Prev, 17, 1988-96.
RICHMAN, E. L., KENFIELD, S. A., STAMPFER, M. J., GIOVANNUCCI, E. L. & CHAN, J. M. 2011. Egg, red meat, and poultry intake and risk of lethal prostate
cancer in the prostate-specific antigen-era: incidence and survival. Cancer Prev Res (Phila), 4, 2110-21.
SAXE, G. A., ROCK, C. L., WICHA, M. S. & SCHOTTENFELD, D. 1999. Diet and risk for breast cancer recurrence and survival. Breast Cancer Res Treat, 53, 241-
53.
SERUR, E., FRUCHTER, R. G., MAIMAN, M., MCGUIRE, J., ARRASTIA, C. D. & GIBBON, D. 1995. Age, substance abuse, and survival of patients with cervical
carcinoma. Cancer, 75, 2530-8.
TALAMINI, R., POLESEL, J., SPINA, M., CHIMIENTI, E., SERRAINO, D., ZUCCHETTO, A., ZANET, E., FRANCESCHI, S. & TIRELLI, U. 2008. The impact of tobacco
smoking and alcohol drinking on survival of patients with non-Hodgkin lymphoma. Int J Cancer, 122, 1624-9.
TANG, L., ZIRPOLI, G. R., GURU, K., MOYSICH, K. B., ZHANG, Y., AMBROSONE, C. B. & MCCANN, S. E. 2010. Intake of cruciferous vegetables modifies bladder
cancer survival. Cancer Epidemiol Biomarkers Prev, 19, 1806-11.
TOMINAGA, K., ANDOW, J., KOYAMA, Y., NUMAO, S., KUROKAWA, E., OJIMA, M. & NAGAI, M. 1998. Family environment, hobbies and habits as
psychosocial predictors of survival for surgically treated patients with breast cancer. Jpn J Clin Oncol, 28, 36-41.
VRIELING, A., BUCK, K., SEIBOLD, P., HEINZ, J., OBI, N., FLESCH-JANYS, D. & CHANG-CLAUDE, J. 2013. Dietary patterns and survival in German
postmenopausal breast cancer survivors. Br J Cancer, 108, 188-92.
WAKAI, K., OHNO, Y., OBATA, K. & AOKI, K. 1993. Prognostic significance of selected lifestyle factors in urinary bladder cancer. Japanese Journal of Cancer
Research, 84, 1223-1229.
YANG, M., KENFIELD, S. A., VAN BLARIGAN, E. L., BATISTA, J. L., SESSO, H. D., MA, J., STAMPFER, M. J. & CHAVARRO, J. E. 2015a. Dietary patterns after
prostate cancer diagnosis in relation to disease-specific and total mortality. Cancer Prev Res (Phila), 8, 545-51.
YANG, M., KENFIELD, S. A., VAN BLARIGAN, E. L., WILSON, K. M., BATISTA, J. L., SESSO, H. D., MA, J., STAMPFER, M. J. & CHAVARRO, J. E. 2015b. Dairy intake
after prostate cancer diagnosis in relation to disease-specific and total mortality. Int J Cancer, 137, 2462-9.
Page 105 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
ZELL, J. A., MCELIGOT, A. J., ZIOGAS, A., HOLCOMBE, R. F. & ANTON-CULVER, H. 2007. Differential effects of wine consumption on colorectal cancer
outcomes based on family history of the disease. Nutrition and Cancer, 59, 36-45.
ZELL, J. A., ZIOGAS, A., BERNSTEIN, L., CLARKE, C. A., DEAPEN, D., LARGENT, J. A., NEUHAUSEN, S. L., STRAM, D. O., URSIN, G. & ANTON-CULVER, H. 2010.
Meat consumption, nonsteroidal anti-inflammatory drug use, and mortality among colorectal cancer patients in the California Teachers Study.
Cancer Prev Res (Phila), 3, 865-75.
ZHU, Y., WU, H., WANG, P. P., SAVAS, S., WOODROW, J., WISH, T., JIN, R., GREEN, R., WOODS, M., ROEBOTHAN, B., BUEHLER, S., DICKS, E., MCLAUGHLIN, J.
R., CAMPBELL, P. T. & PARFREY, P. S. 2013. Dietary patterns and colorectal cancer recurrence and survival: A cohort study. BMJ Open, 3 (2) (no
pagination).
Page 106 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097
For more information, visit www.prisma-statement.org.
PRISMA Flow Diagram – diet and mortality in cancer survivors
Records identified through
database searching
Bladder cancer n= 155
Breast cancer n= 1093
Cervical cancer n= 97
Colorectal cancer n= 755
(Non-)Hodgkin lymphoma n= 233
Kidney cancer n= 174
Larynx cancer n= 153
Multiple myeloma n= 89
Prostate cancer n= 445
Malignant melanoma n= 158
Testicular cancer n= 35
Uterus cancer n= 267
Scr
ee
nin
g
Incl
ud
ed
E
lig
ibil
ity
Id
en
tifi
cati
on
Additional records identified
through other sources
(n= 18)
Records after duplicates removed
(n= 3595)
Records screened
(n= 3595)
Records excluded
(n= 3512)
Full-text articles assessed
for eligibility
(n= 83)
Full-text articles excluded,
with reasons
(n= 21)
Studies included in
qualitative synthesis
(n= 60)
Page 107 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
PRISMA 2009 Checklist
Section/topic # Checklist item Reported on page #
TITLE Title 1 Identify the report as a systematic review, meta-analysis, or both. 1 ABSTRACT Structured summary 2 Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria,
participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number.
2
INTRODUCTION Rationale 3 Describe the rationale for the review in the context of what is already known. 4 Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons,
outcomes, and study design (PICOS). 4
METHODS Protocol and registration 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide
registration information including registration number. 7
Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale.
5-6
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.
5-6
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.
40
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).
5
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.
5
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.
5-6
Risk of bias in individual studies
12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.
7
Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means). 7 Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency
(e.g., I2) for each meta-analysis. NA
Page 1 of 2
Page 108 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
PRISMA 2009 Checklist
Section/topic # Checklist item Reported on page #
Risk of bias across studies 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within studies).
NA
Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-specified.
NA
RESULTS Study selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions
at each stage, ideally with a flow diagram. Figure S1 (in supplement)
Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations.
Table S1-S9 (in supplement)
Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12). Data on risk of bias of each study can be obtained by request – too much information (extremely large table for all included 60 studies) for the manuscript
Results of individual studies 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.
Table S1-S9 (in supplement)
Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency. NA Risk of bias across studies 22 Present results of any assessment of risk of bias across studies (see Item 15). NA
Page 109 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
PRISMA 2009 Checklist
Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]). NA DISCUSSION Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance
to key groups (e.g., healthcare providers, users, and policy makers). 25-30
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).
31
Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research. 33
FUNDING Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for
the systematic review. 33
From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097
For more information, visit: www.prisma-statement.org.
Page 2 of 2
Page 110 of 110
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
The impact of dietary patterns and the main food groups on mortality and recurrence in cancer survivors:
a systematic review of current epidemiological literature
Journal: BMJ Open
Manuscript ID bmjopen-2016-014530.R1
Article Type: Research
Date Submitted by the Author: 15-Jun-2017
Complete List of Authors: Jochems, Sylvia; Maastricht University, NUTRIM; University of Birmingham, Cancer and Genomic Sciences Van Osch, Frits; Maastricht University, NUTRIM; University of
Birmingham, Cancer and Genomic Sciences Bryan, Richard; University of Birmingham Wesselius, Anke; Maastricht University, NUTRIM van Schooten, Frederik; Maastricht University, Toxicology Cheng, Kar Keung; University of Birmingham, Department of Public Health and Epidemiology Zeegers, Maurice; University of Maastricht, NUTRIM School of Nutrition, Metabolism and Toxicology
<b>Primary Subject Heading</b>:
Epidemiology
Secondary Subject Heading: Oncology, Public health, Epidemiology
Keywords: cancer survivors, mortality, cancer recurrence, food, dietary pattern, diet
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open on D
ecember 7, 2020 by guest. P
rotected by copyright.http://bm
jopen.bmj.com
/B
MJ O
pen: first published as 10.1136/bmjopen-2016-014530 on 19 F
ebruary 2018. Dow
nloaded from
For peer review only
The impact of dietary patterns and the main food groups on 1
mortality and recurrence in cancer survivors: 2
a systematic review of current epidemiological literature 3
4
Sylvia H.J. Jochems (1,2), Frits H.M. van Osch (1,2), Richard T. Bryan (1), Anke Wesselius 5
(2), Frederik J. van Schooten (2), K.K. Cheng (3), Maurice P. Zeegers (2,4) 6
7
(1) Institute of Cancer and Genomic Sciences, University of Birmingham, B15 2TT 8
Birmingham, United Kingdom 9
(2) NUTRIM School of Nutrition and Translational Research in Metabolism, Maastricht 10
University, The Netherlands 11
(3) Institute of Applied Health Research, Public Health, Epidemiology and Biostatistics, 12
University of Birmingham, United Kingdom 13
(4) CAPHRI School for Public Health and Primary Care, Maastricht University, 6200 14
MD Maastricht, The Netherlands 15
16
17
Corresponding author contact information: 18
Sylvia H.J. Jochems 19
Institute of Cancer and Genomic Sciences, University of Birmingham, 21
B15 2TT Birmingham, United Kingdom 22
23
Page 1 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
2
ABSTRACT 24
Background: A systematic review of the limited evidence on dietary patterns/indices and 25
whole foods amongst cancer survivors will provide an overview of the status in nutrition 26
research amongst survivors. 27
Objective: To determine whether there is an association between dietary patterns/indices, 28
and foods from the main food groups prior to or after cancer diagnosis and mortality and 29
cancer recurrence in cancer survivors. 30
Data sources: PubMed, Embase and the Cochrane Library were searched from inception to 31
April 2017. Additional studies were identified by searching reference lists of studies. Two 32
authors independently screened titles and abstracts, assessed study quality and extracted data. 33
Participants: Cancer survivors of common cancers with a 10-year survival rate of 50% or 34
more: bladder, bowel, breast, cervical, kidney, laryngeal, prostate, testicular, uterine cancer, 35
malignant melanoma, and (non-)Hodgkin lymphoma. 36
Outcome measure: Mortality (overall, cancer-specific, death from other causes) and cancer 37
recurrence. 38
Results: A total of 38 studies were identified. Apart from the RCTs, the level of evidence is 39
rated low and very low and results should be interpreted with caution. Adherence to a high-40
quality diet and prudent diet after diagnosis appears to decrease the risk of overall mortality, 41
cancer-specific mortality, and death from other causes in breast cancer survivors. Adherence 42
to a Western diet, before and after diagnosis, appears to increase the risk of overall mortality, 43
cancer-specific mortality, and death from other causes amongst breast cancer survivors. For 44
bowel cancer survivors, adherence to a Western diet after diagnosis appears to increase the 45
risk of overall mortality. 46
Conclusion: 47
Page 2 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
3
For many cancer survivors there is little evidence to date to indicate that particular dietary 48
behaviours influence outcomes with regard to mortality and recurrence. Notwithstanding, 49
limited evidence suggests that a high-quality and prudent diet are beneficial for breast cancer 50
survivors, and that a Western diet is detrimental for breast and bowel cancer survivors. 51
52
53
Strengths and limitations 54
- Dietary patterns/indices and whole foods reflect the complexity of dietary intake and 55
capture synergistic relationships between various dietary constituents 56
- Studies investigating dietary patterns/indices and whole foods before diagnosis do not 57
consider potential modifications in dietary intake after cancer diagnosis 58
- Cohort studies provide weaker empirical evidence than RCTs for examining 59
relationships between dietary exposure and mortality and cancer recurrence 60
61
Page 3 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4
INTRODUCTION 62
As cancer survival rates continue to improve, there is an increased need to identify 63
modifiable lifestyle factors amongst cancer survivors in order to improve long-term health. 64
Adherence to a diet rich in fruit and vegetables could decrease the risk of several types of 65
cancer and increase overall life expectancy[1], the suggestion that epigenetic aberrations 66
occurring in cancer could be altered by nutrients makes it plausible that dietary changes after 67
successful cancer treatment could improve prognosis[2,3]. 68
Although cancer survivors are responsive to health promotion[4,5], a recent study has 69
indicated that survivors had poorer diets than individuals without cancer[6]. One possible 70
explanation could be the difficulty for cancer survivors in adopting a healthier diet without 71
clear evidence that it will improve their survival[7]. While guidelines have been well 72
documented for the prevention of cancer, many uncertainties remain for nutrition after cancer 73
treatment[8]. A systematic review, as part of the Continuous Update Project (CUP) of the 74
World Cancer Research Fund International, was published on diet, nutrition, physical activity 75
and survival in breast cancer survivors[9]. The independent panel of scientists concluded that 76
the evidence to date was not strong enough to make specific recommendations for breast 77
cancer survivors[10]. A recent meta-analysis investigating the role of diet on overall 78
mortality and recurrence among cancer survivors concluded that adherence to a Western diet 79
is positively associated with overall mortality, and a high-quality diet / healthy dietary pattern 80
is inversely associated with overall mortality amongst all cancer survivors[11]. 81
In the setting of survivors of cancers with a 10-year survival rate ≥50%, this 82
systematic review provides a structured overview of RCTs and cohort studies addressing the 83
relationship between adherence to dietary patterns/indices and whole foods from the main 84
food groups prior to or after cancer diagnosis and mortality and cancer recurrence. Making 85
healthier dietary changes can be time consuming and challenging, especially after cancer 86
Page 4 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
5
treatment. Survivors of cancers with a high survival rate may be most likely to benefit from 87
such changes to prevent or delay cancer recurrence and improve survival. In this review, we 88
sought to investigate groups of cancer survivors individually. Given the heterogeneous nature 89
of cancer[12], it should be noted that when a particular dietary pattern/indice or food has been 90
evidenced to influence the risk of recurrence and/or mortality for a certain cancer type, this 91
does not necessarily apply for survivors of all cancer types. 92
93
Page 5 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
6
METHODS 94
Search strategy 95
From inception up to April 2017, Pubmed, Embase and the Cochrane Library were 96
searched to find English language articles of original and published randomized trials and 97
observational studies to answer the following research question: do the highest 98
adherence/intake of dietary patterns/indices and foods compared to the lowest 99
adherence/intake prior to or after cancer diagnosis, increase or decrease the risk of mortality 100
or cancer recurrence amongst cancer survivors of common cancers with a 10-year survival 101
rate of 50% or more? This research question was developed using the PICO framework 102
(supporting data review protocol File S1). Search strategies included search terms related to 103
dietary patterns, diet quality, foods from the main food groups, and outcomes of interest, 104
including overall mortality, cancer-specific mortality, death from other causes, and 105
recurrence of cancer. Additionally, studies were identified by searching reference lists of 106
relevant studies, literature reviews and meta-analyses. After the search was completed, 107
articles were screened and selected independently based on the title and abstract by two of the 108
authors (SJ and FvO). The data extraction was performed independently by the same authors 109
(SJ and FvO) and any disagreements about study inclusion were resolved through consensus 110
or a third party. 111
112
Inclusion and exclusion criteria 113
Eligibility criteria included adult survivors of cancer (no sex or age restriction) who were 114
defined as individuals who had been diagnosed with a primary cancer, received cancer 115
therapy, and were in remission or had recovered completely from cancer. Considered cancer 116
types were the commonly-occurring cancers in the Western world with a 10-year net survival 117
of at least 50% (based on cancer diagnoses of men and women during 2010-2011 in England 118
Page 6 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7
and Wales)[13]. These include in decreasing order of net survival: testicular cancer (98%), 119
malignant melanoma (MM) (89%), prostate cancer (84%), Hodgkin lymphoma (HL) (80%), 120
breast cancer (78%), uterine cancer (77%), non-Hodgkin lymphoma (NHL) (63%), cervical 121
cancer (63%), laryngeal cancer (62%), bowel cancer (57% including both colon and rectal 122
cancer), bladder cancer (50%), and kidney cancer (50%). In the statistical analyses 123
adjustments had to be made for at least age and disease stage at baseline and, where possible, 124
for cancer treatment. Exclusion papers did not state hazard ratios (HRs) or relative risks 125
(RRs), nor 95% confidence intervals (95% CI); neither did they provide information on 126
disease stage and/or tumour grade or therapy. Additionally, studies where mortality and 127
cancer recurrence were combined with other outcomes (e.g. cancer progression), confirmed 128
cancer-specific mortality were combined with a diagnosis of metastasis, or where 129
prostate cancer recurrence was determined by a rising PSA level only, were excluded. 130
131
Dietary exposure 132
Dietary patterns/indices that were considered were assessed by index-based methods 133
and data-driven approaches, such as principal component analysis (factor analysis) and 134
cluster analysis[14]. The following diets were considered: the Healthy Eating Index 2005 135
(HEI)[15,16] and the alternate Healthy Eating Index 2010 (AHEI)[17,18], the WCRF/AIRC 136
dietary guidelines adherence score[19] and the American Cancer Society diet-specific 137
recommendations for cancer prevention (ACS)[20], the recommended food score (RFS)[21], 138
the Diet Quality Index-Revised (DQIR)[22], the Dietary Approaches to Stop Hypertension 139
diet (DASH) diet[23], the Healthy Nordic Food Index (HNFI)[24], the alternate 140
Mediterranean diet (aMed)[25,26], a prudent/healthy diet and low-fat diet, and a 141
Western/unhealthy diet. The HEI and AHEI target foods that could possibly reduce the risk 142
of chronic diseases and include fruits, vegetables, fibre, soy, nuts, ratio white and red meat, 143
Page 7 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
8
alcohol, trans fat, saturated fat ratio, and multivitamin use. The RFS includes the foods fruits, 144
vegetables, whole grains, dairy and protein foods low in fat. Diet diversity and moderation 145
was addressed by the DQIR and included fruits, vegetables, cholesterol, total fat, saturated 146
fat, iron, calcium, and fat/sugar moderation. The aMed is based on the original Mediterranean 147
diet score and includes fruits, vegetables, legumes, nuts, whole grains, red and processed 148
meat, moderate alcohol, and the ratio of monounsaturated and saturated fat[27,28]. 149
Nutritious foods of the main food groups (UK Eatwell Guide)[29] that stimulate the 150
consumption of healthier and more sustainable foods were considered. The composition of 151
the investigated groups was as follows: (I) fruit and vegetables including citrus fruits, stone 152
fruits, soft fruits, fleshy fruits, vine fruits, flower vegetables, leafy vegetables, stem 153
vegetables, fruit vegetables, mushrooms, bulbs and roots; (II) grain foods including potatoes, 154
bread, rice, pasta and cereal; (III) protein foods including meat (processed meat, unprocessed 155
meat, red meat, poultry), fish, eggs, tofu, nuts, seeds, pulses, legumes and beans; (IV) dairy 156
and alternative products including yoghurt, milk, cheese; (V) oils and spreads including 157
vegetable oils, spreads. Information on intake of these foods was obtained before and/or after 158
cancer diagnosis with food records, food frequency questionnaires (FFQ) (self-administered 159
or via an interview), or twenty-four-hour recalls, and expressed in servings or (milli)grams 160
per day/week/month. No restrictions were made for time of follow-up, and timing or 161
frequency of dietary intake. 162
163
Mortality and cancer recurrence outcome 164
Considered endpoints were overall mortality, cancer-specific mortality, death from 165
other causes, and cancer recurrence. The cause of death was confirmed via death certificates 166
or the National Death Index in each of the studies. Cancer recurrence was defined as a new 167
occurrence of cancer after a period of time during which the cancer could not be detected at 168
Page 8 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
9
the same or at a different site to the initial primary tumour. Cancer recurrence had to be 169
confirmed by a biopsy, scan, medical record, cancer registry, or treating physician. 170
171
Risk of bias assessment and level of quality 172
The Cochrane Collaboration risk of bias assessment tools were used for appraisal of 173
RCTs [30]and cohort studies[31]. For the two RCTs the RoB 2.0 tool (a revised tool for risk 174
of bias in randomized trials) was used to evaluate the risk of bias. Cohort studies were 175
appraised with an adjusted version of the ROBINS-I tool[30,31]. 176
Levels of quality were determined with the GRADE system[32]; evidence from RCTs 177
or multiple double-upgraded observational studies were considered as high quality, 178
downgraded RCTs or upgraded observational studies were considered as moderate quality, 179
double-downgraded RCTs or observational studies were considered as low quality, and triple-180
downgraded RCTs, downgraded observational studies or case series/case reports were 181
considered as very low quality[32]. Factors reducing the quality of the evidence include 182
limitations in study design, inconsistency of results, indirectness of evidence, imprecision, 183
and publication bias. Factors increasing the quality of the evidence include a large magnitude 184
of effect, correction for all plausible confounding that could reduce the demonstrated effect or 185
increase the effect if no effect was observed, and presence of a dose-response gradient. 186
187
Page 9 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
10
RESULTS 188
The detailed search strategy for bladder cancer survivors in PubMed is shown in Box 189
1. This search strategy was adapted for other types of cancer and for use in Embase and the 190
Cochrane Library. The protocol used for this systematic review that includes the search terms 191
for the other cancer types and the PRISMA flowchart are available in the supporting data 192
(File S1). Additionally, the review was written according to the PRISMA guidelines[33]. 193
The systematic review search resulted in 2883 citations after removal of duplicates. 194
After screening the titles and abstracts, 95 full-text articles were assessed for eligibility. A 195
total of 38 studies were finally included. No studies could be identified for cervical, kidney, 196
testicular, uterine cancer, HL or MM survivors. Therefore, we report only on bladder, bowel, 197
breast, laryngeal, prostate cancer, and NHL survivors. Dietary patterns/indices could be 198
identified for bowel, breast, prostate cancer, and NHL. Whole foods from the main food 199
groups could be identified for bladder, bowel, breast, laryngeal, prostate cancer and NHL 200
survivors. 201
A summary of the number of studies for each cancer type on pre-diagnosis dietary 202
patterns/indices (Table 1) and post-diagnosis dietary patterns/indices (Table 2) and mortality 203
and cancer recurrence is provided. The number of studies for pre-diagnosis and post-204
diagnosis foods on mortality and cancer recurrence are provided in Tables 3 and Table 4. The 205
study characteristics of each included study are provided in the supporting data (File S2). 206
Results for the assessment of the risk of bias for each individual RCT (RoB 2.0) and cohort 207
study (ROBINS-I) will be provided on request. Briefly, the included RCTs investigating a 208
low-fat diet and mortality amongst breast cancer survivors indicated a low risk of bias[34]. 209
The included cohort studies all had an acceptable risk of bias with only some concerns[35]. 210
The templates of the RoB 2.0 and ROBINS-I tools can be found in the supporting data (File 211
S1). An overview of the GRADE ratings with comments can be found in in the supporting 212
Page 10 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
11
data in the summary of findings Tables (File S3). The quality level of the body of evidence of 213
the included studies was rated ‘very low’, ‘low’ and ‘moderate’ by two of the authors (SJ and 214
FvO) when applying the grading system developed by the GRADE collaboration[32]. The 215
quality of the evidence for most of the included comparable cohort studies with 216
corresponding exposures and outcomes and was downgraded from ‘low’ to ‘very low’ due to 217
inconsistency, directness, and publication bias for most studies. Evidence from comparable 218
cohort studies that was not downgraded remained at a low level of evidence. These included 219
for bowel cancer survivors: pre-diagnosis processed meat intake and overall mortality and 220
cancer-specific mortality; and post-diagnosis adherence to a Western diet and overall 221
mortality. For breast cancer survivors these included: pre-diagnosis total fruit intake and 222
overall mortality; pre-diagnosis adherence to a Western diet and overall mortality, cancer-223
specific mortality, and death from other causes; post-diagnosis adherence to a low-fat diet 224
and overall mortality; post-diagnosis adherence to the HEI and AHEI scores and overall 225
mortality, cancer-specific mortality, and death from other causes; post-diagnosis adherence to 226
a prudent diet and overall mortality and death from other causes, and post-diagnosis 227
adherence to a Western diet and overall mortality and death from other causes. Evidence for a 228
low-fat diet from the two included RCTs was downgraded from ‘high’ to ‘moderate’ due to 229
the influence of weight loss of the breast cancer survivors on mortality. We would like to 230
note that study results described as ‘not associated with mortality/cancer recurrence’, are 231
based on statistical ‘significance tests’ performed by the authors of the included studies with 232
the focus on traditional definitions of p values and statistical significance on null hypotheses. 233
Notwithstanding, we believe that a correct interpretation of statistical tests demands critical 234
examining the sizes of effect estimates and confidence limits, p values, and the assumptions 235
and conventions used for the statistical analysis[36]. 236
237
Page 11 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
12
Bladder cancer 238
A total of 1 cohort study could be identified for bladder cancer survivors regarding 239
fruit and vegetable consumption. The study of Tang et al. investigated pre-diagnosis fruit and 240
vegetable consumption with data from 239 male and female bladder cancer survivors from 241
the Roswell Park Cancer Institute (RPCI) Tumor Registry[37]. After an average of 8-year 242
follow-up, no associations were observed between overall mortality or bladder cancer-243
specific mortality when comparing survivors with the highest intakes of total fruit, total 244
vegetables or other cruciferous vegetables (raw or cooked) with those in the lowest intake 245
group. An inverse association was, however, observed for broccoli intake (≥1 versus <1 246
serving per month) and overall mortality (broccoli raw HR=0.57; 95% CI 0.39-0.83, broccoli 247
cooked HR=0.67; 95% CI 0.49-0.91) and bladder cancer-specific mortality (broccoli raw 248
HR=0.43; 95% CI 0.25-0.74). The intake of other raw and cooked vegetables including 249
cabbage, cauliflower, Brussels sprouts, kale, turnip, collard or mustard greens was not related 250
with mortality[37]. 251
In summary, no conclusive evidence of an association with mortality amongst bladder 252
cancer survivors could be provided based on one study. 253
254
Bowel cancer 255
A total of 12 cohort studies could be identified for bowel cancer survivors. Three 256
observational cohort studies could be identified investigating the role of a pre- and post-257
diagnosis prudent diet on mortality in bowel cancer survivors. Results of the Cancer and 258
Leukemia Group B (CALGB) study indicated no associations between a prudent diet after 259
cancer diagnosis and lower mortality[38]. However, in female bowel cancer survivors there 260
was a higher overall mortality associated with the highest post-diagnosis intakes of a Western 261
Page 12 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
13
diet in comparison with female bowel cancer survivors in the lowest category (HR= 2.32; 262
95% CI 1.36-3.96)[38]. When comparing participants in the Familial Bowel Cancer Registry 263
(FBCR) with the highest and lowest intakes of a prudent diet before cancer diagnosis, no 264
associations were found with mortality[39]. In FBCR two dietary patterns were identified to 265
be comparable with a Western diet: a high processed meat pattern and a high sugar pattern 266
diet. No associations were reported for a dietary pattern high in sugar and mortality when 267
comparing the highest to the lowest intake group, whereas a high processed meat diet was 268
specifically related to increased colon cancer mortality (HR=2.13; 95% CI 1.03-4.43)[39]. In 269
the Nurses' Health Study (NHS), no associations were observed between a post-diagnosis 270
prudent diet and mortality when comparing the highest and lowest intakes in bowel cancer 271
survivors[25]. No associations were observed between a Western diet and mortality when 272
comparing the highest and lowest intakes in bowel cancer survivors[25]. Carr et al. reported 273
that red and processed meat consumption was not associated with a poorer survival amongst 274
stage I–III bowel cancer survivors in a follow-up study of the Darmkrebs: Chancen der 275
Verhutung durch Screening (DACHS) study in the Rhine-Neckar region in southwest 276
Germany[40]. However, the authors suggested that major changes in the consumption of red 277
meat measured at 5-year follow-up could influence survival estimates[40]. Similar results 278
were observed in a pooled analysis for pre-diagnosis meat intake amongst 3,789 279
participants[41]. The study of McCullough et al. indicated a detrimental association for 280
mortality when comparing highest versus lowest or never pre-diagnosis red and processed 281
meat consumption for overall mortality (RR= 1.29; 95% CI 1.05- 1.59) and death from other 282
causes than bowel cancer (RR= 1.39; 95% CI 1.00-1.92)[41]. No evidence of an association 283
with mortality was observed for foods from the main food groups, including fruits, 284
vegetables, dairy, or protein foods amongst bowel cancer survivors [41,42]. The consumption 285
of whole grain products was investigated amongst 1,119 Danish, Swedish and Norwegian 286
Page 13 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
14
bowel cancer survivors in the HELGA cohort[43]. Separate analysis for men and women 287
indicated no association between whole grains and overall mortality. Adherence to the HEI 288
diet score was investigated in a large study including 5,727 male and female survivors in the 289
USA and indicated no association between the highest pre-diagnosis adherence score of the 290
HEI diet with overall mortality or cancer-specific mortality[44]. Recently, a German study 291
examined adherence to the Modified Mediterranean Diet Score (MMDS) and the Healthy 292
Nordic Food Index (HNFI) and found that post-diagnosis adherence to this MMDS was 293
associated with a decreased risk of overall mortality amongst bowel cancer survivors (HR= 294
0.48; 95% CI 0.32-0.74)[45]. In The European Prospective Investigation into Cancer and 295
Nutrition (EPIC) study, data from participants of 10 European countries was analysed on 296
adherence to WCRF/AICR diet score, and intake and total dairy, milk, yoghurt, cheese, red 297
meat, processed meat, and poultry[46,47]. Pre-diagnosis adherence to this high-quality diet 298
score indicated a decreased risk of overall mortality amongst bowel cancer survivors (HR= 299
0.79; 95% CI 0.65-0.98)[46]. 300
In summary, no conclusive evidence of an association with mortality was observed 301
for bowel cancer survivors whilst most results are based on only one study. Nevertheless, 302
results carefully suggest that pre-diagnosis processed meat intake and post-diagnosis 303
adherence to a Western dietary pattern is associated with an increased risk of overall 304
mortality amongst bowel cancer survivors. 305
306
Breast cancer 307
A total of 2 RCTs and 16 cohort studies could be identified for breast cancer 308
survivors. Two dietary intervention trials amongst breast cancer survivors met the eligibility 309
criteria of our literature review. The study of Chlebowski et al. aimed to reduce post-310
Page 14 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
15
diagnosis dietary fat intake to almost one sixth of total energy intake in women participating 311
in the Women’s Intervention Nutrition Study (WINS)[48]. Breast cancer survivors in the 312
intervention group were informed extensively on maintaining weight based on energy intake, 313
whilst minimum dietary advice on nutrient intake was provided to breast cancer survivors in 314
the control group. Women in the intervention group had a lower dietary fat intake compared 315
to those in the control group, whereas no differences could be observed for a lower energy or 316
higher dietary fibre intake. According to the authors there was no difference in overall 317
mortality between women adhering to a low-fat diet and women given minimum dietary 318
advice (HR=0.89; 95% CI 0.65-1.21). In the Women’s Healthy Eating and Living (WHEL) 319
study breast cancer survivors in the intervention group received telephone counselling with 320
additional cooking classes and brochures to support adherence to a post-diagnosis diet high in 321
fruit (3 servings/day), high in vegetables (5 servings/day and 16oz of vegetable juice), high in 322
fibre (30g/day), and low in fat (15-20% of energy intake from fat)[49]. In the control group, 323
breast cancer survivors received written advice to eat at least 5 portions of fruit and 324
vegetables each day (5-a-day advice). Differences between the former and latter groups in 325
mean consumption of vegetables (+65%), fruit (+25%), fibre (+30%), and energy from fat (-326
13%) were observed at 4 years. No differences were observed for overall survival comparing 327
women in the intervention group with those in the control group (HR=0.91; 95% CI 0.72-328
1.15)[49]. 329
Post-diagnosis dietary indices were examined in the Health, Eating, Activity, and 330
Lifestyle (HEAL) study[50], Women’s Health Initiative’s Dietary Modification Trial and 331
Observational Study (WHI)[51], Nurses’ Health Study (NHS)[52] and Cancer Prevention 332
Study II Nutrition Cohort (CPS-II)[53]. McCullough et al. investigated additionally pre-333
diagnosis adherence to the ACS diet score in the CPS-II cohort but found no association 334
between pre- nor post-diagnosis diet score adherence and mortality[53]. In the NHS, the post-335
Page 15 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
16
diagnosis dietary DQIR, RFS, aMed, AHEI, and DASH scores were not associated with 336
breast cancer survival[52]. Closer adherence to DASH and AHEI were, however, related to a 337
lower risk of death from other causes than breast cancer[52]. George et al. examined post-338
diagnosis adherence to the HEI-2005 score and concluded that women in the WHI cohort 339
who consumed better quality diets had a 26% lower risk of death from any cause (HR= 0.74; 340
95% CI 0.55-0.99) and a 42% lower risk of death from non-breast cancer related death (HR= 341
0.58; 95% CI 0.38-0.87)[51]. In the HEAL study adherence to the highest post-diagnosis 342
HEI-2005 score was related to a decreased risk of overall mortality and death from breast 343
cancer (HR= 0.40; 95% CI 0.17-0.94 and HR= 0.12; 95% CI 0.02-0.99, respectively)[50]. 344
Results of the NHS study indicated that a post-diagnosis prudent diet was not 345
associated with overall or breast cancer specific mortality whilst death from other causes was 346
associated with a prudent diet after diagnosis when comparing breast cancer survivors of the 347
highest and lowest intake group (HR=0.54; 95% CI 0.31-0.95)[54]. A prudent diet before 348
diagnosis was not associated with mortality amongst breast cancer survivors in the NHS. 349
Both pre- and post-diagnosis adherence to a Western diet in women with the highest 350
adherence were associated with death from other causes (respectively RR=1.95; 95% CI 351
1.06-3.60 and RR=2.31; 95% CI 1.23-4.32). No associations were observed between a pre- or 352
post-diagnosis Western diet and overall or breast cancer specific mortality[54]. In the Life 353
After Cancer Epidemiology (LACE) study investigating a post-diagnosis prudent diet in 354
women with early-stage breast cancer, breast cancer survivors with the highest adherence to a 355
prudent diet had a decreased risk of death from other causes (HR=0.35; 95% CI 0.17-0.73) 356
and overall mortality (HR=0.57; 95% CI 0.36-0.90) compared to women with the lowest 357
adherence to this diet[55]. The study of Vrieling et al. investigated associations between a 358
‘healthy’ and ‘unhealthy’ pre-diagnosis dietary pattern and mortality in German breast cancer 359
survivors in the Mammary carcinoma Risk factor Investigation (MARIE) study[56]. The 360
Page 16 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
17
characteristics of the defined healthy diet are comparable with a prudent diet; nevertheless, no 361
associations between the highest and lowest intake of this defined ‘healthy’ diet before cancer 362
diagnosis and mortality in breast cancer survivors were observed. However, the results did 363
indicate that a higher intake of an ‘unhealthy’ diet could increase the risk of death from other 364
causes (HR=3.69; 95% CI 1.66-8.17) amongst breast cancer survivors compared to those 365
with the lowest intake of this diet[56]. 366
The majority of studies indicated no association between (total) pre- or post-diagnosis 367
fruit and/or vegetable intake and mortality in breast cancer survivors. However, one study 368
found that, when comparing postmenopausal breast cancer survivors in the highest tertile to 369
the lowest tertile group, pre-diagnosis total vegetable intake improved overall survival 370
(HR=0.57; 95% CI 0.35-0.94) - no association was found for total fruit intake and mortality 371
in this cohort of breast cancer survivors[57]. Dal Maso et al. found a difference for total fruit 372
and vegetable consumption and overall mortality (HR= 1.27; 95% CI 1.00-1.61) when 373
comparing survivors of the lowest intake group to the highest intake group[58]. 374
Holmes et al. reported a beneficial association between the highest post-diagnosis 375
poultry consumption and mortality in women once diagnosed with breast cancer (HR=0.70; 376
95% CI 0.50–0.97)[59]. No associations were found for fish or red meat consumption and 377
mortality in this population. Additionally, a high dairy intake before diagnosis amongst 378
female registered nurses who participated in the NHS, was related to overall survival (HR= 379
0.72; 95% CI 0.52–1.00)[59]. Kroenke et al. found that post-diagnosis dairy intake amongst 380
women diagnosed with early-stage invasive breast cancer in the LACE study, was associated 381
with an increased overall mortality (HR=1.39; 95% CI 1.02-1.90)[60]. More specifically, 382
high fat dairy was related to overall mortality and breast cancer specific mortality in these 383
women (respectively HR= 1.64; 95% CI 1.24-2.17 and HR= 1.49; 95% CI 1.00-2.24) whilst 384
low-fat dairy was not[60]. Beasley et al. examined both meat and dairy intake after diagnosis 385
Page 17 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
18
and found no association with survival in the Collaborative Woman’s Longevity Study 386
(CWLS)[61]. Pre-diagnosis intakes of neither bread, sunflower/pumpkin seeds nor 387
sesame/flaxseeds reduced the risk of mortality in the MARIE study[62]. Finally, post-388
diagnosis butter/margarine/lard consumption did increase the risk of breast cancer recurrence 389
in a follow-up study amongst 472 breast cancer survivors enrolled from the Memorial Sloan-390
Kettering Cancer Centre (RR= 1.30; 95% CI 1.03-1.64)[63]. 391
In summary, no conclusive evidence of an association with mortality was observed 392
for most foods of the main food groups, including vegetables, meat, or dairy amongst breast 393
cancer survivors. However, limited evidence appears to show that fruit intake, a high-quality 394
diet and a prudent diet are beneficial for breast cancer survivors whilst a Western diet is 395
detrimental for breast cancer survivors. 396
397
Laryngeal cancer 398
One cohort study could be identified for the association between several foods from 399
the main food groups and mortality amongst laryngeal cancer survivors[64]. Crosignani et.al 400
examined dietary habits and survival in of 215 Italian male laryngeal cancer survivors on pre-401
diagnosis dietary habits and survival. The consumption of total vegetables (HR=0.57; 95% CI 402
0.35-0.94), beef/veal (HR= 0.50; 95% CI 0.30-0.83), and bread (HR= 0.54; 95% CI 0.32-403
0.90) were all associated with a decreased risk of overall mortality when comparing the 404
highest versus the lowest intake group. No associations were found for poultry, fish, eggs, 405
milk, cheese, pasta, potatoes, citrus fruits, other fruits, butter, or olive oil. The authors 406
speculate that the beneficial relation between the highest beef/veal intake and mortality could 407
tentatively be interpreted as an indicator of a good nutritional status of those participants[64]. 408
In summary, no conclusive evidence of an association with mortality amongst 409
laryngeal cancer survivors could be provided based on one study. 410
Page 18 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
19
411
Non-Hodgkin Lymphoma (NHL) 412
A total of 2 cohort studies could be identified for NHL survivors regarding the intake 413
of several food items. One study indicated that pre-diagnosis intakes of total fruit and 414
vegetables and vegetables only (highest versus lowest intake) were associated with decreased 415
overall mortality (respectively HR= 0.68; 95% CI 0.49-0.95 and HR=0.58; 95% CI 0.38-416
0.89) amongst female NHL survivors[65]. Additionally, the highest intakes of citrus fruits 417
and green leafy vegetables compared with the lowest intakes were related to overall mortality 418
amongst survivors with NHL (respectively HR=0.73; 95% CI 0.54-0.99 and HR=0.71; 95% 419
CI 0.51-0.98). No associations were observed for total fruit intake, yellow vegetables, red 420
vegetables or bean vegetables and mortality whilst sub-analysis investigating fruit and 421
vegetables separately for each NHL subtypes did; consumption of citrus fruits improved 422
survival in diffuse large B-cell lymphoma survivors (overall mortality HR=0.40; 95% CI 423
0.22–0.72, cancer-specific mortality HR=0.36; 95% CI 0.16–0.80), and the highest 424
consumption of green leafy vegetables favoured overall mortality in follicular lymphoma 425
survivors (HR= 0.27; 95% CI 0.10–0.76)[65]. Leo et al. found no association between pre-426
diagnosis intakes of fruit, vegetables, meat, fish, legumes, and soy and overall mortality in 427
2,339 male and female NHL survivors[66]. Dairy intake, however, was associated with 428
higher overall mortality (highest versus lowest tertile: HR= 1.14; 95% CI 1.00-1.31) - not 429
with NHL-specific mortality (HR= 1.16; 95% CI 0.98-1.37)[66]. 430
In summary, no conclusive evidence of an association with mortality amongst NHL 431
survivors could be provided based on one study. 432
433
Prostate cancer 434
Page 19 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
20
Adherence to a Western diet after prostate cancer diagnosis was associated with 435
increased overall mortality (HR=1.67; 95% CI 1.16–2.42) and prostate-cancer mortality 436
(HR= 2.53; 95% CI 1.00-6.42) amongst non-metastatic prostate cancer survivors in the 437
Physician’s Health Study (PHS)[67]. The derived Western dietary patterns appeared to be 438
driven by the consumption of processed meat. A prudent diet was investigated (showing 439
overlapping characteristics with the Mediterranean diet examined in the Health Professionals 440
Follow-up Study (HPFS)); adherence to a prudent diet after prostate cancer diagnosis was 441
inversely associated with overall mortality (RR=0.64; 95% CI 0.44–0.93) and appeared to be 442
driven by the use of oil and vinegar dressings[68]. The HPFS reported on a Mediterranean 443
diet and mortality in prostate cancer survivors after diagnosis[69]. Kenfield et al. compared 444
survivors with high adherence to a Mediterranean diet to those with low adherence and 445
demonstrated that post-diagnosis adherence to a Mediterranean diet was associated with 446
decreased overall mortality (HR= 0.78; 95% CI 0.67-0.90); no association was observed for 447
prostate cancer specific mortality and adherence to the Mediterranean diet[69]. A pre-448
diagnosis high fish consumption in men who were diagnosed with prostate cancer while 449
participating in the PHS was related to prolonged survival (HR=0.52; 95% CI 0.30-0.91) 450
according to Chavarro et al.[70]. Another study of Yang et al. investigated post-diagnosis 451
dairy intake amongst prostate cancer survivors[71]. The consumption of total dairy was non-452
beneficially associated with overall mortality (HR=1.76; 95% CI 1.21-2.55). Both high-fat 453
and low-fat dairy consumption contributed to this adverse association and overall mortality 454
(respectively HR=1.22; 95% CI 1.08-1.38 and HR=1.17; 95% CI 1.05-1.29)[71]. 455
In summary, no conclusive evidence of an association with mortality was observed 456
for dietary patterns/indices and foods of the main food groups amongst prostate cancer 457
survivors. 458
459
Page 20 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
21
DISCUSSION 460
This systematic review included the current scientific literature with regard to dietary 461
patterns/indices and foods from the main food groups and mortality and cancer recurrence 462
amongst different groups of cancer survivors. In summary, we have demonstrated 463
inconsistent patters across the cancer types investigated. 464
465
Dietary patterns/indices 466
It could be speculated that the lack of effect in the two identified RCTs investigating a 467
low-fat diet in breast cancer survivors is a consequence of the relatively short follow-up 468
period when using mortality as the primary outcome. The true beneficial effect of diet 469
remains uncertain since increased exercise and weight loss during the intervention may also 470
have advantaged these breast cancer survivors. The beneficial association between a high 471
adherence to a high-quality diet and a prudent diet and the inverse association with the 472
Western diet and mortality could be explained by the general assumption that adherence to a 473
diet rich in fruit and vegetables could increase overall life expectancy. The limited number of 474
studies indicate that additional long-term prospective cohort studies are urgently needed to 475
improve the strength of evidence on the influence of dietary pattern adherence on cancer 476
survival. 477
478
Foods from the main food groups 479
The investigated healthy dietary patterns/indices are characterised by foods of the 480
main food groups. Epidemiological research on fruit and vegetable intake and cancer risk 481
increased rapidly over the last few decades and it has been suggested that people with high 482
intakes of fruit and vegetables, compared to those with low intakes, have a reduced risk of 483
developing cancer[72]. The wide variety of nutrients including vitamins, minerals, 484
Page 21 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
22
phytochemicals and fibre in fruit and vegetables could influence epigenetic processes and 485
potentially via this way improve cancer outcomes[73,74]. However, the exact mechanisms of 486
how diet can alter genetic and epigenetic changes in cancer cells has yet to be established. 487
Even though the majority of the identified studies found no ‘statistically significant’ 488
association, based on a p value that indicates the degree to which the data conform to the 489
pattern predicted by the test hypothesis and all the other assumptions used in the test, between 490
fruit and/or vegetable consumption and survivors’ mortality, two studies, that were evaluated 491
as ‘low’ evidence, suggested that pre-diagnosis fruit intake could decrease overall mortality 492
amongst breast cancer survivors. Adherence to a post-diagnosis high-quality diet and prudent 493
diet, characterized by a high intake of fruit and vegetables, appears to decrease mortality 494
amongst these survivors. Therefore, consumption of a wide variety of fruit and vegetables 495
should be encouraged in breast cancer survivors as they are an important part of a healthy diet 496
to maintain general health and could increase survival after cancer diagnosis. 497
Pre-diagnosis processed meat intake appears to increase mortality amongst bowel 498
cancer survivors. Experts from the World Cancer Research Fund and American Institute for 499
Cancer Research determined that the consumption of processed meat (containing haem, 500
nitrates and nitrites) is a risk factor for the development of bowel cancer[72]. 501
Notwithstanding, red meat provides a useful source of protein, iron and zinc, and eating not 502
more than 70 grams per day (as advised by the UK Department of Health) is compatible with 503
a healthy diet[75]. Adherence to a pre- and post-diagnosis Western diet, characterized by a 504
high intake of processed meat, appears to increase mortality. Hence, a limited intake of 505
processed meats should be encouraged in both bowel and breast cancer survivors to possibly 506
increase survival. With vegetarianism rates increasing, future research could focus on 507
comparing vegetarians with individuals who eat meat to elucidate the relationship between 508
(processed) meat consumption and prolonging breast and bowel cancer survivorship. 509
Page 22 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
23
510
Study strengths and limitations 511
The strengths of this systematic review are the inclusion of dietary patterns/indices 512
and whole foods, and the large total number of cancer survivors investigated. Identifying 513
dietary patterns/indices and whole foods reflects a more real-life condition as most people eat 514
a variety of foods. By examining the whole diet, the intake of nutrients in combination is 515
considered which provides translatable real-life scenarios for clinical recommendations. 516
The limitations of this systematic review were the inclusion of only 2 RCTs, the few 517
studies investigating post-diagnosis intake, the use of FFQs to collect dietary information 518
from participants in most studies, and the considerable heterogeneity in study design and 519
participant characteristics (tumour characteristics (stage/grade), treatment, age, time of 520
follow-up, comorbidity, differences in countries and ethnicity). Due to potential bias, data 521
from observational studies generally provide a lower strength of evidence than from RCTs, 522
even if they were well conducted. Conducting RCTs to investigate dietary intake in cancer 523
survivors with mortality as an outcome can be challenging for cancers with a relatively long 524
survival necessitating adherence to a diet in the long-term. The majority of studies included 525
in this systematic review investigated dietary patterns/indices and foods before cancer 526
diagnosis, with only a few studies in the post-diagnosis setting. Information on dietary intake 527
after diagnosis is valuable for investigating the effect of dietary changes on outcomes 528
amongst cancer survivors. It is too late to amend lifestyle factors before diagnosis but patients 529
may be more receptive to advice after diagnosis, and so further studies are needed in this 530
setting. 531
Although the use of FFQs is an inexpensive approach to capture data from hundreds 532
or thousands of individuals, it may not represent the usual foods or portion sizes chosen by 533
participants, and intake data can be compromised when multiple foods are grouped with 534
Page 23 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
24
single listings. Developments in the screening, diagnosis and treatment of cancers differ 535
greatly between countries and therefore could influence survival. Although most studies are 536
adjusted for tumour stage, age and treatment, often no adjustments could be made for 537
influential lifestyle factors including BMI, physical activity and smoking. It remains a 538
challenge to disentangle the impact of diet from other lifestyle factors, and this should always 539
be taken into consideration when interpreting study results. 540
541
CONCLUSION 542
It is important that the results of well-conducted scientific studies, although limited, reach 543
health care providers so that cancer survivors can be informed of dietary factors that could 544
possibly influence their outcomes. For many cancer survivors, there is little evidence to date 545
to indicate that particular dietary behaviours influence outcomes with regard to mortality and 546
recurrence. Notwithstanding, limited evidence suggests that pre-diagnosis fruit and a high-547
quality and prudent diet are beneficial for breast cancer survivors, and that a Western diet is 548
detrimental for breast and bowel cancer survivors. 549
550
551
Authorship contribution statement 552
SJ drafted the manuscript and worked on the conception, design and interpretation of data. SJ 553
and FvO selected articles, screened titles and abstracts, assessed study quality and extracted 554
data. SJ, FvO, RB and MZ were involved in the interpretation and discussion of the results 555
and critically revised the systematic review for important intellectual content. All authors, SJ, 556
FvO, RB, AW, FJvS, KKC and MZ, approved the final version of the systematic review. SJ 557
is the guarantor. 558
559
Page 24 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
25
Conflict of interest 560
None of the authors have any conflict of interest in connection with this systematic review. 561
562
Funding 563
This research received no grant from any funding agency in the public, commercial or not-564
for-profit sectors. 565
566
Data sharing statement 567
No additional data available. 568
569
Page 25 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
26
LITERATURE 570
1 Parkin DM. 1. The fraction of cancer attributable to lifestyle and environmental factors 571
in the UK in 2010. Br J Cancer 2011;105:S2–5. doi:10.1038/bjc.2011.474 572
2 Leenders M, Sluijs I, Ros MM, et al. Fruit and vegetable consumption and mortality: 573
European prospective investigation into cancer and nutrition. Am J Epidemiol 574
2013;178:590–602. doi:10.1093/aje/kwt006 575
3 Nguyen CT, Pham NM, Lee AH, et al. Prevalence of and Risk Factors for Type 2 576
Diabetes Mellitus in Vietnam: A Systematic Review. Asia Pac J Public Heal 577
2015;27:588–600. doi:10.1177/1010539515595860 578
4 Bours MJ, Beijer S, Winkels RM, et al. Dietary changes and dietary supplement use, 579
and underlying motives for these habits reported by colorectal cancer survivors of the 580
Patient Reported Outcomes Following Initial Treatment and Long-Term Evaluation of 581
Survivorship (PROFILES) registry. Br J Nutr 2015;114:286–96. 582
doi:10.1017/S0007114515001798 583
5 Humpel N, Magee C, Jones SC. The impact of a cancer diagnosis on the health 584
behaviors of cancer survivors and their family and friends. Support Care Cancer 585
2007;15:621–30. doi:10.1007/s00520-006-0207-6 586
6 Zhang FF, Liu S, John EM, et al. Diet quality of cancer survivors and noncancer 587
individuals: Results from a national survey. Cancer 2015;121:4212–21. 588
doi:10.1002/cncr.29488 589
7 D’Avanzo B, La Vecchia C, Negri E, et al. Attributable risks for bladder cancer in 590
northern Italy. Ann Epidemiol 1995;5:427–31. 591
8 Rock CL, Doyle C, Demark-Wahnefried W, et al. Nutrition and physical activity 592
guidelines for cancer survivors. CA Cancer J Clin 2012;62:242–74. 593
doi:10.3322/caac.21142 594
9 World Cancer Research Fund International / American Institute for Cancer Research. 595
Diet, Nutrition, Physical Activity, and Breast Cancer Survivors. 2014. 596
http://www.wcrf.org/sites/default/files/Breast-Cancer-Survivors-2014-Report.pdf 597
(accessed 2 Jun 2017). 598
10 Parsons JK, Pierce JP, Natarajan L, et al. NIH Public Access. 2014;6:1–13. 599
doi:10.1158/1940-6207.CAPR-13-0050.A 600
11 Schwingshackl L, Chaimani A, Bechthold A, et al. Food groups and risk of chronic 601
disease: A protocol for a systematic review and network meta-analysis of cohort 602
studies. Syst Rev 2016;5:1.http://www.systematicreviewsjournal.com/ 603
http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=emed18b&NEWS=604
N&AN=611368739 605
12 Hanahan D, Weinberg RA. Hallmarks of Cancer: The Next Generation. Cell 606
2011;144:646–74. doi:10.1016/j.cell.2011.02.013 607
13 Cancer Research UK. Cancer survival statistics. 608
http://www.cancerresearchuk.org/health-professional/cancer-statistics/survival 609
(accessed 10 Jun 2017). 610
Page 26 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
27
14 Reedy J, Wirfält E, Flood A, et al. Comparing 3 dietary pattern methods--cluster 611
analysis, factor analysis, and index analysis--With colorectal cancer risk: The NIH-612
AARP Diet and Health Study. Am J Epidemiol 2010;171:479–87. 613
doi:10.1093/aje/kwp393 614
15 Guenther PM, Reedy J, Krebs-Smith SM. Development of the Healthy Eating Index-615
2005. J Am Diet Assoc 2008;108:1896–901. doi:10.1016/j.jada.2008.08.016 616
16 Guenther PM, Reedy J, Krebs-Smith SM, et al. Evaluation of the Healthy Eating 617
Index-2005. J Am Diet Assoc 2008;108:1854–64. doi:10.1016/j.jada.2008.08.011 618
17 Chiuve SE, Fung TT, Rimm EB, et al. Alternative dietary indices both strongly predict 619
risk of chronic disease. J Nutr 2012;142:1009–18. doi:10.3945/jn.111.157222 620
18 Guenther PM, Kirkpatrick SI, Reedy J, et al. The Healthy Eating Index-2010 Is a Valid 621
and Reliable Measure of Diet Quality According to the 2010 Dietary Guidelines for 622
Americans. J Nutr 2014;144:399–407. doi:10.3945/jn.113.183079 623
19 Inoue-Choi M, Lazovich D, Prizment AE, et al. Adherence to the World Cancer 624
Research Fund/American Institute for Cancer Research Recommendations for Cancer 625
Prevention Is Associated With Better Health-Related Quality of Life Among Elderly 626
Female Cancer Survivors. J Clin Oncol 2013;31:1758–66. 627
doi:10.1200/JCO.2012.45.4462 628
20 Catsburg C, Miller AB, Rohan TE. Adherence to cancer prevention guidelines and risk 629
of breast cancer. Int J Cancer 2014;135:2444–52. doi:10.1002/ijc.28887 630
21 Kant AK, Schatzkin A, Graubard BI, et al. A prospective study of diet quality and 631
mortality in women. JAMA 2000;283:2109–632
15.http://www.ncbi.nlm.nih.gov/pubmed/10791502 (accessed 10 Jun 2017). 633
22 Haines PS, Siega-Riz AM, Popkin BM. The Diet Quality Index revised: a 634
measurement instrument for populations. J Am Diet Assoc 1999;99:697–704. 635
doi:10.1016/S0002-8223(99)00168-6 636
23 Fung TT, Chiuve SE, McCullough ML, et al. Adherence to a DASH-Style Diet and 637
Risk of Coronary Heart Disease and Stroke in Women. Arch Intern Med 638
2008;168:713. doi:10.1001/archinte.168.7.713 639
24 Olsen A, Egeberg R, Halkjær J, et al. Healthy aspects of the Nordic diet are related to 640
lower total mortality. J Nutr 2011;141:639–44. doi:10.3945/jn.110.131375 641
25 Fung TT, Kashambwa R, Sato K, et al. Post diagnosis diet quality and colorectal 642
cancer survival in women. PLoS One 2014;9:1–13. doi:10.1371/journal.pone.0115377 643
26 Kim EHJ, Willett WC, Fung T, et al. Diet quality indices and postmenopausal breast 644
cancer survival. Nutr Cancer 2011;63:381–8. doi:10.1080/01635581.2011.535963 645
27 Sofi F, Macchi C, Abbate R, et al. Mediterranean diet and health status: an updated 646
meta-analysis and a proposal for a literature-based adherence score. Public Health 647
Nutr 2014;17:2769–82. doi:10.1017/S1368980013003169 648
28 Schwingshackl L, Hoffmann G. Diet Quality as Assessed by the Healthy Eating Index, 649
the Alternate Healthy Eating Index, the Dietary Approaches to Stop Hypertension 650
Page 27 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
28
Score, and Health Outcomes: A Systematic Review and Meta-Analysis of Cohort 651
Studies. J Acad Nutr Diet 2015;115:780–800.e5. doi:10.1016/j.jand.2014.12.009 652
29 Public Health England. The Eatwell Guide. 653
2017.https://www.gov.uk/government/publications/the-eatwell-guide (accessed 10 Jun 654
2017). 655
30 Sterne J, Julian H, Barney R. The Cochrane Collaboration’s tool for assessing risk of 656
bias Description. Handb CochraneTable 85.a 657
2014;:4.http://handbook.cochrane.org/chapter_8/table_8_5_a_the_cochrane_collaborat658
ions_tool_for_assessing.htm 659
31 Sterne JA, Hernán MA, Reeves BC, et al. The Risk Of Bias In Non-randomized 660
Studies – of Interventions (ROBINS-I) assessment tool ROBINS-I tool (Stage I): At 661
protocol stage. 2016;:1–22. doi:10.1136/bmj.i4919 662
32 Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating 663
quality of evidence and strength of recommendations. BMJ 2008;336:924–6. 664
doi:10.1136/bmj.39489.470347.AD 665
33 Moher D. PRISMA 2009 Checklist PRISMA 2009 Checklist. PLoS Med 666
2009;6:e1000097. doi:10.1371/journal.pmed1000097 667
34 Higgins JP, Savović J, Page MJ, et al. Revised Cochrane risk of bias tool for 668
randomized trials (RoB 2.0). 2016;:52. 669
35 Sterne J, Higgins J, Reeves B, et al. A tool for evaluating risk of bias in non- 670
randomized studies of interventions ( ROBINS-I ): development and applications. 671
2015. 672
36 Greenland S, Senn SJ, Rothman KJ, et al. Statistical tests, P values, confidence 673
intervals, and power: a guide to misinterpretations. Eur J Epidemiol 2016;31:337–50. 674
doi:10.1007/s10654-016-0149-3 675
37 Tang L, Zirpoli GR, Guru K, et al. Intake of cruciferous vegetables modifies bladder 676
cancer survival. Cancer Epidemiol Biomarkers Prev 2010;19:1806–11. 677
doi:10.1158/1055-9965.EPI-10-0008 678
38 Meyerhardt JA, Niedzwiecki D, Hollis D, et al. Association of Dietary Patterns With 679
Cancer Recurrence and Survival in Patients With Stage III Colon Cancer. Jama 680
2007;298:754. doi:10.1001/jama.298.7.754 681
39 Zhu Y, Wu H, Wang PP, et al. Dietary patterns and colorectal cancer recurrence and 682
survival: a cohort study. BMJ Open 2013;3:e002270. doi:10.1136/bmjopen-2012-683
002270 684
40 Carr PR, Jansen L, Walter V, et al. Associations of red and processed meat with 685
survival after colorectal cancer and differences according to timing of dietary 686
assessment. Am J Clin Nutr 2016;103:192–200. doi:10.3945/ajcn.115.121145 687
41 McCullough ML, Gapstur SM, Shah R, et al. Association between red and processed 688
meat intake and mortality among colorectal cancer survivors. J Clin Oncol 689
2013;31:2773–82. doi:10.1200/JCO.2013.49.1126 690
Page 28 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
29
42 Yang B, McCullough ML, Gapstur SM, et al. Calcium, vitamin D, dairy products, and 691
mortality among colorectal cancer survivors: The cancer prevention study-II nutrition 692
cohort. J Clin Oncol 2014;32:2335–43. doi:10.1200/JCO.2014.55.3024 693
43 Skeie G, Braaten T, Olsen A, et al. Whole grain intake and survival among 694
Scandinavian colorectal cancer patients. Nutr Cancer 2014;66:6–13. 695
doi:10.1080/01635581.2014.847472 696
44 Pelser C, Arem H, Pfeiffer RM, et al. Prediagnostic lifestyle factors and survival after 697
colon and rectal cancer diagnosis in the National Institutes of Health (NIH)-AARP 698
Diet and Health Study. Cancer 2014;120:1540–7. doi:10.1002/cncr.28573 699
45 Ratjen I, Schafmayer C, di Giuseppe R, et al. Postdiagnostic Mediterranean and 700
Healthy Nordic Dietary Patterns Are Inversely Associated with All-Cause Mortality in 701
Long-Term Colorectal Cancer Survivors. J Nutr 2017;147:636–44. 702
doi:10.3945/jn.116.244129 703
46 Romaguera D, Ward H, Wark PA, et al. Pre-diagnostic concordance with the 704
WCRF/AICR guidelines and survival in European colorectal cancer patients: a cohort 705
study. BMC Med 2015;13:107. doi:10.1186/s12916-015-0332-5 706
47 Ward HA, Norat T, Overvad K, et al. Pre-diagnostic meat and fibre intakes in relation 707
to colorectal cancer survival in the European Prospective Investigation into Cancer and 708
Nutrition. Br J Nutr 2016;116:316–25. doi:10.1017/S0007114516001859 709
48 Chlebowski RT, Blackburn GL, Thomson CA, et al. Dietary fat reduction and breast 710
cancer outcome: interim efficacy results from the Women’s Intervention Nutrition 711
Study. J Natl Cancer Inst 2006;98:1767–76. doi:10.1093/jnci/djj494 712
49 Pierce JP, Natarajan L, Caan BJ, et al. Influence of a Diet Very High in Vegetables, 713
Fruit, and Fiber and Low in Fat on Prognosis Following Treatment for Breast Cancer. 714
Jama 2007;298:289. doi:10.1001/jama.298.3.289 715
50 George SM, Irwin ML, Smith AW, et al. Postdiagnosis diet quality, the combination of 716
diet quality and recreational physical activity, and prognosis after early-stage breast 717
cancer. Cancer Causes Control 2011;22:589–98. doi:10.1007/s10552-011-9732-9 718
51 George SM, Ballard-Barbash R, Shikany JM, et al. Better postdiagnosis diet quality is 719
associated with reduced risk of death among postmenopausal women with invasive 720
breast cancer in the women’s health initiative. Cancer Epidemiol Biomarkers Prev 721
2014;23:575–83. doi:10.1158/1055-9965.EPI-13-1162 722
52 Izano MA, Fung TT, Chiuve SS, et al. Are Diet Quality Scores After Breast Cancer 723
Diagnosis Associated with Improved Breast Cancer Survival? Nutr Cancer 724
2013;65:820–6. doi:10.1080/01635581.2013.804939 725
53 McCullough ML, Gapstur SM, Shah R, et al. Pre- and postdiagnostic diet in relation to 726
mortality among breast cancer survivors in the CPS-II Nutrition Cohort. Cancer 727
Causes Control 2016;27:1303–14. doi:10.1007/s10552-016-0802-x 728
54 Kroenke CH, Fung TT, Hu FB, et al. Dietary patterns and survival after breast cancer 729
diagnosis. J Clin Oncol 2005;23:9295–303. doi:10.1200/JCO.2005.02.0198 730
55 Kwan ML, Weltzien E, Kushi LH, et al. Dietary patterns and breast cancer recurrence 731
Page 29 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
30
and survival among women with early-stage breast cancer. J Clin Oncol 2009;27:919–732
26. doi:10.1200/JCO.2008.19.4035 733
56 Vrieling A, Buck K, Seibold P, et al. Dietary patterns and survival in German 734
postmenopausal breast cancer survivors. Br J Cancer 2013;108:188–92. 735
doi:10.1038/bjc.2012.521 736
57 McEligot AJ, Largent J, Ziogas A, et al. Dietary fat, fiber, vegetable, and 737
micronutrients are associated with overall survival in postmenopausal women 738
diagnosed with breast cancer. Nutr Cancer 2006;55:132–40. 739
doi:10.1207/s15327914nc5502_3 740
58 Dal Maso L, Zucchetto A, Talamini R, et al. Effect of obesity and other lifestyle 741
factors on mortality in women with breast cancer. Int J cancer 2008;123:2188–94. 742
doi:10.1002/ijc.23747 743
59 Holmes MD, Stampfer MJ, Colditz GA, et al. Dietary factors and the survival of 744
women with breast carcinoma. Cancer 1999;86:826–35. doi:10.1002/(SICI)1097-745
0142(19990901)86:5<826::AID-CNCR19>3.0.CO;2-0 746
60 Kroenke CH, Kwan ML, Sweeney C, et al. High-and low-fat dairy intake, recurrence, 747
and mortality after breast cancer diagnosis. J Natl Cancer Inst 2013;105:616–23. 748
doi:10.1093/jnci/djt027 749
61 Beasley JM, Newcomb PA, Trentham-Dietz A, et al. Post-diagnosis dietary factors 750
and survival after invasive breast cancer. Breast Cancer Res Treat 2011;128:229–36. 751
doi:10.1007/s10549-010-1323-z 752
62 Buck K, Zaineddin AK, Vrieling A, et al. Estimated enterolignans, lignan-rich foods, 753
and fibre in relation to survival after postmenopausal breast cancer. Br J Cancer 754
2011;105:1151–7. doi:10.1038/bjc.2011.374 755
63 Hebert JR, Hurley TG, Ma Y. The effect of dietary exposures on recurrence and 756
mortality in early stage breast cancer. Breast Cancer Res. Treat. 1998;51:17–28. 757
doi:10.1023/A:1006056915001 758
64 Crosignani P, Russo A, Tagliabue G, et al. Tobacco and diet as determinants of 759
survival in male laryngeal cancer patients. Int J Cancer 1996;65:308–13. 760
doi:10.1002/(SICI)1097-0215(19960126)65:3<308::AID-IJC5>3.0.CO;2-3 761
65 Han X, Zheng T, Foss F, et al. Vegetable and fruit intake and non-Hodgkin lymphoma 762
survival in Connecticut women. Leuk Lymphoma 2010;51:1047–54. 763
doi:10.3109/10428191003690364 764
66 Leo QJN, Ollberding NJ, Wilkens LR, et al. Nutritional factors and non-Hodgkin 765
lymphoma survival in an ethnically diverse population: the Multiethnic Cohort. Eur J 766
Clin Nutr 2016;70:41–6. doi:10.1038/ejcn.2015.139 767
67 Yang M, Kenfield SA, Van Blarigan EL, et al. Dietary patterns after prostate cancer 768
diagnosis in relation to disease-specific and total mortality. Cancer Prev Res (Phila) 769
2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 770
68 Yang M, Kenfield SA, Van Blarigan EL, et al. Dietary Patterns after Prostate Cancer 771
Diagnosis in Relation to Disease-Specific and Total Mortality. Cancer Prev Res 772
Page 30 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
31
2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 773
69 Kenfield SA, DuPre N, Richman EL, et al. Mediterranean diet and prostate cancer risk 774
and mortality in the Health Professionals Follow-up Study. Eur Urol 2014;65:887–94. 775
doi:10.1016/j.eururo.2013.08.009 776
70 Chavarro JE, Stampfer MJ, Hall MN, et al. A 22-y prospective study of fish intake in 777
relation to prostate cancer incidence and mortality. Am J Clin Nutr 2008;88:1297–778
303.http://www.ncbi.nlm.nih.gov/pubmed/18996866 (accessed 10 Jun 2017). 779
71 Yang M, Kenfield SA, Van Blarigan EL, et al. Dairy intake after prostate cancer 780
diagnosis in relation to disease-specific and total mortality. Int J Cancer 781
2015;137:2462–9. doi:10.1002/ijc.29608 782
72 Wiseman M. The second World Cancer Research Fund/American Institute for Cancer 783
Research expert report. Food, nutrition, physical activity, and the prevention of cancer: 784
a global perspective. Proc Nutr Soc 2008;67:253–6. doi:10.1017/S002966510800712X 785
73 Hardy TM, Tollefsbol TO. Epigenetic diet: impact on the epigenome and cancer. 786
Epigenomics 2011;3:503–18. doi:10.2217/epi.11.71 787
74 Daniel M, Tollefsbol TO. Epigenetic linkage of aging, cancer and nutrition. J Exp Biol 788
2015;218:59–70. doi:10.1242/jeb.107110 789
75 Department of Health UK. Red meat and bowel cancer risk. 790
http://www.nhs.uk/Livewell/Goodfood/Pages/red-meat.aspx (accessed 10 Jun 2017). 791
792
Page 31 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
32
Table 1: Number of studies investigating the association between pre-diagnosis dietary patterns/indices and mortality/cancer recurrence in different populations of cancer survivors
Diet quality indices Prudent / healthy diet Western diet / unhealthy diet
Cancer site /type No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 0 - - - - 0 - - - - 0 - - - -
Bowel 2 - 2 1 - 1 - 1 1 - 1 - 2 2 -
Breast 1 - 1 1 1 2 1 4 4 4 2 1 4 4 4
Cervix 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - -
Larynx 0 - - - - 0 - - - - 0 - - - -
MM 0 - - - - 0 - - - - 0 - - - -
NHL 0 - - - - 0 - - - - 0 - - - -
Prostate 0 - - - - 0 - - - - 0 - - - -
Testes 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - -
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and dietary patterns in the same population
Page 32 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
33
Table 2: Number of studies investigating the association between post-diagnosis dietary patterns/indices and mortality/cancer recurrence in different populations of cancer survivors
Diet quality indices Prudent diet / healthy diet Western diet / unhealthy
Cancer site /type No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 0 - - - - 0 - - - - 0 - - -
Bowel 2 - 5 3 - 2 1 2 1 - 2 1 2 1 -
Breast 7 - 11 9 8 2 1 2 2 2 2 1 2 2 2
Cervix 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - -
Larynx 0 - - - - 0 - - - - 0 - - - -
MM 0 - - - - 0 - - - - 0 - - - -
NHL 0 - - - - 0 - - - - 0 - - - -
Prostate 1 - 1 1 - 1 - 1 1 - 1 - 1 1 -
Testes 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - -
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and dietary patterns in the same population
Page 33 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
34
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and food items in the same population
Table 3: Number of studies investigating the association between pre-diagnosis foods and mortality/cancer recurrence in different populations of cancer survivors
Fruit and vegetables Grain foods Protein foods Dairy and alternative products Oils and spreads
Cancer site
/type
No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 1 - 4 4 - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Bowel 0 - - - - 1 - 5 - - 3 1 8 8 3 2 - 6 6 - 0 - - - -
Breast 4 - 6 4 - 1 - 1 1 - 2 - 5 5 - 1 - 1 1 - 0 - - - -
Cervix 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Larynx 1 - 3 - - 1 - 3 - - 1 - 4 - - 1 - 2 - - 1 - 2 - -
MM 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
NHL 2 - 11 11 - 0 - - - - 1 - 5 5 - 1 - 1 1 - 0 - - - -
Prostate 0 - - - - 0 - - - - 1 - - 1 - 0 - - - - 0 - - - -
Testes 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Page 34 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
35
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and food items in the same population
Table 4: Number of studies investigating the association between post-diagnosis foods and mortality/cancer recurrence in different populations of cancer survivors
Fruit and vegetables Grain foods Protein foods Dairy and alternative products Oils and spreads
Cancer site
/type
No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Bowel 0 - - - - 0 - - - - 1 - 3 3 3 1 - 2 2 - 0 - - - -
Breast 3 1 5 4 - 0 - - - - 3 1 4 5 - 3 3 5 2 - 1 1 - 1 -
Cervix 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Larynx 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
MM 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
NHL 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Prostate 0 - - - - 0 - - - - 0 - - - - 1 - 3 - - 0 - - - -
Testes 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Page 35 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
36
Box 1: literature search for the Pubmed database addressing the relationship between diet and mortality among bladder cancer survivors
(“urinary bladder neoplasms”[Mesh] OR “urinary bladder neoplasm*” OR “bladder cancer*” OR “bladder tumor*” OR “bladder tumour*”)
AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR
“disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-
meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet,
vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat
diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh]
OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR
“red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish
products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR
“nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding
Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR
"Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang]
Page 36 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Online supporting data File S1 for manuscript ‘the impact of dietary patterns and the
main food groups on mortality and recurrence in cancer survivors: systematic review
of current epidemiological literature’
Jochems et al., 31-05-2017
Page 37 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
2
Protocol Systematic Literature Review
INTRODUCTION Title The impact of diet on mortality in cancer survivors: A systematic review of current epidemiological literature Authors Sylvia H.J. Jochems, Frits H.M. van Osch, Richard T. Bryan, Anke Wesselius, Frederik J. van Schooten, K.K. Cheng, Maurice P. Zeegers Article type Systematic literature review (SLR) Language article English
Actual start date 01 December 2014 Updated 07 Mei 2017
- included one additional exclusion criteria: sample size had to be > 200 survivors - exclusion beverages for more clear focus of systematic review - inclusion dietary indices and grain products as an exposure - inclusion cancer recurrence as an outcome (these adjustments had influence on the outcome, exposure and search terms and were accordingly adjusted)
Funding sources No Conflicts of interest No
Research question The aim of this study was to conduct a structured summary and evaluation of randomised controlled trials and observational studies addressing the relationship between the highest versus the lowest intake of dietary patterns/indices and foods of the main food groups and mortality and cancer recurrence amongst groups of survivors of common cancers with a ten-year survival rate of at least 50%. PICO model Population: cancer survivors Intervention/exposure: dietary patterns/indices and foods from the main food groups Comparator/control: highest versus lowest intake Outcome: mortality (overall, cancer-specific, death from other causes) and cancer recurrence
Page 38 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
3
METHODS Database search
Pubmed / Medline (1966 - May 2017)
Embase (1980 – May 2017)
Cochrane Library (1993 – May 2017) Additional search Hand searching and reference tracking on included and related articles, systematic reviews and meta-analyses Study types
Inclusion criteria
Randomized Controlled Trials
Cross-sectional studies
Cohort studies - retrospective & longitudinal
Case-control (including follow-up of cases)
Exclusion criteria
Animal studies
In vitro studies
Gene-nutrient interaction studies Population / participants
Inclusion criteria
Adult population, at least 18 years of age (both men and women)
Survivors of common cancers with a ten-year survival of at least 50% including bladder, bowel, breast, cervical, kidney, laryngeal, prostate, testicular, uterine cancer, malignant melanoma, and (non-)Hodgkin lymphoma
Exclusion criteria
Pre-cancerous conditions of other cancer types
Combination of different types of cancers Study characteristics
Exclusion criteria
Sample size of at least 200 survivors in the analysis [http://www.tandfonline.com/toc/hsem20/current]
Follow-up period of at least 4 years (the risk of cancer recurrence is the greatest within the first three years for most cancers) [http://www.cancerresearchuk.org/about-cancer/what-is-cancer/why-some-cancers-come-back]
Adjustments had to be made for at least age, tumour characteristics (stage/grade), and preferably initial treatment, in the statistical analysis
Page 39 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4
Exposure / intervention Dietary patterns/indices and foods from the main food groups including:
(I) Dietary patterns that were considered were assessed by index-based methods and data-driven approaches, such as principal component analysis (factor analysis) and cluster analysis
(II) fruit and vegetables including citrus fruits, stone fruits, soft fruits, fleshy fruits, vine fruits, flower vegetables, leafy vegetables, stem vegetables, fruit vegetables, mushrooms, bulbs and roots;
(III) grain foods including potatoes, bread, rice, pasta and cereal; (IV) protein foods including meat (processed meat, unprocessed meat, red meat,
poultry), fish, eggs, tofu, nuts, seeds, pulses, legumes and beans; (V) dairy and alternative products including yoghurt, milk, cheese; (VI) oils and spreads including vegetable oils, spreads
Comparators / control Highest compared to lowest intake category to dietary patterns/indices and foods from the main food groups Outcomes Primary outcomes Overall mortality, cancer-specific mortality, death from other causes, and cancer recurrence Secondary outcomes None
Analysis (consideration of a meta-analysis besides a systematic review) We expect a lot of diversity in pre- and post-diagnosis dietary patterns/indices and foods and different cancers, and decided to only consider comparable studies (same timeframe pre- or post-diagnosis and same cancer type) for meta-analysis. If more than 75% of the review will not have 3 or more studies that can be pooled, only a systematic review and no meta-analysis will be conducted.
Article selection Inclusion criteria
Investigate the associations between dietary patterns/indices and foods, and mortality and cancer recurrence in survivors of primary cancer
Report a measure of the effect/association of the exposure on the outcomes
The duration of the exposure/intervention had to be recorded as well as the time between exposure assessment / intervention and outcome assessment
Diet and lifestyle modifications/changes consequent on the disease or its treatment will not be included
Present results of primary and secondary analysis
Present results for any of the following outcomes: • Overall mortality • Cancer-specific mortality • Death from other causes
Page 40 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
5
• Cancer recurrence
Original articles published in peer-reviewed journals
Articles written in the English language Exclusion criteria
Systematic reviews, meta-analysis Reviews, comments, letters, conference abstracts
Data extraction The data extraction will be performed independently by two of the authors (Sylvia Jochems and Frits van Osch) and any disagreements about study inclusion will be resolved through consensus or, if necessary, a third party (Rik Bryan). Information to extract from studies: Author, Study, Country, Number of participants, sex, age, follow-up period, exposure, exposure timeframe, exposure assessment, outcome, results (HR/RR and 95% CI), adjustments in the statistical analysis.
Tools for assessing Risk of Bias and Level of Quality The Cochrane Collaboration risk of bias assessment tools were used for appraisal of RCTs and cohort studies. For RCTs the RoB 2.0 tool (a revised tool for risk of bias in randomized trials) will be used to evaluate the risk of bias. Cohort studies will be evaluated with an adjusted version of the ROBINS-I tool [http://methods.cochrane.org/bias/risk-bias-non-randomized-studies-interventions]. Levels of quality were determined with the GRADE system [http://handbook.cochrane.org/chapter_12/12_2_1_the_grade_approach.htm].
General search terms in Pubmed
1. Searching for all studies relating to cancer and survival: Neoplasms, neoplasm staging, neoplasm recurrence local, neoplasia, tumours, cancer, survivors, survival analysis, recurrence, mortality, survival rate, disease management
2. Searching for all studies relating to dietary modification: Foods from the main food groups, dietary indices, dietary patterns
3. Selecting randomised control trials: Randomized controlled trial, random allocation, double blind method, single-blind method, clinical trial
4. Selecting cohort studies: Epidemiologic studies, cohort studies, follow-up studies, longitudinal studies, prospective studies, retrospective studies
5. Additional filters:
English language, human, full text The search terms will be adapted for use in the Ovid database (EMBASE and Cochrane library).
Page 41 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
6
Specific search terms per cancer site in Pubmed The search strategy will be adapted for use in EMBASE and the Cochrane Library
1. Searching for all studies relating to bladder cancer: (“urinary bladder neoplasms”[Mesh] OR “urinary bladder neoplasm*” OR “bladder cancer*” OR “bladder tumor*” OR “bladder tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 42 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7
2. Searching for all studies relating to breast cancer: (“breast neoplasms/diet therapy”[Mesh] OR “breast neoplasms/mortality”[Mesh] OR “breast neoplasms/prevention and control”[Mesh] OR “mammary neoplasm*” OR “breast neoplasm*” OR “mammary cancer*” OR “breast cancer*” OR “breast carcinoma*” OR “human mammary carcinoma*” OR “breast tumor*” OR “breast tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 43 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
8
3. Searching for all studies relating to cervical cancer: (“uterine cervical neoplasms/diet therapy”[Mesh] OR “uterine cervical neoplasms/mortality”[Mesh] OR “uterine cervical neoplasms/prevention and control”[Mesh] OR “uterine cervical neoplasm*” OR “cervical cancer*” OR “cervical neoplasm*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 44 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
9
4. Searching for all studies relating to bowel cancer:
(“colorectal neoplasms/diet therapy”[Mesh] OR “colorectal neoplasms/mortality”[Mesh] OR “colorectal neoplasms/prevention and control”[Mesh] OR “colorectal neoplasm*” OR “colorectal cancer*” OR “colorectal tumor*” OR “colorectal tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 45 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
10
5. Searching for all studies relating to Hodgkin lymphoma: (“Hodgkin disease/diet therapy”[Mesh] OR “Hodgkin disease/mortality”[Mesh] OR “Hodgkin disease/prevention and control”[Mesh] OR “Hodgkin disease” OR “Hodgkin Lymphoma” OR “Malignant Lymphogranuloma*” OR “Hodgkin lymphoma” OR “nodular lymphocyte predominant Hodgkin's lymphoma” OR “nodular sclerosing Hodgkin's lymphoma” OR “lymphocyte rich classical Hodgkin's lymphoma” OR “mixed cellularity Hodgkin's lymphoma”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 46 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
11
6. Searching for all studies relating to non-Hodgkin lymphoma:
(“lymphoma, non-Hodgkin/diet therapy”[Mesh] OR “lymphoma, non-Hodgkin/mortality”[Mesh] OR “lymphoma, non-Hodgkin/prevention and control”[Mesh] OR “lymphoma, non-Hodgkin” OR “nonhodgkins lymphoma” OR “non-Hodgkins lymphoma”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 47 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
12
7. Searching for all studies relating to kidney cancer: (“Kidney Neoplasms/diet therapy”[Mesh] OR “Kidney Neoplasms/mortality”[Mesh] OR “Kidney Neoplasms/prevention and control”[Mesh] OR “Kidney Neoplasm*” OR “Renal Neoplasm*” OR “Kidney Cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 48 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
13
8. Searching for all studies relating to larynx cancer:
(“laryngeal neoplasms/diet therapy”[Mesh] OR “laryngeal neoplasms/mortality”[Mesh] OR “laryngeal neoplasms/prevention and control”[Mesh] OR “laryngeal neoplasm*” OR “larynx cancer*” OR “larynx neoplasm*” OR “laryngeal cancer”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 49 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
14
9. Searching for all studies relating to multiple myeloma:
(“multiple myeloma/diet therapy”[Mesh] OR “multiple myeloma/mortality”[Mesh] OR “multiple myeloma/prevention and control”[Mesh] OR “multiple myeloma*” OR “plasma cell myeloma*” OR “myelomatosis” OR “myelomatoses” OR “Kahler Disease” OR “myeloma multiple”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 50 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
15
10. Searching for all studies relating to malignant melanoma: ("Melanoma/ diet therapy”[Mesh] OR “Melanoma/mortality”[Mesh] OR “Melanoma/prevention and control”[Mesh] OR “Malignant Melanoma*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 51 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
16
11. Searching for all studies relating to prostate cancer: (“prostatic neoplasms/diet therapy”[Mesh] OR “prostatic neoplasms/mortality”[Mesh] OR “prostatic neoplasms/prevention and control”[Mesh] OR “prostate neoplasm*” OR “prostatic neoplasm*” OR “prostate cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 52 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
17
12. Searching for all studies relating to testicular cancer:
(“testicular neoplasms/diet therapy”[Mesh] OR “testicular neoplasms/mortality”[Mesh] OR “testicular neoplasms/prevention and control”[Mesh] OR “testicular neoplasm*” OR “testicular tumor*” OR “testicular tumour*” OR “testis neoplasm*” OR “testis cancer*” OR “testicular cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 53 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
18
13. Searching for all studies relating to uterus cancer: (“uterine neoplasms/diet therapy”[Mesh] OR “uterine neoplasms/mortality”[Mesh] OR “uterine neoplasms/prevention and control”[Mesh] OR “uterus neoplasm*” OR “uterine neoplasm*” OR “uterus cancer*” OR “uterine cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 54 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
19
The Risk Of Bias In Non-randomized Studies – of Interventions (ROBINS-I) assessment tool (version for cohort-type studies) Version 19 September 2016
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License.
ROBINS-I tool (Stage I): At protocol stage
Specify the review question
Participants
Experimental
intervention
Comparator
Outcomes
List the confounding domains relevant to all or most studies
List co-interventions that could be different between intervention groups and that could
impact on outcomes
Page 55 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
20
ROBINS-I tool (Stage II): For each study
Specify a target randomized trial specific to the study
Design Individually randomized / Cluster randomized / Matched (e.g. cross-
over)
Participants
Experimental
intervention
Comparator
Is your aim for this study…? to assess the effect of assignment to intervention to assess the effect of starting and adhering to intervention
Specify the outcome
Specify which outcome is being assessed for risk of bias (typically from among those earmarked for
the Summary of Findings table). Specify whether this is a proposed benefit or harm of intervention.
Specify the numerical result being assessed
In case of multiple alternative analyses being presented, specify the numeric result (e.g. RR = 1.52
(95% CI 0.83 to 2.77) and/or a reference (e.g. to a table, figure or paragraph) that uniquely defines
the result being assessed.
Page 56 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
21
Preliminary consideration of confounders
Complete a row for each important confounding domain (i) listed in the review protocol; and (ii)
relevant to the setting of this particular study, or which the study authors identified as potentially
important. “Important” confounding domains are those for which, in the context of this study, adjustment is expected to lead to a clinically important change in the estimated effect of the intervention. “Validity” refers to whether the confounding variable or variables fully measure the domain, while “reliability” refers to the precision of the measurement (more measurement error means less reliability).
(i) Confounding domains listed in the review protocol
Confounding
domain
Measured
variable(s)
Is there evidence
that controlling for
this variable was
unnecessary?*
Is the confounding
domain measured
validly and
reliably by this
variable (or these
variables)?
OPTIONAL: Is
failure to adjust for
this variable
(alone) expected to
favour the
experimental
intervention or the
comparator?
Yes / No / No
information
Favour
experimental /
Favour comparator
/ No information
(ii) Additional confounding domains relevant to the setting of this particular study, or which
the study authors identified as important
Confounding
domain
Measured
variable(s)
Is there evidence
that controlling for
this variable was
unnecessary?*
Is the confounding
domain measured
validly and
reliably by this
variable (or these
variables)?
OPTIONAL: Is
failure to adjust for
this variable
(alone) expected to
favour the
experimental
intervention or the
Page 57 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
22
comparator?
Yes / No / No
information
Favour
experimental /
Favour comparator
/ No information
* In the context of a particular study, variables can be demonstrated not to be confounders and so not included in the analysis: (a) if they are not predictive of the outcome; (b) if they are not predictive of intervention; or (c) because adjustment makes no or minimal difference to the estimated
effect of the primary parameter. Note that “no statistically significant association” is not the same as “not predictive”.
Page 58 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
23
Preliminary consideration of co-interventions
Complete a row for each important co-intervention (i) listed in the review protocol; and (ii) relevant
to the setting of this particular study, or which the study authors identified as important. “Important” co-interventions are those for which, in the context of this study, adjustment is expected to lead to a clinically important change in the estimated effect of the intervention.
(i) Co-interventions listed in the review protocol
Co-intervention Is there evidence that controlling
for this co-intervention was
unnecessary (e.g. because it was
not administered)?
Is presence of this co-
intervention likely to favour
outcomes in the experimental
intervention or the comparator
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
(ii) Additional co-interventions relevant to the setting of this particular study, or which the
study authors identified as important
Co-intervention Is there evidence that controlling
for this co-intervention was
unnecessary (e.g. because it was
not administered)?
Is presence of this co-
intervention likely to favour
outcomes in the experimental
intervention or the comparator
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Page 59 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
24
Risk of bias assessment
Responses underlined in green are potential markers for low risk of bias, and responses in red are
potential markers for a risk of bias. Where questions relate only to sign posts to other questions, no
formatting is used.
Signalling questions Description Response
options
Bias due to confounding
1.1 Is there potential for confounding of the effect of
intervention in this study?
If N/PN to 1.1: the study can be considered to be at low risk of
bias due to confounding and no further signalling questions need
be considered
Y / PY / PN / N
If Y/PY to 1.1: determine whether there is a need to assess time-
varying confounding:
1.2. Was the analysis based on splitting participants’ follow
up time according to intervention received?
If N/PN, answer questions relating to baseline
confounding (1.4 to 1.6)
If Y/PY, go to question 1.3.
NA / Y / PY /
PN / N / NI
1.3. Were intervention discontinuations or switches likely to
be related to factors that are prognostic for the outcome?
If N/PN, answer questions relating to baseline
confounding (1.4 to 1.6)
If Y/PY, answer questions relating to both baseline and
time-varying confounding (1.7 and 1.8)
NA / Y / PY /
PN / N / NI
Questions relating to baseline confounding only
1.4. Did the authors use an appropriate analysis method that
controlled for all the important confounding domains?
NA / Y / PY / PN / N / NI
1.5. If Y/PY to 1.4: Were confounding domains that were
controlled for measured validly and reliably by the
variables available in this study?
NA / Y / PY / PN / N / NI
1.6. Did the authors control for any post-intervention
variables that could have been affected by the intervention?
NA / Y / PY / PN / N / NI
Questions relating to baseline and time-varying confounding
1.7. Did the authors use an appropriate analysis method that
controlled for all the important confounding domains and
for time-varying confounding?
NA / Y / PY / PN / N / NI
1.8. If Y/PY to 1.7: Were confounding domains that were
controlled for measured validly and reliably by the
variables available in this study?
NA / Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious /
Critical / NI
Optional: What is the predicted direction of bias due to
confounding?
Favours experimental /
Favours comparator /
Unpredictable
Page 60 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
25
Bias in selection of participants into the study
2.1. Was selection of participants into the study (or into
the analysis) based on participant characteristics
observed after the start of intervention?
If N/PN to 2.1: go to 2.4
Y / PY / PN / N / NI
2.2. If Y/PY to 2.1: Were the post-intervention
variables that influenced selection likely to be
associated with intervention?
2.3 If Y/PY to 2.2: Were the post-intervention
variables that influenced selection likely to be
influenced by the outcome or a cause of the
outcome?
NA / Y / PY / PN / N / NI
NA / Y / PY / PN / N / NI
2.4. Do start of follow-up and start of intervention
coincide for most participants?
Y / PY / PN / N / NI
2.5. If Y/PY to 2.2 and 2.3, or N/PN to 2.4: Were
adjustment techniques used that are likely to correct for
the presence of selection biases?
NA / Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical /
NI
Optional: What is the predicted direction of bias due to
selection of participants into the study?
Favours experimental / Favours
comparator / Towards null /Away
from null / Unpredictable
Bias in classification of interventions
3.1 Were intervention groups clearly defined? Y / PY / PN / N / NI
3.2 Was the information used to define
intervention groups recorded at the start of the
intervention?
Y / PY / PN / N / NI
3.3 Could classification of intervention status have
been affected by knowledge of the outcome or risk
of the outcome?
Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of bias
due to classification of interventions?
Favours experimental / Favours
comparator / Towards null /Away from
null / Unpredictable
Page 61 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
26
Bias due to deviations from intended interventions
If your aim for this study is to assess the effect of assignment to
intervention, answer questions 4.1 and 4.2
4.1. Were there deviations from the intended intervention beyond what would
be expected in usual practice?
Y / PY / PN /
N / NI
4.2. If Y/PY to 4.1: Were these deviations from intended intervention
unbalanced between groups and likely to have affected the outcome?
NA / Y / PY /
PN / N / NI
If your aim for this study is to assess the effect of starting and adhering to
intervention, answer questions 4.3 to 4.6
4.3. Were important co-interventions balanced across intervention groups? Y / PY / PN /
N / NI
4.4. Was the intervention implemented successfully for most participants? Y / PY / PN /
N / NI
4.5. Did study participants adhere to the assigned intervention regimen? Y / PY / PN /
N / NI
4.6. If N/PN to 4.3, 4.4 or 4.5: Was an appropriate analysis used to estimate
the effect of starting and adhering to the intervention?
NA / Y / PY /
PN / N / NI
Risk of bias judgement
Optional: What is the predicted direction of bias due to deviations from the
intended interventions?
Bias due to missing data
5.1 Were outcome data available for all, or nearly all,
participants?
Y / PY / PN / N / NI
5.2 Were participants excluded due to missing data
on intervention status?
Y / PY / PN / N / NI
5.3 Were participants excluded due to missing data
on other variables needed for the analysis?
Y / PY / PN / N / NI
5.4 If PN/N to 5.1, or Y/PY to 5.2 or 5.3: Are the
proportion of participants and reasons for missing
data similar across interventions?
NA / Y / PY / PN / N / NI
5.5 If PN/N to 5.1, or Y/PY to 5.2 or 5.3: Is there
evidence that results were robust to the presence of
missing data?
NA / Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of bias due
to missing data?
Favours experimental / Favours
comparator / Towards null /Away from
null / Unpredictable
Page 62 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
27
Bias in measurement of outcomes
6.1 Could the outcome measure have been
influenced by knowledge of the intervention
received?
Y / PY / PN / N / NI
6.2 Were outcome assessors aware of the
intervention received by study participants?
Y / PY / PN / N / NI
6.3 Were the methods of outcome assessment
comparable across intervention groups?
Y / PY / PN / N / NI
6.4 Were any systematic errors in measurement
of the outcome related to intervention
received?
Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of
bias due to measurement of outcomes?
Favours experimental / Favours comparator /
Towards null /Away from null /
Unpredictable
Bias in selection of the reported result
Is the reported effect estimate likely to be
selected, on the basis of the results, from...
7.1. ... multiple outcome measurements
within the outcome domain?
Y / PY / PN / N / NI
7.2 ... multiple analyses of the intervention-
outcome relationship?
Y / PY / PN / N / NI
7.3 ... different subgroups? Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of
bias due to selection of the reported result?
Favours experimental / Favours comparator /
Towards null /Away from null / Unpredictable
Overall bias
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the overall predicted
direction of bias for this outcome?
Favours experimental / Favours comparator /
Towards null /Away from null / Unpredictable
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License.
Page 63 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
28
The RoB 2.0 tool (individually randomized, parallel group trials)
Assessor name/initials
Study ID and/or reference(s)
Study design
Randomized parallel group trial Cluster-randomized trial Randomized cross-over or other matched design
Specify which outcome is being assessed for risk
of bias
Specify the numerical result being assessed. In
case of multiple alternative analyses being presented,
specify the numeric result (e.g. RR = 1.52 (95% CI
0.83 to 2.77) and/or a reference (e.g. to a table, figure
or paragraph) that uniquely defines the result being
assessed.
Is your aim for this study…? to assess the effect of assignment to intervention to assess the effect of starting and adhering to intervention
Which of the following sources have you obtained to help inform your risk of bias judgements
(tick as many as apply)?
Journal article(s) with results of the trial
Trial protocol
Statistical analysis plan (SAP)
Non-commercial trial registry record (e.g. ClinicalTrials.gov record)
Company-owned trial registry record (e.g. GSK Clinical Study Register record)
“Grey literature” (e.g. unpublished thesis)
Conference abstract(s) about the trial
Regulatory document (e.g. Clinical Study Report, Drug Approval Package)
Research ethics application
Grant database summary (e.g. NIH RePORTER, Research Councils UK Gateway to
Research)
Personal communication with trialist
Personal communication with the sponsor
Page 64 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
29
Page 65 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
30
Domain Signalling questions Response options
Description/Support
for judgement
Bias arising from
the
randomization
process
1.1 Was the allocation
sequence random?
Y / PY / PN / N / NI
1.2 Was the allocation
sequence concealed until
participants were
recruited and assigned to
interventions?
Y / PY / PN / N / NI
1.3 Were there baseline
imbalances that suggest a
problem with the
randomization process?
Y / PY / PN / N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias arising from the
randomization process?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias due to
deviations from
intended
interventions
2.1. Were participants
aware of their assigned
intervention during the
trial?
Y / PY / PN / N / NI
2.2. Were carers and trial
personnel aware of
participants' assigned
intervention during the
trial?
Y / PY / PN / N / NI
2.3. If Y/PY/NI to 2.1 or
2.2: Were important co-
interventions balanced
across intervention
groups?
NA / Y / PY / PN /
N / NI
2.4. Was the intervention
implemented
successfully?
Y / PY / PN / N / NI
2.5. Did study
participants adhere to the
assigned intervention
regimen?
Y / PY / PN / N / NI
Page 66 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
31
Domain Signalling questions Response options
Description/Support
for judgement
2.6. If N/PN/NI to 2.3,
2.4 or 2.5: Was an
appropriate analysis used
to estimate the effect of
starting and adhering to
the intervention?
NA / Y / PY / PN /
N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to deviations
from intended
interventions?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias due to
missing outcome
data
3.1 Were outcome data
available for all, or
nearly all, participants
randomized?
Y / PY / PN / N / NI
3.2 If N/PN/NI to 3.1:
Are the proportions of
missing outcome data
and reasons for missing
outcome data similar
across intervention
groups?
NA / Y / PY / PN /
N / NI
3.3 If N/PN/NI to 3.1: Is
there evidence that
results were robust to the
presence of missing
outcome data?
NA / Y / PY / PN /
N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to missing
outcome data?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias in
measurement of
the outcome
4.1 Were outcome assessors aware of the intervention received by study participants?
Y / PY / PN / N / NI
Page 67 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
32
Domain Signalling questions Response options
Description/Support
for judgement
4.2 If Y/PY/NI to 4.1: Was the assessment of the outcome likely to be influenced by knowledge of intervention received?
NA / Y / PY / PN /
N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to measurement
of the outcome?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias in selection
of the reported
result
Are the reported outcome data likely to have been selected, on the basis of the results, from...
5.1. ... multiple outcome measurements (e.g. scales, definitions, time points) within the outcome domain?
Y / PY / PN / N / NI
5.2 ... multiple analyses of the data?
Y / PY / PN / N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to selection of
the reported result?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Overall bias Risk of bias judgement Low / High / Some
concerns
Optional:
What is the overall
predicted direction of
bias for this outcome?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Page 68 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
33
RESULT SEARCH: PRISMA FLOWCHART
PRISMA Flow Diagram – dietary patterns/indices and foods from
the main food groups and mortality/cancer recurrence amongst
different groups of cancer survivors
Records identified through database searching: Bladder cancer n= 317 Breast cancer n= 2044 Cervical cancer n= 123 Colorectal cancer n= 1157 (Non-)Hodgkin lymphoma n= 245 Kidney cancer n= 134 Larynx cancer n= 94 Malignant melanoma n= 103 Multiple myeloma n= 161 Prostate cancer n= 902 Testicular cancer n= 35 Uterus cancer n= 145
Scre
enin
g In
clu
ded
El
igib
ility
Id
enti
fica
tio
n
Additional records identified through other sources:
(n= 8)
Records after duplicates removed (n= 2883)
Records screened (n= 2883)
Records excluded: on title and abstract (n= 2788)
Full-text articles assessed for eligibility
(n= 95) Full-text articles excluded:
- other outcome than mortality or recurrence (n= 28)
- other exposure than dietary patterns/indices or our selected food items (n= 11)
- Not only cancer survivors (n= 10)
- Excluded because of our own exclusion criteria (sample size, follow-up length) (n= 8)
Studies included in qualitative synthesis
(n= 38)
Page 69 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Online supporting data File S2 for manuscript ‘the impact of dietary patterns and the main food groups on mortality and
recurrence in cancer survivors: systematic review of current epidemiological literature’
Jochems et al., 31-05-2017
Page 70 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Bladder cancer
Table S1: Summary of studies bladder cancer
Author (year)
Study / country Number of participants / sex (age range)
Follow-up period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI)
multivariate adjusted
Adjustment
Tang et al. (2010)
Roswell Park Cancer Institute (RPCI) / United States
239 m/w (not specified) 8.0 Fruit and vegetables
Pre-diagnosis FFQ usual diet in the few years before diagnosis
Overall mortality, cancer-specific mortality
Total fruit: HR1= 0.91; 95% CI 0.62-1.33 HR2= 1.09; 95% CI 0.66-1.81 Total vegetables: HR1= 0.91; 95% CI 0.62-1.36 HR2= 1.06; 95% CI 0.63-1.78 Cruciferous vegetables: HR1= 0.87; 95% CI 0.60-1.26 HR2= 0.89; 95% CI 0.53-1.48 Raw cruciferous vegetables: HR1= 0.73; 95% CI 0.50-1.06 HR2= 0.73; 95% CI 0.44-1.21
age at diagnosis, total meat intake, pack-years of smoking, tumour stage, radiation therapy
Page 71 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Bowel cancer
Table S2: Summary of studies bowel cancer
Author (year) Study / country Number of participants / sex (age at baseline)
Follow-up
period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Meyerhardt et al. (2007)
Cancer and Leukemia Group B (CALGB 89803) adjuvant therapy trial for stage III colon cancer / USA
1,009 m/w (21-85) 5.3 PCA: prudent diet, Western diet
Post-diagnosis
FFQ during and 6 months after adjuvant chemotherapy
Overall mortality, cancer recurrence
Prudent diet: HR1= 1.32; 95% CI 0.86-2.04 HR4= 1.13; 95% CI 0.77-1.67 Western diet: HR1= 2.32; 95% CI 1.36-3.96 HR4= 2.85; 95% CI 1.75-4.63
sex, age, depth of invasion through bowel wall, number of positive lymph nodes, presence of clinical perforation at time of surgery, presence of bowel obstruction at time of surgery, baseline performance status, treatment group, weight change between first and second questionnaire, time-varying BMI, time-varying physical activity level, time-varying total calories
McCullough et al. (2013)
Cancer Prevention Study II (CPSII) Nutrition Cohort / USA
2,315 m/w (40-93) 7.5 Red and processed meat, unprocessed red meat, processed meat
Pre- and post-diagnosis
FFQ usual diet of the year before diagnosis and two times during follow-up
Overall mortality, cancer-specific mortality, death from other causes
Pre-diagnosis: Red and processed meat RR1= 1.29; 95% CI 1.05-1.59 RR2= 1.09; 95% CI 0.79-1.51 RR3= 1.39; 95% CI 1.00-1.92 Unprocessed red meat RR1= 1.12; 95% CI 0.92-1.38 RR2= 1.16; 95% CI 0.84-1.58 RR3= 1.19; 95% CI 0.87-1.64 Processed meat RR1= 1.21; 95% CI 0.99-1.48 RR2= 1.09; 95% CI 0.80-1.48 RR3= 1.27; 95% CI 0.92-1.75 Post-diagnosis: Red and processed meat RR1= 0.94; 95% CI 0.68-1.30 RR2= 1.10; 95% CI 0.61-1.98 RR3= 0.87; 95% CI 0.54-1.41 Unprocessed red meat: RR1= 0.75; 95% CI 0.55-1.03 RR2= 1.13; 95% CI 0.62-2.06 RR3= 0.64; 95% CI 0.40-1.03 Processed meat: RR1= 1.11; 95% CI 0.82-1.49 RR2= 1.06; 95% CI 0.61-1.84 RR3= 1.14; 95% CI 0.73-1.77
pre-diagnosis model: age at diagnosis, sex, tumour stage at diagnosis, 1992 pre-diagnostic energy intake, BMI in 1992, history of diabetes, and history of myocardial infarction. Post-diagnosis model: age at diagnosis, sex, tumour stage at diagnosis, and post-diagnostic energy intake, weight change between 1992 pre-diagnostic, post-diagnostic questionnaires, and 1992 pre-diagnostic meat intake
Page 72 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Zhu et al. (2013)
Familial CRC registry in Newfoundland (FBCR) / Canada
529 m/w (20-75) 6.4 PCA: prudent vegetable pattern, processed meat pattern, high sugar pattern
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
Prudent vegetable pattern: HR1= 1.03; 95% CI 0.61-1.75 HR2= 1.12; 95% CI 0.69-1.84 Processed meat pattern: HR1= 1.53; 95% CI 0.85-2.74 HR2= 1.82; 95% CI 1.07-3.09 High sugar pattern: HR1= 1.27; 95% CI 0.72-2.25 HR2= 1.02; 95% CI 0.62-1.69
total energy intake, sex, age at diagnosis, stage at diagnosis, marital status, family history, reported screening procedure, reported chemo-radiotherapy and microsatellite instability status
Dik et al. (2014)
The European Prospective Investigation into Cancer and Nutrition (EPIC) / pooled analysis data Denmark, Italy, Netherlands, Norway, Spain, Sweden, United Kingdom, France, Germany, Greece
3,859 m/w (25-70) 4.1 Total dairy, milk, yoghurt, cheese
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
Total dairy: HR1= 1.16; 95% CI 0.98-1.36 HR2= 1.17; 95% CI 0.96-1.43 Milk: HR1= 1.21; 95% CI 1.03-1.43 HR2= 1.21; 95% CI 0.99-1.48 Yoghurt: HR1= 1.08; 95% CI 0.92-1.28 HR2= 1.09; 95% CI 0.88-1.34 Cheese: HR1= 0.87; 95% CI 0.74-1.04 HR2= 0.93; 95% CI 0.76-1.14
age at diagnosis, sex, pre-diagnosis BMI, smoking status, energy intake, tumour subsite (colon and rectum), disease stage, differentiation grade; stratified by centre
Fung et al. (2014)
Nurses' Health Study (NHS) / USA
1,201 w (30-55) 11.2 Dietary indices: AHEI, DASH, AMED PCA: prudent diet, Western diet
Post-diagnosis
FFQ at least 6 months after diagnosis
Overall mortality, cancer-specific mortality
AHEI: HR1= 0.71; 95% CI 0.52-0.98 HR2= 0.72; 95% CI 0.43-1.21 DASH: HR1= 0.98; 95% CI 0.71-1.35 HR2= 0.87; 95% CI 0.52-1.45 AMED: HR1= 0.87; 95% CI 0.63-1.21 HR2= 0.84; 95% CI 0.50-1.42 Prudent diet: HR1= 0.93; 95% CI 0.65-1.34 HR2= 0.67; 95% CI 0.37-1.22 Western diet: HR1= 1.32; 95% CI 0.89-1.97 HR2= 1.66; 95% CI 0.85-3.23
age, physical activity, BMI, weight change, cancer grade, chemotherapy, smoking status, energy intake, colon or rectal cancer, stage of disease, and date of colorectal cancer diagnosis
Pelser et al. (2014)
NIH-AARP Diet and Health Study / USA
5,727 m/w (50-71) 5 Dietary indices: HEI
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
HEI: HR1= 0.95; 95% CI 0.78-1.16 HR2= 0.99; 95% CI 0.77-1.27
age, sex, lag time, education, family history colon cancer, cancer stage, primary treatment, BMI, physical activity, alcohol consumption, smoking
Skeie et al. (2014)
HELGA cohort including the Norwegian Women and Cancer Study, the Northern Sweden Health and Disease Study, and the Danish Diet Cancer and Health Study / Denmark, Norway, Sweden
1,119 m/w (30-64) 7 Total whole grains, whole grain wheat, whole grain rye, whole grain oats, whole grain products
Pre-diagnosis
FFQ usual diet before diagnosis
Overall mortality For men Total whole grains: HR1= 1.00; 95% CI 0.67-1.48 Whole grain wheat: HR1= 0.97; 95% CI 0.64-1.49 Whole grain rye: HR1= 0.90; 95% CI 0.60-1.36 Whole grain oats:
age at diagnosis, metastasis, smoking, folate, margarine, energy intake, stratified for country and cancer location. Wheat, rye and oats were also adjusted for the other grains
Page 73 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
HR1= 1.11; 95% CI 0.72-1.70 Whole grain products: HR1= 1.06; 95% CI 0.71-1.56 For women Total whole grains: HR1= 0.91; 95% CI 0.60-1.39 Whole grain wheat: HR1= 1.35; 95% CI 0.72-2.53 Whole grain rye: HR1= 0.93; 95% CI 0.60-1.46 Whole grain oats: HR1= 0.83; 95% CI 0.55-1.26 Whole grain products: HR1= 1.10; 95% CI 0.74-1.64
Yang et al. (2014)
Cancer Prevention Study II (CPSII) Nutrition Cohort / USA
2,284 m/w (40-92) 7.5 Total dairy, milk Pre- and post-diagnosis
FFQ usual diet of the year before diagnosis and two times during follow-up
Overall mortality, cancer-specific mortality
Pre-diagnosis: Total dairy RR1= 0.88; 95% CI 0.72-1.09 RR2= 0.89; 95% CI 0.65-1.22 Milk RR1= 0.95; 95% CI 0.79-1.15 RR2= 0.98; 95% CI 0.73-1.32 Post-diagnosis: Total dairy RR1= 0.75; 95% CI 0.56-1.01 RR2= 0.73; 95% CI 0.44-1.23 Milk RR1= 0.72; 95% CI 0.55-0.94 RR2= 0.93; 95% CI 0.59-1.49
pre-diagnosis: age at diagnosis, sex, tumour stage, pre-diagnosis total energy and total folate intakes. post-diagnosis: age at diagnosis, sex, tumour stage, post-diagnosis total energy and total folate intakes
Carr et al. (2016)
Darmkrebs: chancen der Verhutung durch Screening study (DACHS) / Germany
3,122 m/w (>30) 4.8 Red and processed meat
Pre-diagnosis
FFQ usual diet before diagnosis
Overall mortality, cancer-specific mortality, cancer recurrence
Red and processed meat: HR1= 0.85; 95% CI 0.67-1.09 HR2= 0.83; 95% CI 0.61-1.14 HR4= 1.03; 95% CI 0.80-1.33
age at diagnosis, sex, cancer stage, chemotherapy, surgery, BMI, physical activity, diabetes, stroke, heart failure, myocardial infarction, dairy intake, wholegrain intake, time between diagnosis and interview, and a time-dependent effect of chemotherapy
Romaguera et al. (2016)
The European Prospective Investigation into Cancer and Nutrition (EPIC) / pooled analysis data Denmark, Italy, Netherlands, Norway, Spain, Sweden, United Kingdom, France, Germany, Greece
3,292 m/w (25-70) 4.2 Dietary indices: WCRF/AICR score
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
WCRF/AICR score: HR1= 0.79; 95% CI 0.65-0.98 HR2= 0.70; 95% CI 0.56–0.89
age at diagnosis as entry time and age at death or censoring as exit time, year of diagnosis, tumour stage, tumour grade, tumour site, sex, educational level, and smoking status; stratified by country
Ward et al. (2016)
The European Prospective Investigation into Cancer and Nutrition (EPIC) / pooled analysis data Denmark, Italy,
3,789 m/w (25-70) 4.1 Red meat, processed meat, poultry
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
Red and processed meat: HR1= 1.00; 95% CI 0.83-1.20 HR2= 1.00; 95% CI 0.81-1.23 Unprocessed red meat:
adjusted for age at diagnosis, sex, BMI, smoking status, tumour grade, tumour stage, year of tumour diagnosis,
Page 74 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Netherlands, Norway, Spain, Sweden, United Kingdom, France, Germany, Greece
HR1= 0.95; 95% CI 0.78-1.14 HR2= 0.93; 95% CI 0.75-1.15 Processed meat: HR1= 1.17; 95% CI 0.97-1.42 HR2= 1.12; 95% CI 0.90-1.39 Poultry: HR1= 0.87; 95% CI 0.73-1.03 HR2= 0.91; 95% CI 0.75-1.10
energy intake, calcium intake, folate intake, alcohol intake, education; stratified by country
Ratjen et al. (2017)
Patients with histologically confirmed colorectal cancer recruited by the PopGen biobank / Germany
1,404 m/w (56-67) 7 Dietary indices: Modified Mediterranean Diet Score (MMDS), Healthy Nordic Food Index (HNFI)
Post-diagnosis
FFQ usual diet assessed 6 years (median) after diagnosis
Overall mortality MMDS: HR1= 0.48; 95% CI 0.32-0.74 HNFI: HR1= 0.63; 95% CI 0.39-1.04
sex, age at diet assessment, BMI, physical activity, survival time from CRC diagnosis until diet assessment, tumour location, occurrence of metastases, occurrence of other cancer, chemotherapy, smoking status, total energy intake, time 3 age, time 3 BMI, and time 3 metastases
Page 75 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Breast cancer
Table S3: Summary of studies breast cancer
Author (year)
Study / country Number of participants /
sex (age)
Follow-up
period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI)
multivariate adjusted
Adjustment
Hebert et al. (1998)
Memorial Sloan- Kettering Cancer Center Follow-up Study / USA
472 w (20-80) 8-10 Red meat, butter/margarine/lard
Post-diagnosis
FFQ usual diet before diagnosis
Cancer-specific mortality, cancer recurrence
Red Meat: RR2= 2.60; 95% CI 0.96-7.03 RR4= 1.12; 95% CI 0.66-1.89 Butter/margarine/lard: RR2= 1.03; 95% CI 0.61-1.76 RR4= 1.30; 95% CI 1.03-1.64
disease stage, oestrogen receptor status, age, BMI, menopausal status, energy intake
Holmes et al. (1999)
Nurses' Health Study (NHS) / USA
1,504 w pre-diagnosis (mean age 54) and 1,982 w post-diagnosis)
13.1 Vegetables, poultry, fish, dairy, red meat (processed and unprocessed combined)
Pre- and post-diagnosis
FFQ usual diet after diagnosis
Overall mortality, cancer-specific mortality (results for breast cancer-specific mortality are not shown in paper)
Pre-diagnosis: Vegetables RR1= 0.98; 95% CI 0.62-1.53 Poultry RR1= 0.60; 95% CI 0.39-0.92 Fish RR1= 0.94; 95% CI 0.62-1.43 Dairy RR1= 0.71; 95% CI 0.44-1.14 Red meat (not shown) Post-diagnosis: Vegetables RR1= 0.81; 95% CI 0.59–1.11 Poultry RR1= 0.70; 95% CI 0.50–0.97 Fish RR1= 0.80; 95% CI 0.60–1.07 Dairy RR1= 0.72; 95% CI 0.52–1.00 Red meat RR1= 1.06; 95% CI 0.76–1.49
pre-diagnosis: quantiles of nutrient or food intake prior to diagnosis, previous diet interval, age, diet interval, calendar year of diagnosis, body mass index, oral contraceptive use, menopausal status, postmenopausal hormone use, smoking, age at first birth and parity, number of metastatic lymph nodes, tumour size, and caloric intake post-diagnosis: age, diet interval, calendar year of diagnosis, body mass index, oral contraceptive use, menopausal status, postmenopausal hormone use, smoking, age at first birth and parity, number of metastatic lymph nodes, tumour size, caloric intake
Kroenke et al. (2005)
Nurses' Health Study (NHS) / USA
2,619 w (30-55) 9 PCA: prudent diet, Western diet
Pre- and post-diagnosis
FFQ usual intake 4 years before diagnosis and FFQ at least one year after diagnosis
Overall mortality, cancer-specific mortality, death from other causes
Pre-diagnosis: Prudent diet RR1= non-significant (not shown) RR2= non-significant (not shown) RR3= non-significant (not shown) Western diet RR1= 1.40; 95% CI 0.93-2.09 RR2= 1.01; 95% CI 0.59-1.72 RR3= 1.95; 95% CI 1.06-3.60
age, time since diagnosis, BMI, energy intake, smoking, physical activity, diet missing, age at menarche, oral contraceptive use, menopausal status and use of postmenopausal hormone therapy, age at menopause, tamoxifen, chemotherapy, tumour stage at diagnosis, time between dietary assessment and diagnosis (for pre-diagnosis diet)
Page 76 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Post-diagnosis: Prudent diet RR1= 0.78; 95% CI 0.54-1.12 RR2= 1.07; 95% CI 0.66-1.73 RR3= 0.54; 95% CI 0.31-0.95 Western diet RR1= 1.53; 95% CI 1.03-2.29 RR2= 1.01; 95% CI 0.60-1.70 RR3= 2.09; 95% CI 1.30-3.36
McEligot et al. (2006)
Cancer Surveillance Program of Orange County (CSPOC) / USA
516 w (age >50) 6.7 Fruit, vegetables Pre-diagnosis
FFQ usual diet one year before diagnosis
Overall mortality Total fruit: HR1= 0.63; 95% CI 0.38-1.05 Total vegetables: HR1= 0.57; 95% CI 0.35–0.94
tumour stage, age at diagnosis, BMI, parity, HRT, alcohol intake, multivitamins, energy intake
Chlebowski et al. (2006)
Women’s Intervention Nutrition Study (WINS) RCT / USA
2,437 w 5 Dietary indices: Low fat diet
Post-diagnosis
FFQ with interview on dietary intake after diagnosis
Overall mortality Intervention versus control Low-fat diet: HR1= 0.89; 95% CI 0.65-1.21
nodal status, systemic adjuvant therapy, ER status, tumour size, mastectomy
Pierce et al. (2007a)
Women's Healthy Eating and Living (WHEL) RCT / USA
3,088 w (18-70) 7.3 Dietary indices: Low fat diet
Post-diagnosis
FFQ with interview on dietary intake after diagnosis
Overall mortality Intervention versus control Low-fat diet: HR1= 0.91; 95% CI 0.72-1.15
anti-oestrogen use, bilateral oophorectomy, age, BMI, physical activity, energy intake, tumour characteristics (including hormone receptor status), years from diagnosis to study entry
Dal Maso et al. (2008)
Six Italian Regions Follow-up Study / Italy
1,453 w (23-74) 12.6 Fruit and vegetables Pre-diagnosis
FFQ usual diet year before diagnosis
Overall mortality, cancer-specific mortality
Fruit and vegetables: HR1= 1.27; 95% CI 1.00–1.61 HR2= 1.26; 95% CI 0.96–1.64 (low versus high intake!)
region, age at diagnosis, year of diagnosis, TNM stage, receptor status
Kwan et al. (2009)
Life After Cancer Epidemiology (LACE) study / USA
1,901 w (18-79) 4.2 PCA: prudent diet, Western diet
Post-diagnosis
FFQ usual diet 3 years after diagnosis
Overall mortality, cancer-specific mortality, death from other causes, cancer recurrence
Prudent diet: HR1= 0.57; 95% CI 0.36-0.90 HR2= 0.79; 95% CI 0.43-1.43 HR3= 0.35; 95% CI 0.17-0.73 HR4= 0.95; 95% CI 0.63-1.43 Western diet: HR1= 1.76; 95% CI 1.10-2.81 HR2= 1.20; 95% CI 0.62-2.32 HR3= 2.15; 95% CI 0.97-4.77 HR4= 0.98; 95% CI 0.62-1.54
age at diagnosis, total energy intake, ethnicity, BMI, weight change before diagnosis to baseline, smoking status, menopausal status at diagnosis, stage, hormone receptor status, treatment
Beasley et al. (2011)
Collaborative Women’s Longevity Study (CWLS) / USA
4,441 w (20-79) 5.5 Fruit, vegetables, dairy, meat (poultry, fish, beef, and processed)
Post-diagnosis
FFQ usual diet after diagnosis (1-16 years)
Overall mortality, cancer-specific mortality
Total fruit: HR1= 1.38; 95% CI 0.88-2.17 HR2= 1.39; 95% CI 0.64-2.99 Total vegetables: HR1= 1.44; 95% CI 0.91-2.27 HR2= 0.96; 95% CI 0.38-2.45 Cruciferous vegetables: HR1= 1.02; 95% CI 0.80-1.30 HR2= 0.95; 95% CI 0.59-1.54 Dairy: HR1= 1.18; 95% CI 0.90-1.54 HR2= 0.94; 95% CI 0.56-1.59 Meat (poultry, fish, beef, and processed):
age, state of residence, menopausal status, smoking, breast cancer stage, alcohol, history of hormone replacement therapy at diagnosis, interval between diagnosis and diet assessment, and energy intake, breast cancer treatment, body mass at follow-up
Page 77 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
HR1= 1.12; 95% CI 0.83-1.51 HR2= 0.89; 95% CI 0.50-1.60
Buck et al. (2011)
MARIE study / Germany
2,653 w (50-74) 6.4 Fruit, vegetables, bread, sunflower/pumpkin seeds, sesame/flaxseeds
Pre-diagnosis
FFQ usual diet year before diagnosis
Overall mortality, cancer-specific mortality
Fruit: HR1= 0.84; 95% CI 0.61-1.16 HR2= 0.86; 95% CI 0.59-1.25 Vegetables: HR1= 1.09; 95% CI 0.80-1.48 HR2= 1.01; 95% CI 0.70-1.46 Bread: HR1= 1.31; 95% CI 0.93-1.83 HR2= 1.10; 95% CI 0.74-1.63 Sunflower/pumpkinseeds: HR1= 0.87; 95% CI 0.66-1.15 HR2= 1.12; 95% CI 0.79-1.57 Sesame/flaxseeds: HR1= 0.90; 95% CI 0.68-1.19 HR2= 1.21; 95% CI 0.87-1.68
tumour size, nodal status, metastasis, grade, oestrogen and progesterone receptor status, breast cancer detection type, diabetes, HRT use at diagnosis, study centre, energy intake, age at diagnosis
George et al. (2011)
Health, Eating, Activity, and Lifestyle (HEAL) / USA
670 w (older than 18)
6 Dietary indices: HEI
Post-diagnosis
FFQ usual diet approx. 2,5 yrs after diagnosis
Overall mortality, cancer-specific mortality
HEI: HR1= 0.40; 95% CI 0.17-0.94 HR2= 0.12; 95% CI 0.02-0.99
energy intake, physical activity, race, tumour stage, tamoxifen use, BMI
Kim et al. (2011)
Nurses' Health Study (NHS) / USA
2,729 w (30-55) not stated
Dietary indices: AHEI, DQIR, RFS, AMED
Post-diagnosis
FFQ usual diet around 1 year after diagnosis
Overall mortality, cancer specific mortality, death from other causes
AHEI: RR1= 0.85; 95% CI 0.63-1.17 RR2= 1.53; 95% CI 0.98-2.39 RR3= 0.52; 95% CI 0.32-0.83 DQIR: RR1= 0.78; 95% CI 0.58-1.07 RR2= 0.81; 95% CI 0.53-1.24 RR3= 0.85; 95% CI 0.54-1.34 RFS: RR1= 1.03; 95% CI 0.74-1.42 RR2= 1.54; 95% CI 0.95-2.47 RR3= 0.86; 95% CI 0.54-1.37 AMED: RR1= 0.87; 95% CI 0.64-1.17 RR2= 1.15; 95% CI 0.74-1.77 RR3= 0.80; 95% CI 0.50-1.26
age, time since diagnosis, alcohol intake (only for RFS because alcohol is a component in the other 3 diet quality indices), energy, multivitamin use (except for AHEI because it is a component), BMI, weight change (BMI at time of diet minus BMI just prior to diagnosis), oral contraceptive use, smoking status, physical activity in METs, stage, categories of treatment, age at first birth and parity, menopausal status and postmenopausal hormone use
Izano et al. (2013)
Nurses' Health Study (NHS) / USA
4,103 w (30-55) 9.3 Dietary indices: AHEI, DASH
Post-diagnosis
FFQ usual diet around 1 year after diagnosis
Cancer specific mortality, death from other causes
AHEI: RR2= 1.07; 95% CI 0.77-1.49 RR3= 0.57; 95% CI 0.42-0.77 DASH: RR2= 0.85; 95% CI 0.61-1.19 RR3= 0.72; 95% CI 0.53-0.99
stratified by time since diagnosis, adjusted for age at diagnosis, quintiles of energy intake, BMI and BMI change, age at first birth and parity, oral contraceptive use, menopausal status and HRT use, smoking, stage of disease, radiation treatment, chemotherapy and hormonal treatment, and physical activity
Kroenke et al. (2013)
Life After Cancer Epidemiology (LACE) study / USA
1,893 w (18-70) 11.8 Dairy Post-diagnosis
FFQ diet at diagnosis when cancer recurs before 6 yrs - otherwise 6 years after diagnosis
Overall mortality, cancer recurrence
Total Dairy: HR1= 1.39; 95% CI 1.02-1.90 HR4= 1.13; 95% CI 0.83-1.54 Low-fat dairy: HR1= 1.05; 95% CI 0.80-1.36 HR4= 1.01; 95% CI 0.78-1.32 High-fat dairy:
age, time between diagnosis and dietary assessment, high- and low-fat dairy intake, race, education, cancer stage at diagnosis, tumour size, human epidermal growth receptor 2, nodal and oestrogen receptor status, chemotherapy, radiation, tamoxifen, comorbidity, menopausal status, BMI, physical
Page 78 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
HR1= 1.64; 95% CI 1.24-2.17 HR4= 1.22; 95% CI 0.92-1.55
activity, energy intake, alcohol intake, red meat intake, fibre intake, fruit intake
Nechuta et al. (2013)
After Breast Cancer Pooling Project (includes cohorts SBCSS, LACE, WHEL, NHS) / USA and China
11,390 w (20-83) 9.0 Cruciferous vegetables Post-diagnosis
FFQ approx. 2 yrs after diagnosis
Overall mortality, cancer recurrence
Cruciferous vegetables: HR1= 0.99; 95% CI 0.86-1.13 HR4= 1.10; 95% CI 0.95-1.28
age at diagnosis, ER/PR status, TNM stage, surgery, chemotherapy, radiotherapy, hormonal therapy, smoking, BMI, exercise, menopausal status, race/ethnicity, education
Vrieling et al. (2013)
Mammary carcinoma Risk factor Investigation (MARIE) study / Germany
2,522 w 5.5 PCA: healthy pattern, unhealthy pattern
Pre-diagnosis
FFQ usual diet the year before diagnosis
Overall mortality, cancer specific mortality, death from other causes, cancer recurrence
Healthy pattern: HR1= 0.87; 95% CI 0.61-1.23 HR2= 0.89; 95% CI 0.59-1.35 HR3= 0.81; 95% CI 0.40-1.61 HR4= 0.71; 95% CI 0.48-1.06 Unhealthy pattern: HR1= 1.34; 95% CI 0.93-1.94 HR2= 0.99; 95% CI 0.64-1.52 HR3= 3.69; 95% CI 1.66-8.17 HR4= 0.91; 95% CI 0.61-1.36
tumour size, nodal status, metastases, tumour grade, ERPR status, radiotherapy, HRT use at diagnosis, mode of detection, and total energy intake and stratified by age at diagnosis and study centre
George et al. (2014)
Women’s Health Initiative’s Dietary Modification Trial and Observational Study (WHI) / USA
2,317 w (50-79) 9.6 Dietary indices: HEI
Post-diagnosis
FFQ usual diet approx. 1.5 yrs after diagnosis
Overall mortality, cancer specific mortality, death from other causes
HEI: HR1= 0.74; 95% CI 0.55-0.99 HR2= 0.91; 95% CI 0.60-1.40 HR3= 0.58; 95% CI 0.38-0.87
age at screening visit, WHI component, ethnicity, income, education, stage, estrogen receptor status, progesterone receptor status, time since diagnosis, energy intake in kcals, physical activity in MET, servings of alcohol per week, use of postmenopausal hormone therapy
McCullough et al. (2016)
Cancer Prevention Study II (CPS-II) Nutrition Cohort / USA
4,452 w for pre-diagnosis and 2,152 w for post-diagnosis (mean age 70.7 yrs)
9.8-9.9 Dietary indices: ACS
Pre- and post-diagnosis
FFQ usual diet in 1992 (before diagnosis) and usual diet at least 1 year after diagnosis (after diagnosis)
Overall mortality, cancer-specific mortality, death from other causes
ACS: Pre-diagnosis RR1= 1.00; 95% CI 0.84-1.18 RR2= 1.06; 95% CI 0.79-1.42 RR3= 1.02; 95% CI 0.79-1.31 Post-diagnosis RR1= 0.93; 95% CI 0.73-1.18 RR2= 1.44; 95% CI 0.90-2.30 RR3= 0.78; 95% CI 0.56-1.07
age at diagnosis, diagnosis year, tumour stage at diagnosis, tumour grade at diagnosis, estrogen receptor status, progesterone receptor status, initial treatment (surgery, chemotherapy, radiation, hormone therapy, aromatase inhibitor use and/or Herceptin use), and the following assessed at the time of FFQ completion: BMI, smoking status, physical activity and energy intake
Page 79 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Laryngeal cancer
Table S4: Summary of studies laryngeal cancer
Author (year) Study / country Number of participants / sex
(age)
Follow-up period (yrs)
Exposure Exposure timeframe Exposure assessment Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Crosignani et al. (1996)
Lombardy Cancer Registry (LCR) / Italy
213 m (32-75) 8-10 yr Meat (beef, veal), poultry, fish, eggs, milk, cheese, bread, pasta, potatoes, vegetables, citrus fruits, other fruits, butter, olive oil
Pre-diagnosis Interview usual diet in year before diagnosis
Overall mortality
Citrus fruits: HR1= 0.76; 95% 0.49-1.19 Other fruits: HR1= 0.65; 95% CI 0.39-1.07 Vegetables: HR1= 0.57; 95% CI 0.35-0.94 Meat: HR1= 0.50; 95% CI 0.30-0.83 Poultry: HR1= 0.90; 95% CI 0.55-1.46 Fish: HR1= 0.91; 95% CI 0.59-1.39 Eggs: HR1= 1.22; 95% CI 0.74-2.00 Milk: HR1= 1.58; 95% CI 0.99-2.55 Cheese: HR1= 0.70; 95% CI 0.44-1.12 Bread: HR1= 0.54; 95% CI 0.32-0.90 Pasta: HR1= 1.25; 95% CI 0.76-2.04 Potatoes: HR1= 1.02; 95% CI 0.64-1.64 Butter: HR1= 1.11; 95% CI 0.69-1.80 Olive oil: HR1= 0.71; 95% CI 0.44-1.16
age at diagnosis, clinical stage, occurrence of new primary cancers
Page 80 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Non-Hodgkin lymphoma
Table S5: Summary of studies non-Hodgkin lymphoma
Author (year)
Study / country Number of participants /
sex (age)
Follow-up period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Han et al. (2010)
Yale Connecticut Tumor Registry New York (CTR) / USA
568 w (21-84) 7.7 Fruit, vegetables Pre-diagnosis
FFQ usual diet the year before diagnosis
Overall mortality, cancer-specific mortality
Total fruit and vegetables: HR1= 0.68; 95% CI 0.49-0.95 HR2= 0.70; 95% CI 0.45-1.10 Total fruit: HR1= 0.91; 95% CI 0.70-1.18 HR2= 1.04; 95% CI 0.74-1.45 Total vegetables: HR1= 0.58; 95% CI 0.38-0.89 HR2= 0.58; 95% CI 0.33-1.03 Cruciferous vegetables: HR1= 0.91; 95% CI 0.67–1.24 HR2= 0.75; 95% CI 0.49–1.14 Bean vegetables: HR1= 1.14; 95% CI 0.85-1.54 HR2= 1.05; 95% CI 0.71-1.55 Green leafy vegetables: HR1= 0.71; 95% CI 0.51-0.98 HR2= 0.82; 95% CI 0.54-1.23 Red vegetables: HR1= 1.03; 95% CI 0.76-1.38 HR2= 1.11; 95% CI 0.76-1.62 Yellow vegetables: HR1= 0.93; 95% CI 0.69-1.25 HR2= 1.11; 95% CI 0.77-1.61 Citrus fruits: HR1= 0.73; 95% CI 0.54-0.99 HR2= 0.81; 95% CI 0.54-1.20
age, education, stage, B-symptom, initial treatment, total energy intake
Leo et al. (2015)
Multi-ethnic Cohort (MEC) / USA
2,339 m/w (45-75)
4.5 Fruit, vegetables, dairy, soy, legumes, fish, red meat, processed meat
Pre-diagnosis
FFQ usual diet the year before diagnosis
Overall mortality, cancer specific mortality
Vegetables: HR1= 0.98; 95% CI 0.85-1.12 HR2= 0.98; 95% CI 0.83-1.16 Fruits: HR1= 1.03; 95% CI 0.90-1.19 HR2= 1.04; 95% CI 0.88-1.24 Red meat: HR1= 1.00; 95% CI 0.87-1.15 HR2= 0.95; 95% CI 0.81-1.13 Processed meat: HR1= 0.94; 95% CI 0.82-1.08 HR2= 0.94; 95% CI 0.79-1.12 Fish: HR1= 0.90; 95% CI 0.78-1.03 HR2= 0.91; 95% CI 0.76-1.08 Legumes:
age at NHL diagnosis, BMI, sex, ethnicity, SEER summary stage, NHL subtype, chemo-, radio-, immuno-, and steroid-therapy, smoking status at baseline, alcohol use, education status, energy intake, number of comorbidities
Page 81 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
HR1= 0.88; 95% CI 0.76-1.01 HR2= 0.86; 95% CI 0.72-1.02 Soy foods: HR1= 0.93; 95% CI 0.76-1.14 HR2= 0.92; 95% CI 0.72-1.18 Dairy products: HR1= 1.14; 95% CI 1.00-1.31 HR2= 1.16; 95% CI 0.98-1.37
Page 82 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Prostate cancer
Table S6: Summary of studies prostate cancer
Author (year)
Study / country Number of participants /
sex (age)
Follow-up period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Chavarro et al. (2008)
Physician’s Health Study (PHS) / USA
2,161 m 19 Total fish Pre-diagnosis
FFQ usual diet before diagnosis
Cancer-specific mortality
Total fish: HR2= 0.52; 95% CI 0.30-0.91
age at prostate cancer diagnosis, BMI, physical activity, alcohol use, tomato and dairy products, smoking, ethnicity, multivitamin and vitamin E supplements, random assignment to aspirin or beta-carotene, tumour stage, grade at diagnosis, clinical presentation of case
Kenfield et al. (2014)
Health Professionals Follow-up Study (HPFS) / USA
4,538 m (40-75) 23.2 Dietary indices: Mediterranean diet score (MDS)
Post-diagnosis
FFQ usual diet after diagnosis
Overall mortality, cancer-specific mortality
MDS: HR1= 0.78; 95% CI 0.67-0.90 HR2= 1.01; 95% CI 0.75-1.38
age at diagnosis, time period, time diagnosis to FFQ, energy, BMI, vigorous physical activity, smoking status, clinical stage, Gleason score, treatment
Yang et al. (2015a)
Physician’s Health Study (PHS) / USA
926 m (40-84) 9.6 Total dairy, high-fat dairy, low-fat dairy
Post-diagnosis
FFQ usual diet after diagnosis
Overall mortality Total dairy: HR1 = 1.76; 95% CI 1.21-2.55 HR2 = 2.41; 95% CI 0.96-6.02 High-fat dairy: HR1= 1.22; 95% CI 1.08-1.38 HR2= 1.30; 95% CI 0.97-1.73 Low-fat dairy: HR1= 1.17; 95% CI 1.05-1.29 HR2= 1.16; 95% CI 0.88-1.53
age at diagnosis, total energy intake, BMI, smoking status, exercise, Gleason score, clinical stage, prostate-specific antigen level, time interval between diagnosis and FFQ completion, initial treatment after diagnosis, family history of prostate cancer, and indicators for prudent dietary pattern and Western dietary pattern after excluding dairy products
Yang et al. (2015b)
Physician’s Health Study (PHS) / USA
926 m (40-84) 8.7 PCA: prudent diet, Western diet
Post-diagnosis
FFQ usual after diagnosis
Overall mortality, cancer specific mortality
Prudent diet: RR1= 0.64; 95% CI 0.44-0.93 RR2= 0.46; 95% CI 0.17-1.24 Western diet: RR1= 1.67; 95% CI 1.16-2.42 RR2= 2.53; 95% CI 1.00-6.42
age at diagnosis, total energy intake, BMI, smoking status, vigorous physical activity, Gleason score, clinical stage, prostate-specific antigen level, time interval between diagnosis and FFQ completion, initial treatment, family history of prostate cancer
HR1/RR1= overall mortality
HR2/RR2= cancer-specific mortality
HR3/RR3= death from other causes
HR4/RR4= cancer recurrence
BMI= body mass index
FFQ= food frequency questionnaire
Page 83 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Online supporting data File S3 for manuscript ‘the impact of dietary patterns
and the main food groups on mortality and recurrence in cancer survivors:
systematic review of current epidemiological literature’
Jochems et al., 31-05-2017
Page 84 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
The Grading of Recommendations Assessment, Development and Evaluation (GRADE)
GRADE is a systematic and explicit approach to making judgements about quality of
evidence and strength of recommendations. The focus is on clinical outcomes that patients
themselves are aware of in relation to their condition – in this systematic review these
include overall mortality, cancer-specific mortality, death from other causes, and cancer
recurrence. With the use of GRADE, the evidence is not rated study by study but across
studies for each outcome. Individual study quality was assessed with the Cochrane
Collaboration risk of bias assessment tools; the RoB 2.0 tool for randomised trials and
the ROBINS-I toll for cohort studies. Furthermore, studies were excluded from the
systematic review if the sample size for the analysis was <200 (comparisons containing
less than 200 participants in total are described as sparse data), the follow-up period was
<4 years (for most cancer types, the risk of cancer recurrence is the greatest within the
first three years), no adjustments in the statistical analysis were made for both age and
disease stage and, where possible, for cancer treatment (e.g. studies adjusting for age and
energy intake only were excluded). Additionally, outcomes combining cancer recurrence
with cancer progression, or confirmed cancer-specific mortality combined with a
diagnosis of metastasis, or prostate cancer recurrence is determined by a rising PSA level
only, were excluded. Therefore, methodological flaws within the component studies will
not cause any problems in the GRADE evaluation. Consistency of results across different
studies will.
1. Quality of evidence
Table 1: Quality of Evidence Grades
Grade Definition
High We are very confident that the true effect lies close to that of the estimate of the effect.
Moderate We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect.
Very Low We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
2. Included study design
Table 2: Judgements about the quality of evidence by study design
Study Consequence
Randomised trials without important limitations – high quality evidence (+ 4 points)
Cohort studies without strengths or important limitations – low quality evidence (+ 2 points)
3. Determining the quality of evidence
Table 3: Factors that can reduce or increase the quality of the evidence
Factor Consequence
Limitations in study design or execution ↓ 1 or 2 levels
Inconsistency of results ↓ 1 or 2 levels
Indirectness of evidence ↓ 1 or 2 levels
Imprecision ↓ 1 or 2 levels
Publication bias ↓ 1 or 2 levels
Large magnitude of effect ↑ 1 or 2 levels
Dose-response gradient ↑ 1 level
All plausible confounding would reduce the demonstrated effect or increase the effect if no effect was observed
↑ 1 level
Page 85 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4. Study limitations randomised trials
Table 4: Study limitations in randomised trials Factor Explanation
Lack of allocation concealment Those enrolling patients are aware of the group (or period in a crossover trial) to which the next enrolled patient will be allocated (a major problem in “pseudo” or “quasi” randomized trials with allocation by day of week, birth date, chart number, etc.).
Lack of blinding Patient, caregivers, those recording outcomes, those adjudicating outcomes, or data analysts are aware of the arm to which patients are allocated (or the medication currently being received in a crossover trial).
Incomplete accounting of patients and outcome events Loss to follow-up and failure to adhere to the intention-to-treat principle in superiority trials; or in noninferiority trials, loss to follow-up, and failure to conduct both analyses considering only those who adhered to treatment, and all patients for whom outcome data are available. The significance of particular rates of loss to follow-up, however, varies widely and is dependent on the relation between loss to follow-up and number of events. The higher the proportion lost to follow-up in relation to intervention and control group event rates, and differences between intervention and control groups, the greater the threat of bias.
Selective outcome reporting Incomplete or absent reporting of some outcomes and not others on the basis of the results.
Other limitations Stopping trial early for benefit. Substantial overestimates are likely in trials with fewer than 500 events and that large overestimates are likely in trials with fewer than 200 events. Empirical evidence suggests that formal stopping rules do not reduce this bias. Use of unvalidated outcome measures (e.g. patient-reported outcomes). Carryover effects in crossover trial. Recruitment bias in cluster-randomized trials
5. Study limitations cohort studies
Table 5: Study limitations in observational studies Factor Explanation
Failure to develop and apply appropriate eligibility criteria (inclusion of control population)
Under- or over-matching in case-control studies
Selection of exposed and unexposed in cohort studies from different populations
Flawed measurement of both exposure and outcome Differences in measurement of exposure (e.g. recall bias in case-control studies)
Differential surveillance for outcome in exposed and unexposed in cohort studies
Failure to adequately control confounding Failure of accurate measurement of all known prognostic factors
Failure to match for prognostic factors and/or adjustment in statistical analysis
Incomplete or inadequately short follow-up Especially within prospective cohort studies, both groups should be followed for the same amount of time
6. Grading assessors
A first assessor will grade the quality of evidence for each outcome (cancer recurrence or
overall mortality or cancer-specific mortality or death from other causes) for each cancer type
with data on pre- or post-diagnosis dietary patterns or foods as exposure (bladder cancer, bowel
cancer, breast cancer, laryngeal cancer, prostate cancer). The first assessor will summarize the
findings in summary of findings tables for all evidence obtained. A second assessor will check
the consistency of the ratings of the first assessor. Disagreement about evidence were resolved
through consensus or a third party.
Page 86 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7. Summary of findings Tables
Bladder cancer
Table 1: Bladder cancer and pre-diagnosis fruit and vegetable intake
Outcomes Hazard ratio / Relative risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total fruit: HR= 0.91; 95% CI 0.62-1.33
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total vegetables: HR= 0.91; 95% CI 0.62-1.36
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cruciferous vegetables: HR= 0.87; 95% CI 0.60-1.26
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Raw cruciferous vegetables: HR= 0.73; 95% CI 0.50-1.06
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total fruit: HR= 1.09; 95% CI 0.66-1.81
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total vegetables: HR= 1.06; 95% CI 0.63-1.78
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cruciferous vegetables: HR= 0.89; 95% CI 0.53-1.48
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Raw cruciferous vegetables: HR= 0.73; 95% CI 0.44-1.21
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bowel cancer
Table 2: Bowel cancer and pre-diagnosis diet quality indices
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality
HEI-2005: HR= 0.95; 95% CI 0.78-1.16
5727 (1) + Cohort ++ and downgraded one level; data of only 1 study
WCRF/AICR score: HR= 0.79; 95% CI 0.65-0.98
3292 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
HEI-2005: HR= 0.99; 95% CI 0.77-1.27
5727 (1) + Cohort ++ and downgraded one level; data of only 1 study
WCRF/AICR score: HR= 0.70; 95% CI 0.56–0.89
3292 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis diet quality indices
Overall mortality
AHEI-2010: HR= 0.71; 95% CI 0.52-0.98
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: HR= 0.98; 95% CI 0.71-1.35
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: HR= 0.87; 95% CI 0.63-1.21
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
MMDS: HR= 0.48; 95% CI 0.32-0.74
1404 (1) + Cohort ++ and downgraded one level; data of only 1 study. Although the study has a large estimate HR<0.5, it is based on 1 study only and will therefore not be upgraded
HNFI: HR= 0.63; 95% CI 0.39-1.04
1404 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
AHEI-2010L HR= 0.72; 95% CI 0.43-1.21
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: HR= 0.87; 95% CI 0.52-1.45
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: HR= 0.84; 95% CI 0.50-1.42
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 87 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 3: Bowel cancer and pre-diagnosis prudent/healthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality
Prudent vegetable pattern: HR= 1.03; 95% CI 0.61-1.75
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Prudent vegetable pattern: HR= 1.12; 95% CI 0.69-1.84
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis prudent/healthy diet
Cancer recurrence
Prudent diet: HR= 1.13; 95% CI 0.77-1.67
1009 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality
Prudent diet: HR= 1.32; 95% CI 0.86-2.04 HR= 0.93; 95% CI 0.65-1.34
2210 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Cancer-specific mortality
Prudent diet: HR= 0.67; 95% CI 0.37-1.22
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 4: Bowel cancer and pre-diagnosis Western/unhealthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Processed meat pattern: HR= 1.53; 95% CI 0.85-2.74
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
High sugar pattern: HR= 1.27; 95% CI 0.72-2.25
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Processed meat pattern: HR= 1.82; 95% CI 1.07-3.09
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
High sugar pattern: HR= 1.02; 95% CI 0.62-1.69
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis Western/unhealthy diet
Cancer recurrence
Western diet: HR= 2.85; 95% CI 1.75-4.63
1009 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Western diet: HR= 2.32; 95% CI 1.36-3.96 HR= 1.32; 95% CI 0.89-1.97
2210 (2) ++ Cohort ++; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with an increased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies. Finally, there will be no upgrade of the evidence based on the other study that found an HR larger than >1.0 as it is only true for one of the studies
Cancer-specific mortality
Western diet: HR= 1.66; 95% CI 0.85-3.23
1201(1) + Cohort ++ and downgraded one level; data of only 1 study
Table 5: Bowel cancer and pre-diagnosis grain foods
Outcomes Hazard Ratio / Relative Risk (95% CI) No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Men Total whole grains: HR= 1.00; 95% CI 0.67-1.48 Women Total whole grains: HR= 0.91; 95% CI 0.60-1.39
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain wheat: HR= 0.97; 95% CI 0.64-1.49 Women Whole grain wheat: HR= 1.35; 95% CI 0.72-2.53
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain rye: HR= 0.90; 95% CI 0.60-1.36 Women Whole grain rye: HR= 0.93; 95% CI 0.60-1.46
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain oats: HR= 1.11; 95% CI 0.72-1.70 Women Whole grain oats: HR= 0.83; 95% CI 0.55-1.26
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain products: HR= 1.06; 95% CI 0.71-1.56 Women Whole grain products: HR= 1.10; 95% CI 0.74-1.64
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 88 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 6: Bowel cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Red and processed meat: HR= 1.03; 95% CI 0.80-1.33
3122 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Processed meat: RR= 1.21; 95% CI 0.99-1.48 HR= 1.17; 95% CI 0.97-1.42
6104 (2)
++ Cohort ++; although both studies found a ‘statistically non-significant’ association and estimates are small, there is no reason for downgrading the evidence
Unprocessed red meat: RR= 1.12; 95% CI 0.92-1.38 HR= 0.95; 95% CI 0.78-1.14
6104 (2) + Cohort ++ and downgraded one level; both studies found no statistically significant association, the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Red and processed meat: RR= 1.29; 95% CI 1.05-1.59 HR= 1.00; 95% CI 0.83-1.20 HR= 0.85; 95% CI 0.67-1.09
9226 (3) + Cohort ++ and downgraded one level; both studies found no statistically significant association, the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade one level for consistency lack of agreement between studies)
Poultry: HR= 0.87; 95% CI 0.73-1.03
3789 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Processed meat: RR= 1.09; 95% CI 0.80-1.48 HR= 1.12; 95% CI 0.90-1.39
6104 (2) ++ Cohort ++; although both studies found a ‘statistically non-significant’ association and estimates are small, there is no reason for downgrading the evidence
Unprocessed red meat: RR= 1.16; 95% CI 0.84-1.58 HR= 0.93; 95% CI 0.75-1.15
6104 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Red and processed meat: RR= 1.09; 95% CI 0.79-1.51 HR= 1.00; 95% CI 0.81-1.23 HR= 0.83; 95% CI 0.61-1.14
9226 (3) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade for consistency lack of agreement between studies)
Poultry: HR= 0.91; 95% CI 0.75-1.10
3789 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Processed meat: RR= 1.27; 95% CI 0.92-1.75
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Unprocessed red meat: RR= 1.19; 95% CI 0.87-1.64
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 1.39; 95% CI 1.00-1.92
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis protein foods
Overall mortality Processed meat: RR= 1.11; 95% CI 0.82-1.49
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Unprocessed red meat: RR= 0.75; 95% CI 0.55-1.03
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.94; 95% CI 0.68-1.30
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Processed meat: RR= 1.06; 95% CI 0.61-1.84
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Unprocessed red meat: RR= 1.13; 95% CI 0.62-2.06
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 1.10; 95% CI 0.61-1.98
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Processed meat: RR= 1.14; 95% CI 0.73-1.77
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Unprocessed red meat: RR= 0.64; 95% CI 0.40-1.03
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.87; 95% CI 0.54-1.41
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 89 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 7: Bowel cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy: RR= 0.88; 95% CI 0.72-1.09 HR= 1.16; 95% CI 0.98-1.36
6143 (3) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Milk: HR= 1.21; 95% CI 1.03-1.43 RR= 0.95; 95% CI 0.79-1.15
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Yoghurt: HR= 1.08; 95% CI 0.92-1.28
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cheese: HR= 0.87; 95% CI 0.74-1.04
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: HR= 1.17; 95% CI 0.96-1.43 RR= 0.89; 95% CI 0.65-1.22
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Milk: HR= 1.21; 95% CI 0.99-1.48 RR= 0.98; 95% CI 0.73-1.32
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade one level for consistency lack of agreement between studies)
Yoghurt: HR= 1.09; 95% CI 0.88-1.34
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cheese: HR= 0.93; 95% CI 0.76-1.14
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis dairy and alternatives
Overall mortality Total dairy: RR= 0.75; 95% CI 0.56-1.01
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Milk: RR= 0.72; 95% CI 0.55-0.94
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: RR= 0.73; 95% CI 0.44-1.23
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Milk: RR= 0.93; 95% CI 0.59-1.49
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 90 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Breast cancer
Table 8: Breast cancer and pre-diagnosis diet quality indices
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality ACS: RR= 1.00; 95% CI 0.84-1.18
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
ACS: RR= 1.06; 95% CI 0.79-1.42
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
ACS: RR= 1.02; 95% CI 0.79-1.31
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis diet quality indices
Overall mortality HEI-2005: HR= 0.40; 95% CI 0.17-0.94 HR= 0.74; 95% CI 0.55-0.99
2987 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association and a decreased risk with overall mortality. Both studies have the same direction HR<1.0 (one study even has a large estimate HR<0.5) (no downgrade nor upgrade of the evidence)
AHEI: RR= 0.85; 95% CI 0.63-1.17
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
DQIR: RR= 0.78; 95% CI 0.58-1.07
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
RFS: RR= 1.03; 95% CI 0.74-1.42
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: RR= 0.87; 95% CI 0.64-1.17
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
ACS: RR= 0.93; 95% CI 0.73-1.18
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
HEI-2005: HR= 0.12; 95% CI 0.02-0.99 HR= 0.91; 95% CI 0.60-1.40
2987 (2) ++ Cohort ++; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with a decreased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies. Finally, there will be no upgrade of the evidence based on the other study that found an HR larger than <0.50 as it is only true for one of the studies
AHEI: RR= 1.53; 95% CI 0.98-2.39 RR= 1.07; 95% CI 0.77-1.49
6832 (2) + Cohort ++ and downgraded one level; both studies found a ‘statistically non-significant’ association and a RR> 1.0. However, as the estimate of one study is relatively high (RR=1.53) and the other study RR=1.07, which almost indicates no increased nor decreased risk of cancer-specific mortality, we made the decision to downgrade the evidence with one level for inconsistency of the results
DQIR: RR= 0.81; 95% CI 0.53-1.24
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
RFS: RR= 1.54; 95% CI 0.95-2.47
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: RR= 1.15; 95% CI 0.74-1.77
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: RR= 0.85; 95% CI 0.61-1.19
4103 (1) + Cohort ++ and downgraded one level; data of only 1 study
ACS: RR= 1.44; 95% CI 0.90-2.30
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
HEI-2005: HR= 0.58; 95% CI 0.38-0.87
2317 (1) + Cohort ++ and downgraded one level; data of only 1 study
AHEI: RR= 0.52; 95% CI 0.32-0.83 RR= 0.57; 95% CI 0.42-0.77
6832 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association in the same direction HR>1.0 (one study even has a large estimate HR<0.5). Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
DQIR: RR= 0.85; 95% CI 0.54-1.34
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
RFS: RR= 0.86; 95% CI 0.54-1.37
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: RR= 0.80; 95% CI 0.50-1.26
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: RR= 0.72; 95% CI 0.53-0.99
4103 (1) + Cohort ++ and downgraded one level; data of only 1 study
ACS: RR= 0.78; 95% CI 0.56-1.07
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 91 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 9: Breast cancer and post-diagnosis low-fat diet RCTs
Overall mortality Intervention versus control HR= 0.89; 95% CI 0.65-1.21 HR= 0.91; 95% CI 0.72-1.15
5525 (2) +++ moderate RCT ++++ Downgraded by one level: Although the dietary intervention focused on reducing fat intake, intake of other nutrients changed, as did body weight; influenced outcome. No downgrading for inconsistency of results, indirectness of evidence, imprecision, or publication bias. Both RCTs are not statistically significant, however, both present a HR<1 for overall mortality
Table 10: Breast cancer and pre-diagnosis prudent/healthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Prudent diet: HR= 0.71; 95% CI 0.48-1.06
2522 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Prudent diet: Not shown HR= 0.87; 95% CI 0.61-1.23
5141 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Cancer-specific mortality
Prudent diet: Not shown HR= 0.89; 95% CI 0.59-1.35
5141 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Death from other causes
Prudent diet: Not shown HR= 0.81; 95% CI 0.40-1.61
5141 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Breast cancer and post-diagnosis prudent/healthy diet
Cancer recurrence Prudent diet: HR= 0.95; 95% CI 0.63-1.43
1901 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Prudent diet: RR= 0.78; 95% CI 0.54-1.12 HR= 0.57; 95% CI 0.36-0.90
4520 (2) ++ Cohort ++; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with a decreased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
Cancer-specific mortality
Prudent diet: RR= 1.07; 95% CI 0.66-1.73 HR= 0.79; 95% CI 0.43-1.43
4520 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade one level for consistency lack of agreement between studies)
Death from other causes
Prudent diet: RR= 0.54; 95% CI 0.31-0.95 HR= 0.35; 95% CI 0.17-0.73
4520 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association with a decreased risk of death from other cause in the same direction HR<1.0 (one study even has a large estimate HR<0.5). Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
Page 92 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 11: Breast cancer and pre-diagnosis Western/unhealthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Western diet: HR= 0.91; 95% CI 0.61-1.36
2522 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Western diet: RR= 1.40; 95% CI 0.93-2.09 HR= 1.34; 95% CI 0.93-1.94
5141 (2) ++ Cohort ++; although both studies found a ‘statistically non-significant’ association, there is no reason for downgrading the evidence
Cancer-specific mortality
Western diet: RR= 1.01; 95% CI 0.59-1.72 HR= 0.99; 95% CI 0.64-1.52
5141 (2) ++ Cohort ++; although both studies found a ‘statistically non-significant’ association and indicate no increase nor decrease with cancer-specific mortality, there is no reason for downgrading the evidence
Death from other causes
Western diet: RR= 1.95; 95% CI 1.06-3.60 HR= 3.69; 95% CI 1.66-8.17
5141 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association with an increased risk of death from other cause in the same direction HR>1.0 (one study even has a large estimate HR>2.0). Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies, even though the confidence intervals are broad
Breast cancer and post-diagnosis Western/unhealthy diet
Cancer recurrence Western diet: HR= 0.98; 95% CI 0.62-1.54
1901 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Western diet: RR= 1.53; 95% CI 1.03-2.29 HR= 1.76; 95% CI 1.10-2.81
4520 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association with an increased risk of death from other cause in the same direction HR>1.0. Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
Cancer-specific mortality
Western diet: RR= 1.01; 95% CI 0.60-1.70 HR= 1.20; 95% CI 0.62-2.32
4520 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR=1.0 (downgrade for consistency lack of agreement between studies)
Death from other causes
Western diet: RR= 2.09; 95% CI 1.30-3.36 HR= 2.15; 95% CI 0.97-4.77
4520 (2) ++ Cohort ++; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with a decreased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. Nevertheless, we will not downgrade nor upgrade the evidence with both HRs in the same direction and large (both HR>2.0) and the overlap in confidence intervals of both studies, even though the confidence intervals are broad
Page 93 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 12: Breast cancer and pre-diagnosis fruit and vegetables
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total vegetables: RR= 0.98; 95% CI 0.62-1.53 HR= 0.57; 95% CI 0.35-0.94 HR= 1.09; 95% CI 0.80-1.48
4673 (3) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR= almost 1.0 with 0.98 (downgrade for consistency lack of agreement between studies)
Total fruit: HR= 0.63; 95% CI 0.38-1.05 HR= 0.84; 95% CI 0.61-1.16
3169 (2) ++ Cohort ++; although both studies found a ‘statistically non-significant’ association and estimates are small, there is no reason for downgrading the evidence
Total fruit + vegetables: HR= 1.27; 95% CI 1.00–1.61 (low versus high intake!)
1453 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total vegetables: Not shown HR= 1.01; 95% CI 0.70-1.46
4157 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Total fruit: HR= 0.86; 95% CI 0.59-1.25
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit + vegetables: HR= 1.26; 95% CI 0.96–1.64 (low versus high intake!)
1453 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis fruit and vegetables
Cancer recurrence Cruciferous vegetables: HR= 1.10; 95% CI 0.95-1.28
11390 (1) + Cohort ++ and downgraded one level; data of only 1 study although large sample because participants of 4 cohort studies combined
Overall mortality Total vegetables: RR= 0.81; 95% CI 0.59–1.11 HR= 1.44; 95% CI 0.91-2.27
6423 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Cruciferous vegetables: HR= 1.02; 95% CI 0.80-1.30 HR= 0.99; 95% CI 0.86-1.13
15831 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Total fruit: HR= 1.38; 95% CI 0.88-2.17
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total vegetables: Not shown HR= 0.96; 95% CI 0.38-2.45
6423 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Cruciferous vegetables: HR= 0.95; 95% CI 0.59-1.54
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit: HR= 1.39; 95% CI 0.64-2.99
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 13: Breast cancer and pre-diagnosis grain foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Bread: HR= 1.31; 95% CI 0.93-1.83
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Bread: HR= 1.10; 95% CI 0.74-1.63
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 94 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 14: Breast cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Poultry RR= 0.60; 95% CI 0.39-0.92
1504 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish RR= 0.94; 95% CI 0.62-1.43
1504 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red Meat: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Sunflower/pumpkinseeds: HR= 0.87; 95% CI 0.66-1.15
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sesame/flaxseeds: HR= 0.90; 95% CI 0.68-1.19
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Poultry: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Fish: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Red meat: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Sunflower/pumpkinseeds: HR= 1.12; 95% CI 0.79-1.57
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sesame/flaxseeds: HR= 1.21; 95% CI 0.87-1.68
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis protein foods
Cancer recurrence Red meat: RR= 1.12; 95% CI 0.66-1.89
472 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Red meat: RR= 1.06; 95% CI 0.76–1.49
1982 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total meat (poultry, fish, beef, and processed meat): HR= 1.12; 95% CI 0.83-1.51
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Red meat: RR= 2.60; 95% CI 0.96-7.03 Not shown
2454 (2)
+ Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Poultry: Not show
1982 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Fish: Not shown
1982 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Total meat (poultry, fish, beef, and processed meat): HR= 0.89; 95% CI 0.50-1.60
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 95 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 15: Breast cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy RR= 0.71; 95% CI 0.44-1.14 HR= 1.18; 95% CI 0.90-1.54
5945 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Cancer-specific mortality
Total dairy: HR= 0.94; 95% CI 0.56-1.59
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis dairy and alternatives
Cancer recurrence Total dairy: HR= 1.13; 95% CI 0.83-1.54
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
Low fat dairy: HR= 1.01; 95% CI 0.78-1.32
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
High fat dairy: HR= 1.22; 95% CI 0.92-1.55
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Total dairy RR= 0.72; 95% CI 0.52–1.00 HR= 1.39; 95% CI 1.02-1.90
3875 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Low fat dairy: HR= 1.05; 95% CI 0.80-1.36
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
High fat dairy: HR= 1.64; 95% CI 1.24-2.17
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: Not shown
1982 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Table 16: Breast cancer and post-diagnosis oils and spreads
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Butter/margarine/lard: RR= 1.30; 95% CI 1.03-1.64
472 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Butter/margarine/lard: RR= 1.03; 95% CI 0.61-1.76
472 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 96 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Laryngeal cancer
Table 17: Laryngeal cancer and pre-diagnosis fruit and vegetables
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Citrus fruits: HR= 0.76; 95% 0.49-1.19
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Other fruits: HR= 0.65; 95% CI 0.39-1.07
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Vegetables: HR= 0.57; 95% CI 0.35-0.94
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 18: Laryngeal cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Meat: HR= 0.50; 95% CI 0.30-0.83
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Poultry: HR= 0.90; 95% CI 0.55-1.46
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish: HR= 0.91; 95% CI 0.59-1.39
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Eggs: HR= 1.22; 95% CI 0.74-2.00
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 19: Laryngeal cancer and pre-diagnosis grain foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Bread: HR= 0.54; 95% CI 0.32-0.90
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Pasta: HR= 1.25; 95% CI 0.76-2.04
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Potatoes: HR= 1.02; 95% CI 0.64-1.64
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 20: Laryngeal cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Milk: HR= 1.58; 95% CI 0.99-2.55
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cheese: HR= 0.70; 95% CI 0.44-1.12
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 21: Laryngeal cancer and pre-diagnosis oils and spreads
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Butter: HR= 1.11; 95% CI 0.69-1.80
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Olive oil: HR= 0.71; 95% CI 0.44-1.16
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 97 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Non-Hodgkin Lymphoma
Table 22: Non-Hodgkin Lymphoma and pre-diagnosis fruit and vegetables
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total fruit and vegetables: HR= 0.68; 95% CI 0.49-0.95
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit: HR= 0.91; 95% CI 0.70-1.18 HR= 1.03; 95% CI 0.90-1.19
2907 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies). Additionally, one study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Total vegetables: HR= 0.58; 95% CI 0.38-0.89 HR= 0.98; 95% CI 0.85-1.12
2907 (2) + Cohort ++; One study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Cruciferous vegetables: HR= 0.91; 95% CI 0.67–1.24
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bean vegetables: HR= 1.14; 95% CI 0.85-1.54
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Green leafy vegetables: HR= 0.71; 95% CI 0.51-0.98
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red vegetables: HR= 1.03; 95% CI 0.76-1.38
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Yellow vegetables: HR= 0.93; 95% CI 0.69-1.25
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Citrus fruits: HR= 0.73; 95% CI 0.54-0.99
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality Total fruit and vegetables: HR= 0.70; 95% CI 0.45-1.10
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit: HR= 1.04; 95% CI 0.74-1.45 HR= 1.04; 95% CI 0.88-1.24
2907 (2) + Cohort ++; One study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Total vegetables: HR= 0.58; 95% CI 0.33-1.03 HR= 0.98; 95% CI 0.83-1.16
2907 (2) + Cohort ++; One study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Cruciferous vegetables: HR= 0.75; 95% CI 0.49–1.14
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bean vegetables: HR= 1.05; 95% CI 0.71-1.55
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Green leafy vegetables: HR= 0.82; 95% CI 0.54-1.23
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red vegetables: HR= 1.11; 95% CI 0.76-1.62
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Yellow vegetables: HR= 1.11; 95% CI 0.77-1.61
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Citrus fruits: HR= 0.81; 95% CI 0.54-1.20
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 23: Non-Hodgkin Lymphoma and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Red meat: HR= 1.00; 95% CI 0.87-1.15
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Processed meat: HR= 0.94; 95% CI 0.82-1.08
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish: HR= 0.90; 95% CI 0.78-1.03
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Legumes: HR= 0.88; 95% CI 0.76-1.01
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality Red meat: HR= 0.95; 95% CI 0.81-1.13
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Processed meat: HR= 0.94; 95% CI 0.79-1.12
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish: HR= 0.91; 95% CI 0.76-1.08
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Legumes: HR= 0.86; 95% CI 0.72-1.02
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 98 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 24: Non-Hodgkin Lymphoma and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Soy foods: HR= 0.93; 95% CI 0.76-1.14
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total dairy: HR= 1.14; 95% CI 1.00-1.31
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality Soy foods: HR= 0.92; 95% CI 0.72-1.18
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total dairy: HR= 1.16; 95% CI 0.98-1.37
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 99 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Prostate cancer
Table 25: Prostate cancer and post-diagnosis diet quality indices
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Med diet score: HR= 0.78; 95% CI 0.67-0.90
4538 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Med diet score: HR= 1.01; 95% CI 0.75-1.38
4538 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 26: Prostate cancer and post-diagnosis prudent/healthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Prudent diet: RR= 0.64; 95% CI 0.44-0.93
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Prudent diet: RR= 0.46; 95% CI 0.17-1.24
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 27: Prostate cancer and post-diagnosis Western/unhealthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Western diet: RR= 1.67; 95% CI 1.16-2.42
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Western diet: RR= 2.53; 95% CI 1.00-6.42
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 28: Prostate cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer-specific mortality
Fish: HR= 0.52; 95% CI 0.30-0.91
2161 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 29: Prostate cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy: HR = 1.76; 95% CI 1.21-2.55
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
High-fat dairy: HR= 1.22; 95% CI 1.08-1.38
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Low-fat dairy: HR= 1.17; 95% CI 1.05-1.29
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: HR = 2.41; 95% CI 0.96-6.02
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
High-fat dairy: HR= 1.30; 95% CI 0.97-1.73
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Low-fat dairy: HR= 1.16; 95% CI 0.88-1.53
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 100 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
PRISMA Checklist
Section/topic # Checklist item Reported on page #
TITLE
Title 1 Identify the report as a systematic review, meta-analysis, or both. 1
ABSTRACT
Structured summary 2 Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number.
2
INTRODUCTION
Rationale 3 Describe the rationale for the review in the context of what is already known. 4
Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).
6
METHODS
Protocol and registration 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.
10
Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered,
language, publication status) used as criteria for eligibility, giving rationale. 6-8
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.
6-8
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.
36
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable,
included in the meta-analysis). 6
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.
6
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.
6-8
Risk of bias in individual studies
12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.
9
Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means). 7
Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e.g., I2) for each meta-analysis.
9
Page 1 of 2
Page 101 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
PRISMA Checklist
Section/topic # Checklist item Reported on page #
Risk of bias across studies 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within studies).
9
Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done,
indicating which were pre-specified. NA
RESULTS
Study selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.
10 and flowchart in supplement S1
Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations.
Supplement S2
Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12). Data on risk of bias of each study can be obtained on request
Results of individual studies 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.
Supplement S2
Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency. NA
Risk of bias across studies 22 Present results of any assessment of risk of bias across studies (see Item 15). Supplement S3
Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]). NA
DISCUSSION
Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).
21-22
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).
23
Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research. 24
FUNDING
Page 102 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
PRISMA Checklist
Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for the systematic review.
25
From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097
For more information, visit: www.prisma-statement.org.
Page 2 of 2
Page 103 of 103
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
The impact of dietary patterns and the main food groups on mortality and recurrence in cancer survivors:
a systematic review of current epidemiological literature
Journal: BMJ Open
Manuscript ID bmjopen-2016-014530.R2
Article Type: Research
Date Submitted by the Author: 18-Aug-2017
Complete List of Authors: Jochems, Sylvia; Maastricht University, NUTRIM School of Nutrition, Metabolism and Toxicology; University of Birmingham, Cancer and Genomic Sciences
Van Osch, Frits; Maastricht University, NUTRIM School of Nutrition, Metabolism and Toxicology; University of Birmingham, Cancer and Genomic Sciences Bryan, Richard; University of Birmingham, Cancer and Genomic Sciences Wesselius, Anke; Maastricht University, NUTRIM School of Nutrition, Metabolism and Toxicology van Schooten, Frederik; Maastricht University, NUTRIM School of Nutrition, Metabolism and Toxicology Cheng, Kar Keung; University of Birmingham, Public Health and Epidemiology Zeegers, Maurice; University of Maastricht, NUTRIM School of Nutrition, Metabolism and Toxicology; Maastricht University, CAPHRI Care and Public
Health Research Institute
<b>Primary Subject Heading</b>:
Epidemiology
Secondary Subject Heading: Oncology, Public health, Epidemiology
Keywords: cancer survivors, mortality, cancer recurrence, food, dietary pattern, diet
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open on D
ecember 7, 2020 by guest. P
rotected by copyright.http://bm
jopen.bmj.com
/B
MJ O
pen: first published as 10.1136/bmjopen-2016-014530 on 19 F
ebruary 2018. Dow
nloaded from
For peer review only
The impact of dietary patterns and the main food groups on 1
mortality and recurrence in cancer survivors: 2
a systematic review of current epidemiological literature 3
4
Sylvia H.J. Jochems (1,2), Frits H.M. van Osch (1,2), Richard T. Bryan (1), Anke Wesselius 5
(2), Frederik J. van Schooten (2), K.K. Cheng (3), Maurice P. Zeegers (2,4) 6
7
(1) Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, 8
United Kingdom 9
(2) NUTRIM School of Nutrition and Translational Research in Metabolism, Maastricht 10
University, Maastricht, The Netherlands 11
(3) Institute of Applied Health Research, Public Health, Epidemiology and Biostatistics, 12
University of Birmingham, Birmingham, United Kingdom 13
(4) CAPHRI School for Public Health and Primary Care, Maastricht University, 14
Maastricht, The Netherlands 15
16
17
Corresponding author contact information: 18
Sylvia H.J. Jochems 19
Institute of Cancer and Genomic Sciences, University of Birmingham, 21
B15 2TT Birmingham, United Kingdom 22
23
Page 1 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
2
ABSTRACT 24
Objective: To determine whether there is an association between dietary patterns/indices and 25
foods from the main food groups (highest versus lowest intakes) prior to or after cancer 26
diagnosis and mortality and cancer recurrence in cancer survivors. 27
Participants: Survivors of common cancers with a 10-year survival rate of 50% or more: 28
bladder, bowel, breast, cervical, kidney, laryngeal, prostate, testicular, uterine cancer, 29
malignant melanoma, and (non-)Hodgkin lymphoma. 30
Outcome measures: Mortality (overall, cancer-specific, from other causes) and cancer 31
recurrence. 32
Information sources: PubMed, Embase and the Cochrane Library were searched from 33
inception to April 2017. Additional studies were identified by searching reference lists. Two 34
authors independently screened titles and abstracts, assessed study quality, and extracted the 35
data. 36
Results: A total of 38 studies were included. The risk of bias was rated low for the included 37
RCTs and moderate for the cohort studies. The quality of evidence was assessed with the 38
GRADE approach and was rated moderate (RCTs), and (very)low (cohort studies). Reducing 39
the amount of fat after diagnosis appears to decrease the risk of breast cancer recurrence. 40
Adherence to a high-quality diet and prudent diet after diagnosis appears to decrease the risk 41
of death from other causes (and overall mortality for high-quality diet) in breast cancer 42
survivors. Adherence to a Western diet, before and after diagnosis, appears to increase the 43
risk of overall mortality and death from other causes amongst breast cancer survivors. 44
Evidence from studies amongst other cancer survivors were too limited or could not be 45
identified. 46
Conclusion: For many cancer survivors, there is little evidence to date to indicate that 47
particular dietary behaviours influence outcomes with regard to recurrence and mortality. 48
Page 2 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
3
Notwithstanding, limited evidence suggests that a low-fat diet, a high-quality diet, and a 49
prudent diet are beneficial for breast cancer survivors, whilst a Western diet is detrimental for 50
breast cancer survivors. 51
52
Strengths and limitations 53
- Dietary patterns/indices and whole foods reflect the complexity of dietary intake and 54
capture synergistic relationships between various dietary constituents 55
- Most studies investigating dietary patterns/indices and foods before diagnosis do not 56
consider potential modifications in dietary intake after cancer diagnosis 57
- Cohort studies provide weaker empirical evidence than RCTs for examining 58
relationships between dietary exposure and mortality and cancer recurrence 59
60
Page 3 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4
INTRODUCTION 61
As cancer survival rates continue to improve, there is an increased need to identify 62
modifiable lifestyle factors amongst cancer survivors in order to improve long-term health. 63
Adherence to a diet rich in fruit and vegetables could decrease the risk of several types of 64
cancer and increase overall life expectancy[1,2], the suggestion that epigenetic aberrations 65
occurring in cancer could be altered by nutrients makes it plausible that dietary changes after 66
successful cancer treatment could improve prognosis[3,4]. 67
Although cancer survivors are responsive to health promotion[5,6], a recent study has 68
indicated that survivors had poorer diets than individuals without cancer[7]. One possible 69
explanation could be the difficulty for cancer survivors in adopting a healthier diet without 70
clear evidence that it will improve their survival[8]. While guidelines have been well 71
documented for the prevention of cancer, many uncertainties remain for nutrition after cancer 72
treatment[9]. A systematic review, as part of the Continuous Update Project (CUP) of the 73
World Cancer Research Fund International, was published on diet, nutrition, physical activity 74
and survival in breast cancer survivors[10]. The independent panel of scientists concluded 75
that the evidence to date was not strong enough to make specific recommendations for breast 76
cancer survivors[11]. A recent meta-analysis investigating the role of diet on overall 77
mortality and recurrence among cancer survivors concluded that adherence to a Western diet 78
is positively associated with overall mortality, and a high-quality diet / healthy dietary pattern 79
is inversely associated with overall mortality amongst all cancer survivors[12]. 80
In the setting of survivors of cancers with a 10-year survival rate ≥50%, this 81
systematic review provides a structured overview of RCTs and cohort studies addressing the 82
relationship between adherence to dietary patterns/indices and intake of foods from the main 83
food groups, prior to or after cancer diagnosis, and health outcomes including cancer 84
Page 4 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
5
recurrence, cancer-specific mortality, overall mortality, and death from other causes than 85
their cancer. 86
Given that these survivors have the potential for long-term survival, they may be most 87
likely to benefit from dietary changes to prevent or delay cancer recurrence and improve 88
survival. Notwithstanding, many of these survivors will die from other causes such as 89
cardiovascular disease – even if the dietary exposures identified will not help the investigated 90
outcomes, it could be desirable to follow a diet that could help reduce other conditions. 91
92
METHODS 93
Search strategy 94
From inception up to April 2017, Pubmed, Embase and the Cochrane Library were 95
searched to find English language articles of original and published randomized trials and 96
observational studies to answer the following research question: does adherence to/intake of 97
dietary patterns/indices and foods (highest versus lowest adherence/intake) prior to or after 98
cancer diagnosis, increase or decrease the risk of mortality and cancer recurrence amongst 99
cancer survivors of common cancers with a 10-year survival rate of 50% or more? This 100
research question was developed using the PICO framework (supporting data review protocol 101
File S1). Search strategies included search terms related to dietary patterns, dietary indices, 102
diet quality, foods from the main food groups, and outcomes of interest, including overall 103
mortality, cancer-specific mortality, death from other causes, and recurrence of cancer. 104
Additionally, studies were identified by searching reference lists of relevant studies, literature 105
reviews and meta-analyses. After the search was completed, articles were screened and 106
selected independently based on the title and abstract by two of the authors (SJ and FvO). 107
The data extraction was performed independently by the same authors (SJ and FvO) and any 108
disagreements about study inclusion were resolved through consensus or a third party. 109
Page 5 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
6
110
Inclusion and exclusion criteria 111
Eligibility criteria included adult survivors of cancer (no sex or age restriction) who 112
were defined as individuals who had been diagnosed with a primary cancer, received cancer 113
therapy, and were in remission or had recovered completely from cancer. Considered cancer 114
types were the commonly-occurring cancers in the Western world with a 10-year net survival 115
of at least 50% (based on cancer diagnoses of men and women during 2010-2011 in England 116
and Wales)[13]. These include in decreasing order of net survival: testicular cancer (98%), 117
malignant melanoma (MM) (89%), prostate cancer (84%), Hodgkin lymphoma (HL) (80%), 118
breast cancer (78%), uterine cancer (77%), non-Hodgkin lymphoma (NHL) (63%), cervical 119
cancer (63%), laryngeal cancer (62%), bowel cancer (57% including both colon and rectal 120
cancer), bladder cancer (50%), and kidney cancer (50%). In the statistical analyses 121
adjustments had to be made for at least age and disease stage at baseline and, where possible, 122
for cancer treatment. Excluded papers did not state hazard ratios (HRs) or relative risks 123
(RRs), nor 95% confidence intervals (95% CI); neither did they provide information on 124
disease stage or tumour grade or therapy. Additionally, studies were excluded when outcomes 125
were combined, such as mortality and cancer progression, mortality and diagnosed 126
metastasis, or where prostate cancer recurrence was determined by a rising PSA level 127
only. 128
129
Dietary exposure 130
Dietary patterns/indices that were considered were assessed by index-based methods 131
and data-driven approaches, such as principal component analysis (factor analysis) and 132
cluster analysis[14]. The following diet scores were considered: the Healthy Eating Index 133
2005 (HEI-2005)[15,16], the alternate Healthy Eating Index 2010 (AHEI)[17], the World 134
Page 6 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7
Cancer Research Fund and the American Institute for Cancer Research (WCRF/AIRC) 135
dietary guidelines adherence score[18] and the American Cancer Society diet-specific 136
recommendations for cancer prevention (ACS)[19], the recommended food score (RFS)[20], 137
the Diet Quality Index-Revised (DQIR)[21], the Dietary Approaches to Stop Hypertension 138
diet (DASH) diet[22], the Healthy Nordic Food Index (HNFI)[23], and the alternate 139
Mediterranean diet (aMed)[24,25]; empirical patterns reviewed included a low-fat diet, a 140
prudent/healthy diet, and a Western/unhealthy diet. The HEI-2005 was developed by the US 141
Department of Agriculture and targets foods that could possibly reduce the risk of chronic 142
diseases and include fruits, vegetables, fibre, soy, nuts, ratio white and red meat, alcohol, 143
trans fat, saturated fat ratio, and multivitamin use[15]. Five years later, the AHEI was 144
introduced, which differs from the HEI-2005 by distinguishing quality within food groups 145
and recognizing health benefits of unsaturated oils[26]. The RFS includes the foods fruits, 146
vegetables, whole grains, dairy and protein foods low in fat. Diet diversity and moderation 147
was addressed by the DQIR and included fruits, vegetables, cholesterol, total fat, saturated 148
fat, iron, calcium, and fat/sugar moderation. The aMed is based on the original Mediterranean 149
diet score and includes fruits, vegetables, legumes, nuts, whole grains, red and processed 150
meat, moderate alcohol, and the ratio of monounsaturated and saturated fat[27,28]. In 151
addition, whole foods of the main food groups (UK Eatwell Guide)[29] were considered. The 152
composition of the investigated groups was as follows: (I) fruit and vegetables including 153
citrus fruits, stone fruits, soft fruits, fleshy fruits, vine fruits, flower vegetables, leafy 154
vegetables, stem vegetables, fruit vegetables, mushrooms, bulbs and roots; (II) grain foods 155
including potatoes, bread, rice, pasta and cereal; (III) protein foods including unprocessed 156
meat, red meat, poultry, fish, eggs, tofu, nuts, seeds, pulses, legumes and beans; (IV) dairy 157
and alternative products including yoghurt, milk, cheese; (V) oils and spreads including 158
vegetable oils, spreads. Although processed (red) meats are not included in the main food 159
Page 7 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
8
groups recommended by the UK Eatwell Guide, lean red meats (rich in protein, iron, zinc, 160
selenium and B vitamins) can be part of a healthy diet. Studies that made no distinction 161
between (lean) red meats and processed meats in their estimates, were still included in this 162
systematic review – they will however, be interpreted with caution. Information on intake of 163
food was obtained before or after cancer diagnosis with food records, food frequency 164
questionnaires (FFQ) (self-administered or via an interview), or twenty-four-hour recalls, and 165
expressed in servings or (milli)grams per day/week/month. No restrictions were made for 166
time of follow-up, and timing or frequency of dietary intake. 167
168
Mortality and cancer recurrence 169
Considered endpoints were overall mortality, cancer-specific mortality, death from 170
other causes, and cancer recurrence. The cause of death was confirmed via death certificates 171
or the National Death Index in each of the studies. Cancer recurrence was defined as a new 172
occurrence of cancer after a period of time during which the cancer could not be detected at 173
the same or at a different site to the initial primary tumour. Cancer recurrence had to be 174
confirmed by a biopsy, scan, medical record, cancer registry, or treating physician. 175
176
Assessment risk of bias and level of quality 177
The Cochrane Collaboration risk of bias assessment tools were used for appraisal of 178
RCTs [30]and cohort studies[31]. For the two RCTs the RoB 2.0 tool (a revised tool for risk 179
of bias in randomized trials) was used to evaluate the risk of bias. Cohort studies were 180
appraised with an adjusted version of the ROBINS-I tool[30,31]. Levels of quality were 181
determined with the GRADE approach[32]; evidence from RCTs or multiple double-182
upgraded observational studies were considered as high quality, downgraded RCTs or 183
upgraded observational studies were considered as moderate quality, double-downgraded 184
Page 8 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
9
RCTs or observational studies were considered as low quality, and triple-downgraded RCTs, 185
downgraded observational studies or case series/case reports were considered as very low 186
quality[32]. Factors reducing the quality of the evidence include limitations in study design, 187
inconsistency between study results, indirectness of evidence, imprecision, and publication 188
bias. Factors increasing the quality of the evidence include a large magnitude of effect, 189
correction for all plausible confounding that could reduce the demonstrated effect or increase 190
the effect if no effect was observed, and presence of a dose-response gradient. For 191
observational studies, this could intent controlling for key knows risk factors and 192
confounders. GRADE separates the process of assessing the quality of evidence from making 193
recommendations. To determine whether evidence for an association between dietary 194
patterns/indices or foods and mortality or cancer recurrence amongst cancer survivors was 195
conclusive, the risk of bias and levels of quality were considered. 196
197
RESULTS 198
The search resulted in 2883 citations after removal of duplicates. After screening the 199
titles and abstracts, 95 full-text articles were assessed for eligibility - a total of 2 RCTs and 36 200
cohort studies were included in this systematic review. No studies could be identified for 201
cervical, kidney, testicular, uterine cancer, HL or MM survivors. Dietary patterns/indices 202
could be identified for bowel, breast, prostate cancer, and NHL. Whole foods from the main 203
food groups could be identified for bladder, bowel, breast, laryngeal, prostate cancer and 204
NHL survivors. 205
The protocol used for this systematic review is available in the supporting data (File 206
S1). A detailed search strategy is provided in Table 1 and the search was adapted accordingly 207
for the individual cancers and databases (File S1). The review was written according to the 208
PRISMA guidelines[33]. A summary of the number of studies for pre-diagnosis dietary 209
Page 9 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
10
patterns/indices (Table 2) and post-diagnosis dietary patterns/indices (Table 3) and mortality 210
and cancer recurrence is provided. Additionally, tables with the number of studies for pre-211
diagnosis food intake (Table 4) and post-diagnosis food intake (Table 5) and mortality and 212
cancer recurrence is given. The study characteristics including the HRs/RRs with their 213
corresponding 95% CI are provided in the supporting data (File S2). 214
Templates of the RoB 2.0 and ROBINS-I tools can be found in the supporting data 215
(File S1). Results for the assessment of the risk of bias for each individual RCT (RoB 2.0) 216
and cohort study (ROBINS-I) will be provided on request. Briefly, the included RCTs 217
investigating a low-fat diet and mortality amongst breast cancer survivors indicated a low risk 218
of bias[34]; the included cohort studies all had a moderate risk of bias[35]. 219
An overview of the GRADE ratings with comments can be found in the supporting 220
data (File S3). As the risk of bias was rated ‘low’ and ‘moderate’, there was no reason to 221
downgrade the quality of evidence on this matter. The quality level of the body of evidence 222
of the studies was rated ‘very low’, ‘low’ and ‘moderate’ by two of the authors (SJ and FvO) 223
when applying the grading system developed by the GRADE collaboration[32]. The evidence 224
for a low-fat diet from the included RCTs was downgraded from ‘high’ to ‘moderate’ due to 225
confounding factors. Evidence from cohort studies with corresponding exposures and 226
outcomes, remained at a ‘low’ level of evidence or was downgraded from ‘low’ to ‘very low’ 227
due to inconsistency, directness, and publication bias. 228
229
Bladder cancer 230
A total of 1 cohort study could be identified for bladder cancer survivors regarding 231
fruit and vegetable consumption. The study of Tang et al. investigated pre-diagnosis fruit and 232
vegetable consumption with data from 239 male and female bladder cancer survivors from 233
the Roswell Park Cancer Institute (RPCI) Tumor Registry[36]. After an average of 8-year 234
Page 10 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
11
follow-up, no associations were observed between overall mortality or bladder cancer-235
specific mortality when comparing survivors with the highest intakes of total fruit, total 236
vegetables or other cruciferous vegetables (raw or cooked) with those in the lowest intake 237
group. An association was, however, observed for broccoli intake (≥1 versus <1 serving per 238
month) with overall mortality (broccoli raw HR=0.57; 95% CI 0.39-0.83, broccoli cooked 239
HR=0.67; 95% CI 0.49-0.91) and bladder cancer-specific mortality (broccoli raw HR=0.43; 240
95% CI 0.25-0.74). The intake of other raw and cooked vegetables including cabbage, 241
cauliflower, Brussels sprouts, kale, turnip, collard or mustard greens was not related with 242
mortality[36]. 243
In summary, no conclusive evidence for an association between vegetable and fruit 244
intake and mortality amongst bladder cancer survivors could be provided, as evidence for 245
each exposure and outcome was based on the results of one study only. 246
247
Bowel cancer 248
A total of 12 cohort studies could be identified for bowel cancer survivors. Three 249
observational cohort studies could be identified investigating the role of a pre- and post-250
diagnosis prudent diet on mortality in bowel cancer survivors. Results of the Cancer and 251
Leukemia Group B (CALGB) study indicated no associations between a prudent diet after 252
cancer diagnosis and decreased mortality[37]. However, there was a higher overall mortality 253
amongst these survivors with the highest post-diagnosis intakes of a Western diet in 254
comparison with those in the lowest category (HR=2.32; 95% CI 1.36-3.96)[37]. When 255
comparing participants in the Familial Bowel Cancer Registry (FBCR) with the highest and 256
lowest intakes of a prudent diet before cancer diagnosis, no associations were found with 257
mortality[38]. Besides a prudent diet, two other dietary patterns comparable with a Western 258
diet were identified in this study: a high processed meat pattern and a high sugar pattern diet. 259
Page 11 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
12
No associations were reported for the pattern high in sugar and mortality when comparing the 260
highest to the lowest intake group, whereas a high processed meat pattern was specifically 261
related to increased colon cancer mortality (HR=2.13; 95% CI 1.03-4.43). This relationship 262
between a processed meat pattern and bowel cancer survival was modified by sex[38]. In the 263
Nurses' Health Study (NHS), no associations were observed between adherence to the AHEI, 264
DASH, or AMED score, a prudent diet, or a Western diet after diagnosis and mortality in 265
these bowel cancer survivors[24]. It should be noted, however, that even though there was 266
‘no statistically significant’ result for the role of a post-diagnosis Western diet in this study, 267
the HR was >1 (HR= 1.31; 95% CI 0.89-1.97)[39] as observed in the earlier described study 268
of Meyerhardt et al. (HR= 2.32; 95% CI 1.36-3.96)[37]. Adherence to the HEI diet score was 269
investigated in a large study including 5,727 male and female survivors in the USA and 270
indicated no association between pre-diagnosis adherence to the HEI-2005 score with overall 271
mortality or cancer-specific mortality[40]. Recently, a German study examined adherence to 272
the Modified Mediterranean Diet Score (MMDS) and the Healthy Nordic Food Index (HNFI) 273
and found that post-diagnosis adherence to this MMDS was associated with a decreased risk 274
of overall mortality amongst bowel cancer survivors (HR=0.48; 95% CI 0.32-0.74)[41]. In 275
the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort, data from 276
participants of 10 European countries was analysed on adherence to WCRF/AICR diet scores 277
and intake of total dairy, milk, yoghurt, cheese, red meat, and poultry[42–44]. Pre-diagnosis 278
adherence to this high-quality diet score indicated a decreased risk of overall mortality 279
amongst bowel cancer survivors (HR=0.79; 95% CI 0.65-0.98)[43]. No evidence of an 280
association with mortality was observed for foods from the main food groups, including 281
fruits, vegetables, dairy, or protein foods amongst these bowel cancer survivors[42,44]. The 282
study by Yang et al. indicated a protective association with milk consumption and overall 283
mortality after a diagnosis of bowel cancer (RR=0.72; 95% CI 0.55-0.94)[45]. Additionally, 284
Page 12 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
13
no association could be found for total dairy intake and mortality in this study[45]. Whole 285
gains, another food group investigated in bowel cancer survivors, were not associated with 286
overall mortality amongst 1,119 Danish, Swedish and Norwegian bowel cancer survivors in 287
the HELGA cohort[46]. Carr et al. reported that red and processed meat consumption was not 288
associated with a poorer survival amongst stage I–III bowel cancer survivors in a follow-up 289
study of the Darmkrebs: Chancen der Verhutung durch Screening (DACHS) study[47]. 290
However, it should be noted that the authors investigated red and processed meat combined 291
and they suggest that major changes in the consumption of red meat measured at 5-year 292
follow-up could have influenced survival[47]. The study of McCullough et al. indicated an 293
association with mortality when comparing highest versus lowest pre- and post-diagnosis red 294
and processed meat consumption for overall mortality (RR=1.29; 95% CI 1.05-1.59) and 295
death from other causes than bowel cancer (RR=1.39; 95% CI 1.00-1.92)[48]. It should be 296
noted that the authors combined the consumption of red and processed meat in these 297
estimates, and that there were no associations found for ‘fresh’ meats and mortality[48]. 298
In summary, no conclusive evidence for an association between adherence to a high-299
quality diet, a prudent diet, a Western diet, and the consumption of fruits, vegetables, meats 300
or dairy and mortality in bowel cancer survivors could be provided, as evidence for each 301
exposure and outcome was based on the results of one study only or on inconsistent results. 302
303
Breast cancer 304
A total of 2 RCTs and 16 cohort studies could be identified for breast cancer 305
survivors. Two dietary intervention trials amongst breast cancer survivors met the inclusion 306
criteria[49,50]. The study of Chlebowski et al. aimed to reduce post-diagnosis dietary fat 307
intake to almost one sixth of total energy intake while maintaining nutritional adequacy in 308
Page 13 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
14
women participating in the Women’s Intervention Nutrition Study (WINS)[49]. Breast cancer 309
survivors in the intervention group were informed extensively on maintaining weight based 310
on energy intake, whilst minimum dietary advice on nutrient intake was provided to breast 311
cancer survivors in the control group. Women in the intervention group had a lower dietary 312
fat intake compared to those in the control group, whereas no differences could be observed 313
for a lower energy or higher dietary fibre intake. According to the authors of this RCT, there 314
was no association with overall mortality between women adhering to a low-fat diet and 315
women given minimum dietary advice (HR=0.89; 95% CI 0.65-1.21). However, for relapse 316
events (including local, regional, distant, or ipsilateral breast cancer recurrence or new 317
contralateral breast cancer) the HR of an event in the intervention group compared to the 318
control group was HR=0.76; 95% CI 0.60-0.98. This could indicate that a lifestyle 319
intervention reducing dietary fat intake, could improve relapse-free survival of breast cancer 320
survivors[49]. In the Women’s Healthy Eating and Living (WHEL) study breast cancer 321
survivors in the intervention group received telephone counselling with additional cooking 322
classes and brochures to support adherence to a post-diagnosis diet high in fruit (3 323
servings/day), high in vegetables (5 servings/day and 16oz of vegetable juice), high in fibre 324
(30g/day), and low in fat (15-20% of energy intake from fat)[50]. In the control group, breast 325
cancer survivors received written advice to eat at least 5 portions of fruit and vegetables each 326
day (5-a-day advice). Differences between the former and latter groups in mean consumption 327
of vegetables (+65%), fruit (+25%), fibre (+30%), and energy from fat (-13%) were observed 328
at 4 years. The authors of this trial reported that no associations were observed for overall 329
survival when comparing women in the intervention group with those in the control group 330
(HR=0.91; 95% CI 0.72-1.15)[50]. Although the results for overall mortality in the trials 331
were statistically non-significant, the HRs of both studies were <1 (HR=0.89; 95% CI 0.65-332
1.21 [49] and HR=0.91; 95% CI 0.72-1.15 [50]). 333
Page 14 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
15
Post-diagnosis dietary indices were examined in the Health, Eating, Activity, and Lifestyle 334
(HEAL) study[51], Women’s Health Initiative’s Dietary Modification Trial and 335
Observational Study (WHI)[52], Nurses’ Health Study (NHS)[25,53] and Cancer Prevention 336
Study II Nutrition Cohort (CPS-II)[54]. McCullough et al. demonstrated that pre- and post-337
diagnosis adherence to the ACS diet amongst breast cancer survivors in the CPS-II cohort 338
was not association with breast cancer-specific mortality[54]. It should be noted, however, 339
that an inverse association was observed for the continuous post-diagnosis diet scores and 340
other causes of death RR=0.88; 95% CI 0.79-0.99)[54]. Whilst no associations were found 341
between pre- and post-diagnosis fruit and vegetable intake and the intake of whole grains, 342
detrimental associations were found with post-diagnosis red and processed meat consumption 343
and overall mortality and death from other causes (respectively RR=0.64; 95% CI 0.49-0.84 344
and RR=0.57; 95% CI 0.39-0.82)[54]. In the NHS, post-diagnosis dietary DQIR, RFS, aMed, 345
AHEI, and DASH scores were not associated with overall mortality or breast cancer-specific 346
mortality[25,53]. Closer adherence to DASH and AHEI were, however, related to a lower 347
risk of death from other causes than breast cancer (respectively RR=0.72; 95% CI 0.53-0.99 348
and RR=0.57; 95% CI 0.42-0.77)[53]. George et al. examined post-diagnosis adherence to 349
the HEI-2005 scores and concluded that there was an association with a decreased risk of 350
mortality (overall mortality HR=0.40; 95% CI 0.17-0.94 and breast cancer-specific mortality 351
HR=0.12; 95% CI 0.02-0.99)[51]. In the WHI cohort, results of post-diagnosis adherence to 352
the HEI-2005 scores indicated that women who consumed better quality diets had a 26% 353
lower risk of overall mortality (HR=0.74; 95% CI 0.55-0.99) and a 42% lower risk of death 354
from non-breast cancer related death (HR=0.58; 95% CI 0.38-0.87)[52]. Even though the 355
result for breast cancer-specific mortality and adherence to the HEI-2005 score in this study 356
was statistically non-significant (HR= 0.91; 95% CI 0.60-1.40), the HR is <1, as observed for 357
women in the HEAL study regarding cancer-specific mortality [51]. Results of the NHS 358
Page 15 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
16
study indicated that a post-diagnosis prudent diet was not associated with overall or breast 359
cancer-specific mortality whilst death from other causes was associated with a prudent diet 360
after diagnosis when comparing breast cancer survivors of the highest and lowest intake 361
group (HR=0.54; 95% CI 0.31-0.95)[55]- adherence to a prudent diet before diagnosis was 362
not associated with mortality amongst breast cancer survivors in the NHS[55]. Both pre- and 363
post-diagnosis adherence to a Western diet was associated with death from other causes 364
(respectively RR=1.95; 95% CI 1.06-3.60 and RR=2.31; 95% CI 1.23-4.32)[55]. The study of 365
Kwan et al. concludes no associations between adherence to a pre- or post-diagnosis Western 366
diet and overall mortality, breast cancer-specific mortality or cancer recurrence[56]. The HR 367
for a Western diet and death from other causes was, however, >1 (HR=2.15; 95% CI 0.97-368
4.77)[56], and therefore in agreement with the HR for a Western diet and death from other 369
causes observed in the study of Kroenke et al. (RR= 2.09; 95% CI 1.30-3.36)[55]. In the Life 370
After Cancer Epidemiology (LACE) study, post-diagnosis adherence to a prudent diet in 371
women with early-stage breast cancer resulted in a decreased risk of death from other causes 372
(HR=0.35; 95% CI 0.17-0.73) and overall mortality (HR=0.57; 95% CI 0.36-0.90)[56]. The 373
study of Vrieling et al. investigated associations between a ‘healthy’ and ‘unhealthy’ pre-374
diagnosis dietary pattern and mortality in German breast cancer survivors in the Mammary 375
carcinoma Risk factor Investigation (MARIE) study[57]. The characteristics of the defined 376
healthy diet are comparable with a prudent diet; nevertheless, no associations between the 377
highest and lowest intake of this defined ‘healthy’ diet before cancer diagnosis and mortality 378
in breast cancer survivors were observed. However, the results did indicate that a higher 379
intake of an ‘unhealthy’ diet could increase the risk of death from other causes (HR=3.69; 380
95% CI 1.66-8.17) amongst breast cancer survivors compared to those with the lowest intake 381
of this diet[57]. 382
Page 16 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
17
The majority of studies investigating pre- or post-diagnosis fruit and vegetable intake 383
indicated no association with mortality in breast cancer survivors. However, one study found 384
that, when comparing postmenopausal breast cancer survivors in the highest tertile to the 385
lowest tertile group, pre-diagnosis total vegetable intake improved overall survival 386
(HR=0.57; 95% CI 0.35-0.94) - no association was found for total fruit intake and mortality 387
in this cohort of breast cancer survivors[58]. In addition, Dal Maso et al. found an association 388
with total fruit and vegetable consumption and overall mortality (HR=1.27; 95% CI 1.00-389
1.61) when comparing survivors of the lowest intake group to the highest intake group[59]. 390
Results from the After Breast Cancer Pooling Project, combining data from 4 cohort studies, 391
indicated no association between post-diagnosis intakes of cruciferous vegetables and 392
survival amongst 11,390 breast cancer survivors[60]. Holmes et al. reported an association 393
between the highest post-diagnosis poultry consumption and mortality in women once 394
diagnosed with breast cancer (HR=0.70; 95% CI 0.50–0.97)[61]. No associations were found 395
for fish or red meat consumption and mortality in this population. Additionally, a high dairy 396
intake before diagnosis amongst female registered nurses who participated in the NHS, was 397
related to overall survival (HR=0.72; 95% CI 0.52–1.00)[61]. Kroenke et al. found that post-398
diagnosis dairy intake amongst women diagnosed with early-stage invasive breast cancer in 399
the LACE study, was associated with an increased overall mortality (HR=1.39; 95% CI 1.02-400
1.90)[62]. More specifically, high fat dairy was related to overall mortality and breast cancer-401
specific mortality in these women (respectively HR=1.64; 95% CI 1.24-2.17 and HR=1.49; 402
95% CI 1.00-2.24) whilst low-fat dairy was not[62]. Beasley et al. examined both meat and 403
dairy intake after diagnosis and found no association with survival in the Collaborative 404
Woman’s Longevity Study (CWLS)[63]. Pre-diagnosis intakes of neither bread, 405
sunflower/pumpkin seeds nor sesame/flaxseeds reduced the risk of mortality in the MARIE 406
study[64]. Finally, post-diagnosis butter/margarine/lard consumption did increase the risk of 407
Page 17 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
18
breast cancer recurrence in a follow-up study amongst 472 breast cancer survivors enrolled 408
from the Memorial Sloan-Kettering Cancer Centre (RR=1.30; 95% CI 1.03-1.64)[65]. 409
In summary, no conclusive evidence could be provided for an association between 410
most foods of the main food groups, including fruits, vegetables, meat, or dairy, and cancer 411
recurrence or mortality - evidence for each exposure and outcome was based on the results of 412
one study only or on inconsistent results. However, limited evidence appears to indicate that 413
the reduction of dietary fat after breast cancer diagnosis could increase relapse-free survival 414
amongst breast cancer survivors, adherence to the HEI-2005 score after diagnosis is 415
associated with decreased overall mortality, adherence to the AHEI diet after diagnosis is 416
associated with decreased death from other causes, and that adherence to a prudent diet after 417
diagnosis is associated with decreased death from other causes amongst breast cancer 418
survivors. Adherence to a pre-diagnosis Western diet is associated with death from other 419
causes whilst post-diagnosis adherence to a Western diet is associated with an increased risk 420
of overall mortality in breast cancer survivors. 421
422
Laryngeal cancer 423
One cohort study could be identified for the association between several foods from 424
the main food groups and mortality amongst laryngeal cancer survivors[66]. Crosignani et.al 425
examined dietary habits and survival in of 215 Italian male laryngeal cancer survivors on pre-426
diagnosis dietary habits and survival. The consumption of total vegetables (HR=0.57; 95% CI 427
0.35-0.94), beef/veal (HR=0.50; 95% CI 0.30-0.83), and bread (HR=0.54; 95% CI 0.32-0.90) 428
were all associated with a decreased risk of overall mortality when comparing the highest 429
versus the lowest intake group. No associations were found for poultry, fish, eggs, milk, 430
cheese, pasta, potatoes, citrus fruits, other fruits, butter, or olive oil. The authors speculate 431
Page 18 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
19
that the association between the highest beef/veal intakes and mortality could tentatively be 432
interpreted as an indicator of a good nutritional status of those participants[66]. 433
In summary, no conclusive evidence for an association between fruits, vegetables, 434
protein foods, grain foods, dairy, or oils and spreads, and mortality amongst laryngeal cancer 435
survivors could be provided, as evidence for each exposure and outcome was based on the 436
results of one study only. 437
438
Non-Hodgkin Lymphoma (NHL) 439
A total of 2 cohort studies could be identified for NHL survivors regarding the intake 440
of food items. One study indicated that pre-diagnosis intakes of total fruit and vegetables and 441
vegetables only (highest versus lowest intake) were associated with decreased overall 442
mortality (respectively HR=0.68; 95% CI 0.49-0.95 and HR=0.58; 95% CI 0.38-0.89) 443
amongst female NHL survivors[67]. Additionally, the highest intakes of citrus fruits and 444
green leafy vegetables compared with the lowest intakes were related to overall mortality 445
amongst survivors with NHL (respectively HR=0.73; 95% CI 0.54-0.99 and HR=0.71; 95% 446
CI 0.51-0.98). No associations were observed for total fruit intake, yellow vegetables, red 447
vegetables or bean vegetables and mortality whilst sub-analysis investigating fruit and 448
vegetables separately for each NHL subtypes did; consumption of citrus fruits improved 449
survival in diffuse large B-cell lymphoma survivors (overall mortality HR=0.40; 95% CI 450
0.22–0.72, cancer-specific mortality HR=0.36; 95% CI 0.16–0.80), and the highest 451
consumption of green leafy vegetables favoured overall mortality in follicular lymphoma 452
survivors (HR=0.27; 95% CI 0.10–0.76)[67]. Although Leo et al. found no association 453
between pre-diagnosis intakes of fruit, vegetables, meat, fish, or legumes, and mortality in 454
2,339 NHL survivors[68], dairy intake did appear to be associated with a higher overall 455
Page 19 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
20
mortality (HR=1.14; 95% CI 1.00-1.31), yet not with NHL-specific mortality (HR=1.16; 95% 456
CI 0.98-1.37)[68]. 457
In summary, no conclusive evidence for an association between intakes of fruit, 458
vegetables, protein foods, or dairy and mortality in NHL survivors could be provided, as 459
evidence for each exposure and outcome was based on the results of one study only or on 460
inconsistent results. 461
462
Prostate cancer 463
For prostate cancer 4 cohort studies could be identified. Adherence to a Western diet 464
after prostate cancer diagnosis was associated with increased overall mortality (HR=1.67; 465
95% CI 1.16–2.42) and prostate cancer-specific mortality (HR=2.53; 95% CI 1.00-6.42) 466
amongst non-metastatic prostate cancer survivors in the Physician’s Health Study (PHS)[69]. 467
The derived Western dietary patterns appeared to be driven by the consumption of processed 468
meat[69]. A prudent diet was investigated (showing overlapping characteristics with the 469
Mediterranean diet examined in the Health Professionals Follow-up Study (HPFS)); 470
adherence to a prudent diet after prostate cancer diagnosis was inversely associated with 471
overall mortality (RR=0.64; 95% CI 0.44–0.93) and appeared to be driven by the use of oil 472
and vinegar dressings[70]. The HPFS reported on a Mediterranean diet and mortality in 473
prostate cancer survivors after diagnosis[71]. Kenfield et al. demonstrated that post-diagnosis 474
adherence to a Mediterranean diet was associated with decreased overall mortality (HR=0.78; 475
95% CI 0.67-0.90); no association was observed for prostate cancer-specific mortality and 476
adherence to the Mediterranean diet[71]. A pre-diagnosis high fish consumption in men who 477
were diagnosed with prostate cancer while participating in the PHS was related to prolonged 478
survival (HR=0.52; 95% CI 0.30-0.91) according to Chavarro et al.[72]. Another study of 479
Yang et al. investigated post-diagnosis dairy intake amongst prostate cancer survivors[73]. 480
Page 20 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
21
The consumption of total dairy was associated with increased overall mortality (HR=1.76; 481
95% CI 1.21-2.55). Both high-fat and low-fat dairy consumption contributed to this adverse 482
association and overall mortality (respectively HR=1.22; 95% CI 1.08-1.38 and HR=1.17; 483
95% CI 1.05-1.29)[73]. 484
In summary, no conclusive evidence for an association between a Mediterranean diet 485
score, adherence to a prudent or Western diet, fish, or dairy, and mortality in prostate cancer 486
survivors could be provided, as evidence for each exposure and outcome was based on the 487
results of one study only. 488
489
DISCUSSION 490
This systematic review summarizes current scientific literature regarding dietary 491
patterns/indices and foods from the main food groups and health outcomes amongst different 492
groups of cancer survivors. Limited evidence appears to indicate that the reduction of dietary 493
fat after breast cancer diagnosis could increase relapse-free survival amongst breast cancer 494
survivors, adherence to the HEI-2005 score after diagnosis is associated with decreased 495
overall mortality, adherence to the AHEI diet after diagnosis is associated with decreased 496
death from other causes, and that adherence to a prudent diet after diagnosis is associated 497
with decreased death from other causes amongst breast cancer survivors. Adherence to a pre-498
diagnosis Western diet is associated with death from other causes whilst post-diagnosis 499
adherence to a Western diet is associated with an increased risk of overall mortality in breast 500
cancer survivors. Although no conclusive evidence could be provided for other survivors than 501
of breast cancer, the results of available studies investigating dietary patterns/indices and 502
food in other cancer survivors were described in detail. 503
504
Dietary patterns/indices 505
Page 21 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
22
It could be speculated that the lack of effect in the two identified RCTs investigating a 506
low-fat diet in breast cancer survivors is a consequence of the relatively short follow-up 507
period when using mortality as the primary outcome[49,50]. It did appear, however, that a 508
reduction in dietary fat intake could increase relapse-free survival amongst these 509
survivors[49]. Nevertheless, the true beneficial effect of dietary intake in this trial remains 510
uncertain since increased exercise and weight loss during the intervention may also have 511
advantaged these breast cancer survivors[49]. Adherence to a high-quality diet or a prudent 512
diet and the increase in survival could be explained by the effects of fruit and vegetables on 513
health in general. This could also clarify the increase in mortality amongst survivors with 514
adherence to a Western diet, as it is characterised by low intakes of vegetables and fruits. It 515
remains difficult, however, to disentangle the beneficial effect of fruit and vegetables from 516
other foods in the diet – it could even be speculated that not the consumption of fruit and 517
vegetables in a high-quality and prudent diet decrease mortality, but eating less amounts of 518
sugars, salt, and saturated fats, could explain the associations found with mortality and 519
relapse-free survival. 520
Besides the evidence for a potential role of a low-fat diet in breast cancer recurrence, 521
most studies showed an association with overall mortality and death from other causes; not 522
with cancer-specific mortality or cancer recurrence. Even if the exposures identified cannot 523
help these cancer outcomes, given the survivors of the investigated cancers have potential for 524
long-term survival, it is desirable for them to follow a diet that could help reduce other 525
conditions such as cardiovascular disease and increase overall life expectancy. The limited 526
number of studies indicate that additional long-term prospective studies are urgently needed 527
to improve the strength of evidence on the influence of dietary pattern/indices adherence on 528
cancer survival. 529
530
Page 22 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
23
Foods from the main food groups 531
The investigated healthy dietary patterns/indices are characterised by foods of the 532
main food groups. Epidemiological research on fruit and vegetable intake and cancer risk 533
increased rapidly over the last few decades and it has been suggested that people with high 534
intakes of fruit and vegetables, compared to those with low intakes, have a reduced risk of 535
developing cancer[74]. The wide variety of nutrients including vitamins, minerals, 536
phytochemicals and fibre in fruit and vegetables could influence epigenetic processes and 537
potentially via this way improve cancer outcomes[75,76]. However, the exact mechanisms of 538
how diet can alter genetic and epigenetic changes in cancer cells has yet to be established. 539
The majority of the identified studies found statistically non-significant results, based on a p-540
value that indicates the degree to which the data conform to the pattern predicted by the test 541
hypothesis and all the other assumptions used in the test. Nonetheless, the HRs<1 of two 542
studies investigating pre-diagnosis fruit intake overall mortality [64,77], although statically 543
non-significant results, could strengthen the evidence that adherence to a high-quality diet, 544
characterised by high intakes of fruit and vegetables, could decrease overall mortality in 545
breast cancer survivors. The consumption of fruits could, therefore, be encouraged in breast 546
cancer survivors as they are an important part of a high-quality diet to increase overall life 547
expectancy. Studies investigating the role of fruit after diagnosis in cancer survivors are 548
urgently needed. 549
550
Study strengths and limitations 551
The strengths of this systematic review are the inclusion of dietary patterns/indices 552
and whole foods, and the large total number of cancer survivors investigated. By examining 553
the whole diet, the intake of nutrients in combination is considered which provides 554
translatable real-life scenarios for clinical recommendations. 555
Page 23 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
24
The limitations of this systematic review were the inclusion of only 2 RCTs, the few 556
studies investigating post-diagnosis intake, the use of FFQs to collect dietary information 557
from participants in most studies, and the considerable heterogeneity in study design and 558
participant characteristics (tumour characteristics (stage/grade), treatment, age, time of 559
follow-up, comorbidity, differences in countries and ethnicity). Due to potential bias, data 560
from observational studies generally provide a lower strength of evidence than from RCTs, 561
even if they were well conducted. Conducting RCTs to investigate dietary intake in cancer 562
survivors with mortality as an outcome can be challenging for cancers with a relatively long 563
survival necessitating adherence to a diet in the long-term. The majority of studies included 564
in this systematic review investigated foods before cancer diagnosis, with only a few studies 565
in the post-diagnosis setting. Information on food intake after diagnosis is valuable for 566
investigating the effect of dietary changes on health outcomes amongst cancer survivors – 567
even though it is too late to amend lifestyle factors from before diagnosis, patients are more 568
receptive to advice after diagnosis. Although the use of FFQs is an inexpensive approach to 569
capture data from hundreds or thousands of individuals, it may not represent the usual foods 570
or portion sizes chosen by participants, and intake data can be compromised when multiple 571
foods are grouped with single listings. Developments in the screening, diagnosis and 572
treatment of cancers differ greatly between countries and therefore could influence survival. 573
Although most studies are adjusted for tumour stage, age and treatment, often no adjustments 574
could be made for influential lifestyle factors including BMI, physical activity and smoking. 575
It remains a challenge to disentangle the impact of diet from other lifestyle factors, and this 576
should always be taken into consideration when interpreting study results. 577
578
CONCLUSION 579
Page 24 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
25
To conclude, the reduction of dietary fat after breast cancer diagnosis could increase 580
relapse-free survival amongst breast cancer survivors, adherence to a high-quality diet may 581
protect against overall mortality and death from other causes amongst breast cancer 582
survivors, whilst adherence to a prudent diet may protect against death from other causes 583
amongst breast cancer survivors. Adherence to a Western diet before diagnosis may be 584
detrimental for breast cancer survivors concerning death from other causes. Adherence to a 585
Western diet after diagnosis may increase overall mortality amongst breast cancer survivors. 586
It is important that the results of well-conducted studies reach health care providers so that 587
survivors can be informed of dietary factors that could possibly influence their health 588
outcomes. 589
590
Authorship contribution statement 591
SJ drafted the manuscript and worked on the conception, design and interpretation of data. SJ 592
and FvO selected articles, screened titles and abstracts, assessed study quality and extracted 593
data. SJ, FvO, RB and MZ were involved in the interpretation and discussion of the results 594
and critically revised the systematic review for important intellectual content. All authors, SJ, 595
FvO, RB, AW, FJvS, KKC and MZ, approved the final version of the systematic review. SJ 596
is the guarantor. 597
598
Conflict of interest 599
None of the authors have any conflict of interest in connection with this systematic review. 600
601
Funding 602
Page 25 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
26
This research received no grant from any funding agency in the public, commercial or not-603
for-profit sectors. This systematic review has not as yet been registered. 604
605
Data sharing statement 606
No additional data available. 607
608
Page 26 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
27
LITERATURE 609
1 Schwingshackl L, Hoffmann G. Adherence to Mediterranean diet and risk of cancer: 610
an updated systematic review and meta-analysis of observational studies. Cancer Med 611
2015;4:1933–47. doi:10.1002/cam4.539 612
2 Schwingshackl L, Chaimani A, Bechthold A, et al. Food groups and risk of chronic 613
disease: A protocol for a systematic review and network meta-analysis of cohort 614
studies. Syst Rev 2016;5:1.http://www.systematicreviewsjournal.com/ 615
http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=reference&D=emed18b&NEWS=616
N&AN=611368739 617
3 Leenders M, Sluijs I, Ros MM, et al. Fruit and vegetable consumption and mortality: 618
European prospective investigation into cancer and nutrition. Am J Epidemiol 619
2013;178:590–602. doi:10.1093/aje/kwt006 620
4 Nguyen CT, Pham NM, Lee AH, et al. Prevalence of and Risk Factors for Type 2 621
Diabetes Mellitus in Vietnam: A Systematic Review. Asia Pac J Public Heal 622
2015;27:588–600. doi:10.1177/1010539515595860 623
5 Bours MJ, Beijer S, Winkels RM, et al. Dietary changes and dietary supplement use, 624
and underlying motives for these habits reported by colorectal cancer survivors of the 625
Patient Reported Outcomes Following Initial Treatment and Long-Term Evaluation of 626
Survivorship (PROFILES) registry. Br J Nutr 2015;114:286–96. 627
doi:10.1017/S0007114515001798 628
6 Humpel N, Magee C, Jones SC. The impact of a cancer diagnosis on the health 629
behaviors of cancer survivors and their family and friends. Support Care Cancer 630
2007;15:621–30. doi:10.1007/s00520-006-0207-6 631
7 Zhang FF, Liu S, John EM, et al. Diet quality of cancer survivors and noncancer 632
individuals: Results from a national survey. Cancer 2015;121:4212–21. 633
doi:10.1002/cncr.29488 634
8 D’Avanzo B, La Vecchia C, Negri E, et al. Attributable risks for bladder cancer in 635
northern Italy. Ann Epidemiol 1995;5:427–31. 636
9 Rock CL, Doyle C, Demark-Wahnefried W, et al. Nutrition and physical activity 637
guidelines for cancer survivors. CA Cancer J Clin 2012;62:242–74. 638
doi:10.3322/caac.21142 639
10 World Cancer Research Fund International / American Institute for Cancer Research. 640
Diet, Nutrition, Physical Activity, and Breast Cancer Survivors. 2014. 641
http://www.wcrf.org/sites/default/files/Breast-Cancer-Survivors-2014-Report.pdf 642
(accessed 2 Jun 2017). 643
11 Parsons JK, Pierce JP, Natarajan L, et al. NIH Public Access. 2014;6:1–13. 644
doi:10.1158/1940-6207.CAPR-13-0050.A 645
12 Schwedhelm C, Boeing H, Hoffmann G, et al. Effect of diet on mortality and cancer 646
recurrence among cancer survivors: a systematic review and meta-analysis of cohort 647
studies. Nutr Rev 2016;74:737–48. doi:10.1093/nutrit/nuw045 648
13 Cancer Research UK. Cancer survival statistics. 649
Page 27 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
28
http://www.cancerresearchuk.org/health-professional/cancer-statistics/survival 650
(accessed 10 Jun 2017). 651
14 Reedy J, Wirfält E, Flood A, et al. Comparing 3 dietary pattern methods--cluster 652
analysis, factor analysis, and index analysis--With colorectal cancer risk: The NIH-653
AARP Diet and Health Study. Am J Epidemiol 2010;171:479–87. 654
doi:10.1093/aje/kwp393 655
15 Guenther PM, Reedy J, Krebs-Smith SM. Development of the Healthy Eating Index-656
2005. J Am Diet Assoc 2008;108:1896–901. doi:10.1016/j.jada.2008.08.016 657
16 Guenther PM, Reedy J, Krebs-Smith SM, et al. Evaluation of the Healthy Eating 658
Index-2005. J Am Diet Assoc 2008;108:1854–64. doi:10.1016/j.jada.2008.08.011 659
17 Chiuve SE, Fung TT, Rimm EB, et al. Alternative dietary indices both strongly predict 660
risk of chronic disease. J Nutr 2012;142:1009–18. doi:10.3945/jn.111.157222 661
18 Inoue-Choi M, Lazovich D, Prizment AE, et al. Adherence to the World Cancer 662
Research Fund/American Institute for Cancer Research Recommendations for Cancer 663
Prevention Is Associated With Better Health-Related Quality of Life Among Elderly 664
Female Cancer Survivors. J Clin Oncol 2013;31:1758–66. 665
doi:10.1200/JCO.2012.45.4462 666
19 Catsburg C, Miller AB, Rohan TE. Adherence to cancer prevention guidelines and risk 667
of breast cancer. Int J Cancer 2014;135:2444–52. doi:10.1002/ijc.28887 668
20 Kant AK, Schatzkin A, Graubard BI, et al. A prospective study of diet quality and 669
mortality in women. JAMA 2000;283:2109–670
15.http://www.ncbi.nlm.nih.gov/pubmed/10791502 (accessed 10 Jun 2017). 671
21 Haines PS, Siega-Riz AM, Popkin BM. The Diet Quality Index revised: a 672
measurement instrument for populations. J Am Diet Assoc 1999;99:697–704. 673
doi:10.1016/S0002-8223(99)00168-6 674
22 Fung TT, Chiuve SE, McCullough ML, et al. Adherence to a DASH-Style Diet and 675
Risk of Coronary Heart Disease and Stroke in Women. Arch Intern Med 676
2008;168:713. doi:10.1001/archinte.168.7.713 677
23 Olsen A, Egeberg R, Halkjær J, et al. Healthy aspects of the Nordic diet are related to 678
lower total mortality. J Nutr 2011;141:639–44. doi:10.3945/jn.110.131375 679
24 Fung TT, Kashambwa R, Sato K, et al. Post diagnosis diet quality and colorectal 680
cancer survival in women. PLoS One 2014;9:1–13. doi:10.1371/journal.pone.0115377 681
25 Kim EHJ, Willett WC, Fung T, et al. Diet quality indices and postmenopausal breast 682
cancer survival. Nutr Cancer 2011;63:381–8. doi:10.1080/01635581.2011.535963 683
26 McCullough ML, Willett WC. Evaluating adherence to recommended diets in adults: 684
the Alternate Healthy Eating Index. Public Health Nutr 2006;9:152–685
7.http://www.ncbi.nlm.nih.gov/pubmed/16512963 (accessed 12 Aug 2017). 686
27 Sofi F, Macchi C, Abbate R, et al. Mediterranean diet and health status: an updated 687
meta-analysis and a proposal for a literature-based adherence score. Public Health 688
Nutr 2014;17:2769–82. doi:10.1017/S1368980013003169 689
Page 28 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
29
28 Schwingshackl L, Hoffmann G. Diet Quality as Assessed by the Healthy Eating Index, 690
the Alternate Healthy Eating Index, the Dietary Approaches to Stop Hypertension 691
Score, and Health Outcomes: A Systematic Review and Meta-Analysis of Cohort 692
Studies. J Acad Nutr Diet 2015;115:780–800.e5. doi:10.1016/j.jand.2014.12.009 693
29 Public Health England. The Eatwell Guide. 694
2017.https://www.gov.uk/government/publications/the-eatwell-guide (accessed 10 Jun 695
2017). 696
30 Sterne J, Julian H, Barney R. The Cochrane Collaboration’s tool for assessing risk of 697
bias Description. Handb CochraneTable 85.a 698
2014;:4.http://handbook.cochrane.org/chapter_8/table_8_5_a_the_cochrane_collaborat699
ions_tool_for_assessing.htm 700
31 Sterne JA, Hernán MA, Reeves BC, et al. The Risk Of Bias In Non-randomized 701
Studies – of Interventions (ROBINS-I) assessment tool ROBINS-I tool (Stage I): At 702
protocol stage. 2016;:1–22. doi:10.1136/bmj.i4919 703
32 Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating 704
quality of evidence and strength of recommendations. BMJ 2008;336:924–6. 705
doi:10.1136/bmj.39489.470347.AD 706
33 Moher D. PRISMA 2009 Checklist PRISMA 2009 Checklist. PLoS Med 707
2009;6:e1000097. doi:10.1371/journal.pmed1000097 708
34 Higgins JP, Savović J, Page MJ, et al. Revised Cochrane risk of bias tool for 709
randomized trials (RoB 2.0). 2016;:52. 710
35 Sterne J, Higgins J, Reeves B, et al. A tool for evaluating risk of bias in non- 711
randomized studies of interventions ( ROBINS-I ): development and applications. 712
2015. 713
36 Tang L, Zirpoli GR, Guru K, et al. Intake of cruciferous vegetables modifies bladder 714
cancer survival. Cancer Epidemiol Biomarkers Prev 2010;19:1806–11. 715
doi:10.1158/1055-9965.EPI-10-0008 716
37 Meyerhardt JA, Niedzwiecki D, Hollis D, et al. Association of Dietary Patterns With 717
Cancer Recurrence and Survival in Patients With Stage III Colon Cancer. Jama 718
2007;298:754. doi:10.1001/jama.298.7.754 719
38 Zhu Y, Wu H, Wang PP, et al. Dietary patterns and colorectal cancer recurrence and 720
survival: a cohort study. BMJ Open 2013;3:e002270. doi:10.1136/bmjopen-2012-721
002270 722
39 Fung TT, Kashambwa R, Sato K, et al. Post Diagnosis Diet Quality and Colorectal 723
Cancer Survival in Women. PLoS One 2014;9:e115377. 724
doi:10.1371/journal.pone.0115377 725
40 Pelser C, Arem H, Pfeiffer RM, et al. Prediagnostic lifestyle factors and survival after 726
colon and rectal cancer diagnosis in the National Institutes of Health (NIH)-AARP 727
Diet and Health Study. Cancer 2014;120:1540–7. doi:10.1002/cncr.28573 728
41 Ratjen I, Schafmayer C, di Giuseppe R, et al. Postdiagnostic Mediterranean and 729
Healthy Nordic Dietary Patterns Are Inversely Associated with All-Cause Mortality in 730
Page 29 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
30
Long-Term Colorectal Cancer Survivors. J Nutr 2017;147:636–44. 731
doi:10.3945/jn.116.244129 732
42 Dik VK, Murphy N, Siersema PD, et al. Prediagnostic Intake of Dairy Products and 733
Dietary Calcium and Colorectal Cancer Survival--Results from the EPIC Cohort 734
Study. Cancer Epidemiol Biomarkers Prev 2014;23:1813–23. doi:10.1158/1055-735
9965.EPI-14-0172 736
43 Romaguera D, Ward H, Wark PA, et al. Pre-diagnostic concordance with the 737
WCRF/AICR guidelines and survival in European colorectal cancer patients: a cohort 738
study. BMC Med 2015;13:107. doi:10.1186/s12916-015-0332-5 739
44 Ward HA, Norat T, Overvad K, et al. Pre-diagnostic meat and fibre intakes in relation 740
to colorectal cancer survival in the European Prospective Investigation into Cancer and 741
Nutrition. Br J Nutr 2016;116:316–25. doi:10.1017/S0007114516001859 742
45 Yang B, McCullough ML, Gapstur SM, et al. Calcium, vitamin D, dairy products, and 743
mortality among colorectal cancer survivors: The cancer prevention study-II nutrition 744
cohort. J Clin Oncol 2014;32:2335–43. doi:10.1200/JCO.2014.55.3024 745
46 Skeie G, Braaten T, Olsen A, et al. Whole grain intake and survival among 746
Scandinavian colorectal cancer patients. Nutr Cancer 2014;66:6–13. 747
doi:10.1080/01635581.2014.847472 748
47 Carr PR, Jansen L, Walter V, et al. Associations of red and processed meat with 749
survival after colorectal cancer and differences according to timing of dietary 750
assessment. Am J Clin Nutr 2016;103:192–200. doi:10.3945/ajcn.115.121145 751
48 McCullough ML, Gapstur SM, Shah R, et al. Association between red and processed 752
meat intake and mortality among colorectal cancer survivors. J Clin Oncol 753
2013;31:2773–82. doi:10.1200/JCO.2013.49.1126 754
49 Chlebowski RT, Blackburn GL, Thomson CA, et al. Dietary fat reduction and breast 755
cancer outcome: interim efficacy results from the Women’s Intervention Nutrition 756
Study. J Natl Cancer Inst 2006;98:1767–76. doi:10.1093/jnci/djj494 757
50 Pierce JP, Natarajan L, Caan BJ, et al. Influence of a Diet Very High in Vegetables, 758
Fruit, and Fiber and Low in Fat on Prognosis Following Treatment for Breast Cancer. 759
Jama 2007;298:289. doi:10.1001/jama.298.3.289 760
51 George SM, Irwin ML, Smith AW, et al. Postdiagnosis diet quality, the combination of 761
diet quality and recreational physical activity, and prognosis after early-stage breast 762
cancer. Cancer Causes Control 2011;22:589–98. doi:10.1007/s10552-011-9732-9 763
52 George SM, Ballard-Barbash R, Shikany JM, et al. Better postdiagnosis diet quality is 764
associated with reduced risk of death among postmenopausal women with invasive 765
breast cancer in the women’s health initiative. Cancer Epidemiol Biomarkers Prev 766
2014;23:575–83. doi:10.1158/1055-9965.EPI-13-1162 767
53 Izano MA, Fung TT, Chiuve SS, et al. Are Diet Quality Scores After Breast Cancer 768
Diagnosis Associated with Improved Breast Cancer Survival? Nutr Cancer 769
2013;65:820–6. doi:10.1080/01635581.2013.804939 770
54 McCullough ML, Gapstur SM, Shah R, et al. Pre- and postdiagnostic diet in relation to 771
Page 30 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
31
mortality among breast cancer survivors in the CPS-II Nutrition Cohort. Cancer 772
Causes Control 2016;27:1303–14. doi:10.1007/s10552-016-0802-x 773
55 Kroenke CH, Fung TT, Hu FB, et al. Dietary patterns and survival after breast cancer 774
diagnosis. J Clin Oncol 2005;23:9295–303. doi:10.1200/JCO.2005.02.0198 775
56 Kwan ML, Weltzien E, Kushi LH, et al. Dietary patterns and breast cancer recurrence 776
and survival among women with early-stage breast cancer. J Clin Oncol 2009;27:919–777
26. doi:10.1200/JCO.2008.19.4035 778
57 Vrieling A, Buck K, Seibold P, et al. Dietary patterns and survival in German 779
postmenopausal breast cancer survivors. Br J Cancer 2013;108:188–92. 780
doi:10.1038/bjc.2012.521 781
58 McEligot AJ, Largent J, Ziogas A, et al. Dietary fat, fiber, vegetable, and 782
micronutrients are associated with overall survival in postmenopausal women 783
diagnosed with breast cancer. Nutr Cancer 2006;55:132–40. 784
doi:10.1207/s15327914nc5502_3 785
59 Dal Maso L, Zucchetto A, Talamini R, et al. Effect of obesity and other lifestyle 786
factors on mortality in women with breast cancer. Int J cancer 2008;123:2188–94. 787
doi:10.1002/ijc.23747 788
60 Nechuta S, Caan BJ, Chen WY, et al. Postdiagnosis Cruciferous Vegetable 789
Consumption and Breast Cancer Outcomes: A Report from the After Breast Cancer 790
Pooling Project. Cancer Epidemiol Biomarkers Prev 2013;22:1451–6. 791
doi:10.1158/1055-9965.EPI-13-0446 792
61 Holmes MD, Stampfer MJ, Colditz GA, et al. Dietary factors and the survival of 793
women with breast carcinoma. Cancer 1999;86:826–35. doi:10.1002/(SICI)1097-794
0142(19990901)86:5<826::AID-CNCR19>3.0.CO;2-0 795
62 Kroenke CH, Kwan ML, Sweeney C, et al. High-and low-fat dairy intake, recurrence, 796
and mortality after breast cancer diagnosis. J Natl Cancer Inst 2013;105:616–23. 797
doi:10.1093/jnci/djt027 798
63 Beasley JM, Newcomb PA, Trentham-Dietz A, et al. Post-diagnosis dietary factors 799
and survival after invasive breast cancer. Breast Cancer Res Treat 2011;128:229–36. 800
doi:10.1007/s10549-010-1323-z 801
64 Buck K, Zaineddin AK, Vrieling A, et al. Estimated enterolignans, lignan-rich foods, 802
and fibre in relation to survival after postmenopausal breast cancer. Br J Cancer 803
2011;105:1151–7. doi:10.1038/bjc.2011.374 804
65 Hebert JR, Hurley TG, Ma Y. The effect of dietary exposures on recurrence and 805
mortality in early stage breast cancer. Breast Cancer Res. Treat. 1998;51:17–28. 806
doi:10.1023/A:1006056915001 807
66 Crosignani P, Russo A, Tagliabue G, et al. Tobacco and diet as determinants of 808
survival in male laryngeal cancer patients. Int J Cancer 1996;65:308–13. 809
doi:10.1002/(SICI)1097-0215(19960126)65:3<308::AID-IJC5>3.0.CO;2-3 810
67 Han X, Zheng T, Foss F, et al. Vegetable and fruit intake and non-Hodgkin lymphoma 811
survival in Connecticut women. Leuk Lymphoma 2010;51:1047–54. 812
Page 31 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
32
doi:10.3109/10428191003690364 813
68 Leo QJN, Ollberding NJ, Wilkens LR, et al. Nutritional factors and non-Hodgkin 814
lymphoma survival in an ethnically diverse population: the Multiethnic Cohort. Eur J 815
Clin Nutr 2016;70:41–6. doi:10.1038/ejcn.2015.139 816
69 Yang M, Kenfield SA, Van Blarigan EL, et al. Dietary patterns after prostate cancer 817
diagnosis in relation to disease-specific and total mortality. Cancer Prev Res (Phila) 818
2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 819
70 Yang M, Kenfield SA, Van Blarigan EL, et al. Dietary Patterns after Prostate Cancer 820
Diagnosis in Relation to Disease-Specific and Total Mortality. Cancer Prev Res 821
2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 822
71 Kenfield SA, DuPre N, Richman EL, et al. Mediterranean diet and prostate cancer risk 823
and mortality in the Health Professionals Follow-up Study. Eur Urol 2014;65:887–94. 824
doi:10.1016/j.eururo.2013.08.009 825
72 Chavarro JE, Stampfer MJ, Hall MN, et al. A 22-y prospective study of fish intake in 826
relation to prostate cancer incidence and mortality. Am J Clin Nutr 2008;88:1297–827
303.http://www.ncbi.nlm.nih.gov/pubmed/18996866 (accessed 10 Jun 2017). 828
73 Yang M, Kenfield SA, Van Blarigan EL, et al. Dairy intake after prostate cancer 829
diagnosis in relation to disease-specific and total mortality. Int J Cancer 830
2015;137:2462–9. doi:10.1002/ijc.29608 831
74 Wiseman M. The second World Cancer Research Fund/American Institute for Cancer 832
Research expert report. Food, nutrition, physical activity, and the prevention of cancer: 833
a global perspective. Proc Nutr Soc 2008;67:253–6. doi:10.1017/S002966510800712X 834
75 Hardy TM, Tollefsbol TO. Epigenetic diet: impact on the epigenome and cancer. 835
Epigenomics 2011;3:503–18. doi:10.2217/epi.11.71 836
76 Daniel M, Tollefsbol TO. Epigenetic linkage of aging, cancer and nutrition. J Exp Biol 837
2015;218:59–70. doi:10.1242/jeb.107110 838
77 Dal Maso L, Zucchetto A, Talamini R, et al. Effect of obesity and other lifestyle 839
factors on mortality in women with breast cancer. Int J Cancer 2008;123:2188–94. 840
doi:10.1002/ijc.23747 841
Page 32 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
33
842 Table 1: literature search for the Pubmed database addressing the relationship between diet and mortality among bladder cancer survivors
(“urinary bladder neoplasms”[Mesh] OR “urinary bladder neoplasm*” OR “bladder cancer*” OR “bladder tumor*” OR “bladder tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang]
843
844
Page 33 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
34
Table 2: Number of studies investigating the association between pre-diagnosis dietary patterns/indices and mortality/cancer recurrence in different populations of cancer survivors
Diet quality indices Prudent / healthy diet Western diet / unhealthy diet
Cancer site /type No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 0 - - - - 0 - - - - 0 - - - -
Bowel 2 - 2 1 - 1 - 1 1 - 1 - 2 2 -
Breast 1 - 1 1 1 2 1 4 4 4 2 1 4 4 4
Cervix 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - -
Larynx 0 - - - - 0 - - - - 0 - - - -
MM 0 - - - - 0 - - - - 0 - - - -
NHL 0 - - - - 0 - - - - 0 - - - -
Prostate 0 - - - - 0 - - - - 0 - - - -
Testes 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - -
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and several dietary patterns in the same
population
Page 34 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
35
Table 3: Number of studies investigating the association between post-diagnosis dietary patterns/indices and mortality/cancer recurrence in different populations of cancer survivors
Diet quality indices Prudent diet / healthy diet Western diet / unhealthy
Cancer site /type No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 0 - - - - 0 - - - - 0 - - -
Bowel 2 - 5 3 - 2 1 2 1 - 2 1 2 1 -
Breast 7 1 11 9 8 2 1 2 2 2 2 1 2 2 2
Cervix 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - -
Larynx 0 - - - - 0 - - - - 0 - - - -
MM 0 - - - - 0 - - - - 0 - - - -
NHL 0 - - - - 0 - - - - 0 - - - -
Prostate 1 - 1 1 - 1 - 1 1 - 1 - 1 1 -
Testes 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - -
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and several dietary patterns in the same
population
Page 35 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
36
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and several food items in the same population
Table 4: Number of studies investigating the association between pre-diagnosis foods and mortality/cancer recurrence in different populations of cancer survivors
Fruit and vegetables Grain foods Protein foods Dairy and alternatives Oils and spreads
Cancer site
/type
No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 1 - 4 4 - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Bowel 0 - - - - 1 - 5 - - 3 1 6 6 2 2 - 6 6 - 0 - - - -
Breast 5 - 7 5 1 1 - 1 1 - 3 - 6 3 1 1 - 1 1 - 0 - - - -
Cervix 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Larynx 1 - 3 - - 1 - 3 - - 1 - 4 - - 1 - 2 - - 1 - 2 - -
MM 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
NHL 2 - 11 11 - 0 - - - - 1 - 3 3 - 1 - 1 1 - 0 - - - -
Prostate 0 - - - - 0 - - - - 1 - - 1 - 0 - - - - 0 - - - -
Testes 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Page 36 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
37
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and several food items in the same population
Table 5: Number of studies investigating the association between post-diagnosis foods and mortality/cancer recurrence in different populations of cancer survivors
Fruit and vegetables Grain foods Protein foods Dairy and alternatives Oils and spreads
Cancer site
/type
No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Bowel 0 - - - - 0 - - - - 1 - 2 2 2 1 - 2 2 - 0 - - - -
Breast 4 1 6 4 1 0 - - - - 4 1 5 6 1 3 3 5 2 - 1 1 - 1 -
Cervix 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Larynx 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
MM 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
NHL 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Prostate 0 - - - - 0 - - - - 0 - - - - 1 - 3 - - 0 - - - -
Testes 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Page 37 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Online supporting data File S1 for manuscript ‘the impact of dietary patterns and the
main food groups on mortality and recurrence in cancer survivors: systematic review
of current epidemiological literature’
Jochems et al., 31-05-2017
Page 38 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
2
Protocol Systematic Literature Review
INTRODUCTION Title The impact of diet on mortality in cancer survivors: A systematic review of current epidemiological literature Authors Sylvia H.J. Jochems, Frits H.M. van Osch, Richard T. Bryan, Anke Wesselius, Frederik J. van Schooten, K.K. Cheng, Maurice P. Zeegers Article type Systematic literature review (SLR) Language article English
Actual start date 01 December 2014 Updated 07 Mei 2017
- inclusion one additional exclusion criteria: sample size had to be > 200 survivors - exclusion of beverages for more clear focus of systematic review - inclusion of dietary indices and grain products as an exposure - inclusion of cancer recurrence as an outcome (these adjustments had influence on the outcome, exposure and search terms of the systematic review and were therefore adjusted accordingly)
Funding sources No Conflicts of interest No
Research question The aim of this study was to conduct a structured summary and evaluation of randomised controlled trials and observational studies addressing the relationship between the highest versus the lowest intake of dietary patterns/indices and foods of the main food groups and mortality and cancer recurrence amongst groups of survivors of common cancers with a ten-year survival rate of at least 50%. PICO model Population: cancer survivors Intervention/exposure: dietary patterns/indices and foods from the main food groups Comparator/control: highest versus lowest intake Outcome: mortality (overall, cancer-specific, and death from other causes) and cancer recurrence
Page 39 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
3
METHODS Database search
Pubmed / Medline (1966 - May 2017)
Embase (1980 – May 2017)
Cochrane Library (1993 – May 2017) Additional search Reference tracking of included and related articles, systematic reviews and meta-analyses Study types
Inclusion criteria
Randomized Controlled Trials
Cross-sectional studies
Cohort studies - retrospective & longitudinal
Case-control (including follow-up of cases)
Exclusion criteria
Animal studies
In vitro studies
Gene-nutrient interaction studies Population / participants
Inclusion criteria
Adult population, at least 18 years of age (both men and women)
Survivors of common cancers with a ten-year survival of at least 50% including bladder, bowel, breast, cervical, kidney, laryngeal, prostate, testicular, uterine cancer, malignant melanoma, and (non-)Hodgkin lymphoma
Exclusion criteria
Pre-cancerous conditions of other cancer types
Combination of different types of cancers Study characteristics
Exclusion criteria
Sample size of at least 200 survivors in the analysis [http://www.tandfonline.com/toc/hsem20/current]
Follow-up period of at least 4 years (the risk of cancer recurrence is the greatest within the first three years for most included cancers) [http://www.cancerresearchuk.org/about-cancer/what-is-cancer/why-some-cancers-come-back]
Adjustments had to be made for at least age, tumour characteristics (stage/grade), and preferably initial treatment, in the statistical analysis
Page 40 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4
Exposure / intervention Dietary patterns/indices and foods from the main food groups including:
(I) Dietary patterns that were considered were assessed by index-based methods and data-driven approaches, such as principal component analysis (factor analysis) and cluster analysis
(II) fruit and vegetables including citrus fruits, stone fruits, soft fruits, fleshy fruits, vine fruits, flower vegetables, leafy vegetables, stem vegetables, fruit vegetables, mushrooms, bulbs and roots;
(III) grain foods including potatoes, bread, rice, pasta and cereal; (IV) protein foods including meat (processed meat, unprocessed meat, red meat,
poultry), fish, eggs, tofu, nuts, seeds, pulses, legumes and beans; (V) dairy and alternative products including yoghurt, milk, cheese; (VI) oils and spreads including vegetable oils, spreads
Comparators / control Highest compared to lowest intake category of dietary patterns/indices and foods from the main food groups described under exposure Outcomes Primary outcomes Overall mortality, cancer-specific mortality, death from other causes, and cancer recurrence Secondary outcomes None
Analysis (consideration of a meta-analysis besides a systematic review) We expect diversity in pre- and post-diagnosis dietary patterns/indice and foods, as well as the different cancers, and therefore decided to only consider comparable studies (same timeframe pre- or post-diagnosis and same cancer type) for meta-analysis. If more than 75% of the review will not have 3 or more studies that can be pooled under these conditions, only a systematic review and no meta-analysis will be conducted.
Article selection Inclusion criteria
Investigate the associations between dietary patterns/indices and foods, and mortality and cancer recurrence in survivors of primary cancer
Report a measure of the effect/association of the exposure on the outcomes
The duration of the exposure/intervention had to be recorded as well as the time between exposure assessment / intervention and outcome assessment
Diet and lifestyle modifications/changes consequent on the disease or its treatment will not be included
Present results of primary and secondary analysis
Present results for any of the following outcomes: • Overall mortality • Cancer-specific mortality • Death from other causes • Cancer recurrence
Page 41 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
5
Original articles published in peer-reviewed journals
Articles written in the English language Exclusion criteria
Systematic reviews, meta-analysis Reviews, comments, letters, conference abstracts
Data extraction The data extraction will be performed independently by two of the authors (Sylvia Jochems and Frits van Osch) and any disagreements about study inclusion will be resolved through consensus or, if necessary, a third party (Rik Bryan). Information to extract from studies: Author, Study, Country, Number of participants, sex, age, follow-up period, exposure, exposure timeframe, exposure assessment, outcome, results (HR/RR and 95% CI), adjustments in the statistical analysis. Statistical significance tests used in individual studies The authors of this review would like to note that results from the individual studies described as ‘not associated with mortality or cancer recurrence’, are mostly based on statistical significance tests with the focus on traditional definitions of p-values and statistical significance on null hypotheses. Notwithstanding, it is our believe that a correct interpretation of statistical tests demands critical examining the sizes of effect estimates and confidence limits, p-values, and the assumptions and conventions used for the statistical analyses.
Tools for assessing Risk of Bias and Level of Quality The Cochrane Collaboration risk of bias assessment tools were used for appraisal of RCTs and cohort studies. For RCTs the RoB 2.0 tool (a revised tool for risk of bias in randomized trials) will be used to evaluate the risk of bias. Cohort studies will be evaluated with an adjusted version of the ROBINS-I tool [http://methods.cochrane.org/bias/risk-bias-non-randomized-studies-interventions]. Levels of quality were determined with the GRADE system [http://handbook.cochrane.org/chapter_12/12_2_1_the_grade_approach.htm].
General search terms in Pubmed 1. Searching for all studies relating to cancer and survival:
Neoplasms, neoplasm staging, neoplasm recurrence local, neoplasia, tumours, cancer, survivors, survival analysis, recurrence, mortality, survival rate, disease management
2. Searching for all studies relating to dietary modification: Foods from the main food groups, dietary indices, dietary patterns
3. Selecting randomised control trials: Randomized controlled trial, random allocation, double blind method, single-blind method, clinical trial
4. Selecting cohort studies: epidemiologic studies, cohort studies, follow-up studies, longitudinal studies, prospective studies, retrospective studies
5. Additional filters: English language, human, full text These search terms will be adapted for use in the Ovid database (EMBASE and the Cochrane library).
Page 42 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
6
Specific search terms per cancer site in Pubmed The search strategy will be adapted for use in EMBASE and the Cochrane Library
1. Searching for all studies relating to bladder cancer: (“urinary bladder neoplasms”[Mesh] OR “urinary bladder neoplasm*” OR “bladder cancer*” OR “bladder tumor*” OR “bladder tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 43 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7
2. Searching for all studies relating to breast cancer: (“breast neoplasms/diet therapy”[Mesh] OR “breast neoplasms/mortality”[Mesh] OR “breast neoplasms/prevention and control”[Mesh] OR “mammary neoplasm*” OR “breast neoplasm*” OR “mammary cancer*” OR “breast cancer*” OR “breast carcinoma*” OR “human mammary carcinoma*” OR “breast tumor*” OR “breast tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 44 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
8
3. Searching for all studies relating to cervical cancer: (“uterine cervical neoplasms/diet therapy”[Mesh] OR “uterine cervical neoplasms/mortality”[Mesh] OR “uterine cervical neoplasms/prevention and control”[Mesh] OR “uterine cervical neoplasm*” OR “cervical cancer*” OR “cervical neoplasm*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 45 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
9
4. Searching for all studies relating to bowel cancer:
(“colorectal neoplasms/diet therapy”[Mesh] OR “colorectal neoplasms/mortality”[Mesh] OR “colorectal neoplasms/prevention and control”[Mesh] OR “colorectal neoplasm*” OR “colorectal cancer*” OR “colorectal tumor*” OR “colorectal tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 46 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
10
5. Searching for all studies relating to Hodgkin lymphoma: (“Hodgkin disease/diet therapy”[Mesh] OR “Hodgkin disease/mortality”[Mesh] OR “Hodgkin disease/prevention and control”[Mesh] OR “Hodgkin disease” OR “Hodgkin Lymphoma” OR “Malignant Lymphogranuloma*” OR “Hodgkin lymphoma” OR “nodular lymphocyte predominant Hodgkin's lymphoma” OR “nodular sclerosing Hodgkin's lymphoma” OR “lymphocyte rich classical Hodgkin's lymphoma” OR “mixed cellularity Hodgkin's lymphoma”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 47 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
11
6. Searching for all studies relating to non-Hodgkin lymphoma:
(“lymphoma, non-Hodgkin/diet therapy”[Mesh] OR “lymphoma, non-Hodgkin/mortality”[Mesh] OR “lymphoma, non-Hodgkin/prevention and control”[Mesh] OR “lymphoma, non-Hodgkin” OR “nonhodgkins lymphoma” OR “non-Hodgkins lymphoma”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 48 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
12
7. Searching for all studies relating to kidney cancer: (“Kidney Neoplasms/diet therapy”[Mesh] OR “Kidney Neoplasms/mortality”[Mesh] OR “Kidney Neoplasms/prevention and control”[Mesh] OR “Kidney Neoplasm*” OR “Renal Neoplasm*” OR “Kidney Cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 49 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
13
8. Searching for all studies relating to larynx cancer:
(“laryngeal neoplasms/diet therapy”[Mesh] OR “laryngeal neoplasms/mortality”[Mesh] OR “laryngeal neoplasms/prevention and control”[Mesh] OR “laryngeal neoplasm*” OR “larynx cancer*” OR “larynx neoplasm*” OR “laryngeal cancer”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 50 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
14
9. Searching for all studies relating to multiple myeloma:
(“multiple myeloma/diet therapy”[Mesh] OR “multiple myeloma/mortality”[Mesh] OR “multiple myeloma/prevention and control”[Mesh] OR “multiple myeloma*” OR “plasma cell myeloma*” OR “myelomatosis” OR “myelomatoses” OR “Kahler Disease” OR “myeloma multiple”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 51 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
15
10. Searching for all studies relating to malignant melanoma: ("Melanoma/ diet therapy”[Mesh] OR “Melanoma/mortality”[Mesh] OR “Melanoma/prevention and control”[Mesh] OR “Malignant Melanoma*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 52 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
16
11. Searching for all studies relating to prostate cancer: (“prostatic neoplasms/diet therapy”[Mesh] OR “prostatic neoplasms/mortality”[Mesh] OR “prostatic neoplasms/prevention and control”[Mesh] OR “prostate neoplasm*” OR “prostatic neoplasm*” OR “prostate cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 53 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
17
12. Searching for all studies relating to testicular cancer:
(“testicular neoplasms/diet therapy”[Mesh] OR “testicular neoplasms/mortality”[Mesh] OR “testicular neoplasms/prevention and control”[Mesh] OR “testicular neoplasm*” OR “testicular tumor*” OR “testicular tumour*” OR “testis neoplasm*” OR “testis cancer*” OR “testicular cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 54 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
18
13. Searching for all studies relating to uterus cancer: (“uterine neoplasms/diet therapy”[Mesh] OR “uterine neoplasms/mortality”[Mesh] OR “uterine neoplasms/prevention and control”[Mesh] OR “uterus neoplasm*” OR “uterine neoplasm*” OR “uterus cancer*” OR “uterine cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 55 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
19
The Risk Of Bias In Non-randomized Studies – of Interventions (ROBINS-I) assessment tool (version for cohort-type studies) Version 19 September 2016
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License.
ROBINS-I tool (Stage I): At protocol stage
Specify the review question
Participants
Experimental
intervention
Comparator
Outcomes
List the confounding domains relevant to all or most studies
List co-interventions that could be different between intervention groups and that could
impact on outcomes
Page 56 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
20
ROBINS-I tool (Stage II): For each study
Specify a target randomized trial specific to the study
Design Individually randomized / Cluster randomized / Matched (e.g. cross-
over)
Participants
Experimental
intervention
Comparator
Is your aim for this study…? to assess the effect of assignment to intervention to assess the effect of starting and adhering to intervention
Specify the outcome
Specify which outcome is being assessed for risk of bias (typically from among those earmarked for
the Summary of Findings table). Specify whether this is a proposed benefit or harm of intervention.
Specify the numerical result being assessed
In case of multiple alternative analyses being presented, specify the numeric result (e.g. RR = 1.52
(95% CI 0.83 to 2.77) and/or a reference (e.g. to a table, figure or paragraph) that uniquely defines
the result being assessed.
Page 57 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
21
Preliminary consideration of confounders
Complete a row for each important confounding domain (i) listed in the review protocol; and (ii)
relevant to the setting of this particular study, or which the study authors identified as potentially
important. “Important” confounding domains are those for which, in the context of this study, adjustment is expected to lead to a clinically important change in the estimated effect of the intervention. “Validity” refers to whether the confounding variable or variables fully measure the domain, while “reliability” refers to the precision of the measurement (more measurement error means less reliability).
(i) Confounding domains listed in the review protocol
Confounding
domain
Measured
variable(s)
Is there evidence
that controlling for
this variable was
unnecessary?*
Is the confounding
domain measured
validly and
reliably by this
variable (or these
variables)?
OPTIONAL: Is
failure to adjust for
this variable
(alone) expected to
favour the
experimental
intervention or the
comparator?
Yes / No / No
information
Favour
experimental /
Favour comparator
/ No information
(ii) Additional confounding domains relevant to the setting of this particular study, or which
the study authors identified as important
Confounding
domain
Measured
variable(s)
Is there evidence
that controlling for
this variable was
unnecessary?*
Is the confounding
domain measured
validly and
reliably by this
variable (or these
variables)?
OPTIONAL: Is
failure to adjust for
this variable
(alone) expected to
favour the
experimental
intervention or the
Page 58 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
22
comparator?
Yes / No / No
information
Favour
experimental /
Favour comparator
/ No information
* In the context of a particular study, variables can be demonstrated not to be confounders and so not included in the analysis: (a) if they are not predictive of the outcome; (b) if they are not predictive of intervention; or (c) because adjustment makes no or minimal difference to the estimated
effect of the primary parameter. Note that “no statistically significant association” is not the same as “not predictive”.
Page 59 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
23
Preliminary consideration of co-interventions
Complete a row for each important co-intervention (i) listed in the review protocol; and (ii) relevant
to the setting of this particular study, or which the study authors identified as important. “Important” co-interventions are those for which, in the context of this study, adjustment is expected to lead to a clinically important change in the estimated effect of the intervention.
(i) Co-interventions listed in the review protocol
Co-intervention Is there evidence that controlling
for this co-intervention was
unnecessary (e.g. because it was
not administered)?
Is presence of this co-
intervention likely to favour
outcomes in the experimental
intervention or the comparator
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
(ii) Additional co-interventions relevant to the setting of this particular study, or which the
study authors identified as important
Co-intervention Is there evidence that controlling
for this co-intervention was
unnecessary (e.g. because it was
not administered)?
Is presence of this co-
intervention likely to favour
outcomes in the experimental
intervention or the comparator
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Page 60 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
24
Risk of bias assessment
Responses underlined in green are potential markers for low risk of bias, and responses in red are
potential markers for a risk of bias. Where questions relate only to sign posts to other questions, no
formatting is used.
Signalling questions Description Response
options
Bias due to confounding
1.1 Is there potential for confounding of the effect of
intervention in this study?
If N/PN to 1.1: the study can be considered to be at low risk of
bias due to confounding and no further signalling questions need
be considered
Y / PY / PN / N
If Y/PY to 1.1: determine whether there is a need to assess time-
varying confounding:
1.2. Was the analysis based on splitting participants’ follow
up time according to intervention received?
If N/PN, answer questions relating to baseline
confounding (1.4 to 1.6)
If Y/PY, go to question 1.3.
NA / Y / PY /
PN / N / NI
1.3. Were intervention discontinuations or switches likely to
be related to factors that are prognostic for the outcome?
If N/PN, answer questions relating to baseline
confounding (1.4 to 1.6)
If Y/PY, answer questions relating to both baseline and
time-varying confounding (1.7 and 1.8)
NA / Y / PY /
PN / N / NI
Questions relating to baseline confounding only
1.4. Did the authors use an appropriate analysis method that
controlled for all the important confounding domains?
NA / Y / PY / PN / N / NI
1.5. If Y/PY to 1.4: Were confounding domains that were
controlled for measured validly and reliably by the
variables available in this study?
NA / Y / PY / PN / N / NI
1.6. Did the authors control for any post-intervention
variables that could have been affected by the intervention?
NA / Y / PY / PN / N / NI
Questions relating to baseline and time-varying confounding
1.7. Did the authors use an appropriate analysis method that
controlled for all the important confounding domains and
for time-varying confounding?
NA / Y / PY / PN / N / NI
1.8. If Y/PY to 1.7: Were confounding domains that were
controlled for measured validly and reliably by the
variables available in this study?
NA / Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious /
Critical / NI
Optional: What is the predicted direction of bias due to
confounding?
Favours experimental /
Favours comparator /
Unpredictable
Page 61 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
25
Bias in selection of participants into the study
2.1. Was selection of participants into the study (or into
the analysis) based on participant characteristics
observed after the start of intervention?
If N/PN to 2.1: go to 2.4
Y / PY / PN / N / NI
2.2. If Y/PY to 2.1: Were the post-intervention
variables that influenced selection likely to be
associated with intervention?
2.3 If Y/PY to 2.2: Were the post-intervention
variables that influenced selection likely to be
influenced by the outcome or a cause of the
outcome?
NA / Y / PY / PN / N / NI
NA / Y / PY / PN / N / NI
2.4. Do start of follow-up and start of intervention
coincide for most participants?
Y / PY / PN / N / NI
2.5. If Y/PY to 2.2 and 2.3, or N/PN to 2.4: Were
adjustment techniques used that are likely to correct for
the presence of selection biases?
NA / Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical /
NI
Optional: What is the predicted direction of bias due to
selection of participants into the study?
Favours experimental / Favours
comparator / Towards null /Away
from null / Unpredictable
Bias in classification of interventions
3.1 Were intervention groups clearly defined? Y / PY / PN / N / NI
3.2 Was the information used to define
intervention groups recorded at the start of the
intervention?
Y / PY / PN / N / NI
3.3 Could classification of intervention status have
been affected by knowledge of the outcome or risk
of the outcome?
Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of bias
due to classification of interventions?
Favours experimental / Favours
comparator / Towards null /Away from
null / Unpredictable
Page 62 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
26
Bias due to deviations from intended interventions
If your aim for this study is to assess the effect of assignment to
intervention, answer questions 4.1 and 4.2
4.1. Were there deviations from the intended intervention beyond what would
be expected in usual practice?
Y / PY / PN /
N / NI
4.2. If Y/PY to 4.1: Were these deviations from intended intervention
unbalanced between groups and likely to have affected the outcome?
NA / Y / PY /
PN / N / NI
If your aim for this study is to assess the effect of starting and adhering to
intervention, answer questions 4.3 to 4.6
4.3. Were important co-interventions balanced across intervention groups? Y / PY / PN /
N / NI
4.4. Was the intervention implemented successfully for most participants? Y / PY / PN /
N / NI
4.5. Did study participants adhere to the assigned intervention regimen? Y / PY / PN /
N / NI
4.6. If N/PN to 4.3, 4.4 or 4.5: Was an appropriate analysis used to estimate
the effect of starting and adhering to the intervention?
NA / Y / PY /
PN / N / NI
Risk of bias judgement
Optional: What is the predicted direction of bias due to deviations from the
intended interventions?
Bias due to missing data
5.1 Were outcome data available for all, or nearly all,
participants?
Y / PY / PN / N / NI
5.2 Were participants excluded due to missing data
on intervention status?
Y / PY / PN / N / NI
5.3 Were participants excluded due to missing data
on other variables needed for the analysis?
Y / PY / PN / N / NI
5.4 If PN/N to 5.1, or Y/PY to 5.2 or 5.3: Are the
proportion of participants and reasons for missing
data similar across interventions?
NA / Y / PY / PN / N / NI
5.5 If PN/N to 5.1, or Y/PY to 5.2 or 5.3: Is there
evidence that results were robust to the presence of
missing data?
NA / Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of bias due
to missing data?
Favours experimental / Favours
comparator / Towards null /Away from
null / Unpredictable
Page 63 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
27
Bias in measurement of outcomes
6.1 Could the outcome measure have been
influenced by knowledge of the intervention
received?
Y / PY / PN / N / NI
6.2 Were outcome assessors aware of the
intervention received by study participants?
Y / PY / PN / N / NI
6.3 Were the methods of outcome assessment
comparable across intervention groups?
Y / PY / PN / N / NI
6.4 Were any systematic errors in measurement
of the outcome related to intervention
received?
Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of
bias due to measurement of outcomes?
Favours experimental / Favours comparator /
Towards null /Away from null /
Unpredictable
Bias in selection of the reported result
Is the reported effect estimate likely to be
selected, on the basis of the results, from...
7.1. ... multiple outcome measurements
within the outcome domain?
Y / PY / PN / N / NI
7.2 ... multiple analyses of the intervention-
outcome relationship?
Y / PY / PN / N / NI
7.3 ... different subgroups? Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of
bias due to selection of the reported result?
Favours experimental / Favours comparator /
Towards null /Away from null / Unpredictable
Overall bias
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the overall predicted
direction of bias for this outcome?
Favours experimental / Favours comparator /
Towards null /Away from null / Unpredictable
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License.
Page 64 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
28
The RoB 2.0 tool (individually randomized, parallel group trials)
Assessor name/initials
Study ID and/or reference(s)
Study design
Randomized parallel group trial Cluster-randomized trial Randomized cross-over or other matched design
Specify which outcome is being assessed for risk
of bias
Specify the numerical result being assessed. In
case of multiple alternative analyses being presented,
specify the numeric result (e.g. RR = 1.52 (95% CI
0.83 to 2.77) and/or a reference (e.g. to a table, figure
or paragraph) that uniquely defines the result being
assessed.
Is your aim for this study…? to assess the effect of assignment to intervention to assess the effect of starting and adhering to intervention
Which of the following sources have you obtained to help inform your risk of bias judgements
(tick as many as apply)?
Journal article(s) with results of the trial
Trial protocol
Statistical analysis plan (SAP)
Non-commercial trial registry record (e.g. ClinicalTrials.gov record)
Company-owned trial registry record (e.g. GSK Clinical Study Register record)
“Grey literature” (e.g. unpublished thesis)
Conference abstract(s) about the trial
Regulatory document (e.g. Clinical Study Report, Drug Approval Package)
Research ethics application
Grant database summary (e.g. NIH RePORTER, Research Councils UK Gateway to
Research)
Personal communication with trialist
Personal communication with the sponsor
Page 65 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
29
Page 66 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
30
Domain Signalling questions Response options
Description/Support
for judgement
Bias arising from
the
randomization
process
1.1 Was the allocation
sequence random?
Y / PY / PN / N / NI
1.2 Was the allocation
sequence concealed until
participants were
recruited and assigned to
interventions?
Y / PY / PN / N / NI
1.3 Were there baseline
imbalances that suggest a
problem with the
randomization process?
Y / PY / PN / N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias arising from the
randomization process?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias due to
deviations from
intended
interventions
2.1. Were participants
aware of their assigned
intervention during the
trial?
Y / PY / PN / N / NI
2.2. Were carers and trial
personnel aware of
participants' assigned
intervention during the
trial?
Y / PY / PN / N / NI
2.3. If Y/PY/NI to 2.1 or
2.2: Were important co-
interventions balanced
across intervention
groups?
NA / Y / PY / PN /
N / NI
2.4. Was the intervention
implemented
successfully?
Y / PY / PN / N / NI
2.5. Did study
participants adhere to the
assigned intervention
regimen?
Y / PY / PN / N / NI
Page 67 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
31
Domain Signalling questions Response options
Description/Support
for judgement
2.6. If N/PN/NI to 2.3,
2.4 or 2.5: Was an
appropriate analysis used
to estimate the effect of
starting and adhering to
the intervention?
NA / Y / PY / PN /
N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to deviations
from intended
interventions?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias due to
missing outcome
data
3.1 Were outcome data
available for all, or
nearly all, participants
randomized?
Y / PY / PN / N / NI
3.2 If N/PN/NI to 3.1:
Are the proportions of
missing outcome data
and reasons for missing
outcome data similar
across intervention
groups?
NA / Y / PY / PN /
N / NI
3.3 If N/PN/NI to 3.1: Is
there evidence that
results were robust to the
presence of missing
outcome data?
NA / Y / PY / PN /
N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to missing
outcome data?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias in
measurement of
the outcome
4.1 Were outcome assessors aware of the intervention received by study participants?
Y / PY / PN / N / NI
Page 68 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
32
Domain Signalling questions Response options
Description/Support
for judgement
4.2 If Y/PY/NI to 4.1: Was the assessment of the outcome likely to be influenced by knowledge of intervention received?
NA / Y / PY / PN /
N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to measurement
of the outcome?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias in selection
of the reported
result
Are the reported outcome data likely to have been selected, on the basis of the results, from...
5.1. ... multiple outcome measurements (e.g. scales, definitions, time points) within the outcome domain?
Y / PY / PN / N / NI
5.2 ... multiple analyses of the data?
Y / PY / PN / N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to selection of
the reported result?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Overall bias Risk of bias judgement Low / High / Some
concerns
Optional:
What is the overall
predicted direction of
bias for this outcome?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Page 69 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
33
RESULT SEARCH: PRISMA FLOWCHART
PRISMA Flow Diagram – dietary patterns/indices and foods from
the main food groups and mortality/cancer recurrence amongst
different groups of cancer survivors
Records identified through database searching: Bladder cancer n= 317 Breast cancer n= 2044 Cervical cancer n= 123 Colorectal cancer n= 1157 (Non-)Hodgkin lymphoma n= 245 Kidney cancer n= 134 Larynx cancer n= 94 Malignant melanoma n= 103 Multiple myeloma n= 161 Prostate cancer n= 902 Testicular cancer n= 35 Uterus cancer n= 145
Scre
enin
g In
clu
ded
El
igib
ility
Id
enti
fica
tio
n
Additional records identified through other sources:
(n= 8)
Records after duplicates removed (n= 2883)
Records screened (n= 2883)
Records excluded: on title and abstract (n= 2788)
Full-text articles assessed for eligibility
(n= 95) Full-text articles excluded:
- other outcome than mortality or recurrence (n= 28)
- other exposure than dietary patterns/indices or our selected food items (n= 11)
- Not only cancer survivors (n= 10)
- Excluded because of our own exclusion criteria (sample size, follow-up length) (n= 8)
Studies included in qualitative synthesis
(n= 38)
Page 70 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Online supporting data File S2 for manuscript ‘the impact of dietary patterns and the main food groups on mortality and
recurrence in cancer survivors: systematic review of current epidemiological literature’
Jochems et al., 11-08-2017
Page 71 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Bladder cancer
Table S1: Summary of studies bladder cancer
Author (year)
Study / country Number of participants / sex (age range)
Follow-up period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI)
multivariate adjusted
Adjustment
Tang et al. (2010)
Roswell Park Cancer Institute (RPCI) / United States
239 m/w (not specified) 8.0 Fruit and vegetables
Pre-diagnosis FFQ usual diet in the few years before diagnosis
Overall mortality, cancer-specific mortality
Total fruit: HR1= 0.91; 95% CI 0.62-1.33 HR2= 1.09; 95% CI 0.66-1.81 Total vegetables: HR1= 0.91; 95% CI 0.62-1.36 HR2= 1.06; 95% CI 0.63-1.78 Cruciferous vegetables: HR1= 0.87; 95% CI 0.60-1.26 HR2= 0.89; 95% CI 0.53-1.48 Raw cruciferous vegetables: HR1= 0.73; 95% CI 0.50-1.06 HR2= 0.73; 95% CI 0.44-1.21
age at diagnosis, total meat intake, pack-years of smoking, tumour stage, radiation therapy
Page 72 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Bowel cancer
Table S2: Summary of studies bowel cancer
Author (year) Study / country Number of participants / sex (age at baseline)
Follow-up
period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Meyerhardt et al. (2007)
Cancer and Leukemia Group B (CALGB 89803) adjuvant therapy trial for stage III colon cancer / USA
1,009 m/w (21-85) 5.3 PCA: prudent diet, Western diet
Post-diagnosis
FFQ during and 6 months after adjuvant chemotherapy
Overall mortality, cancer recurrence
Prudent diet: HR1= 1.32; 95% CI 0.86-2.04 HR4= 1.13; 95% CI 0.77-1.67 Western diet: HR1= 2.32; 95% CI 1.36-3.96 HR4= 2.85; 95% CI 1.75-4.63
sex, age, depth of invasion through bowel wall, number of positive lymph nodes, presence of clinical perforation at time of surgery, presence of bowel obstruction at time of surgery, baseline performance status, treatment group, weight change between first and second questionnaire, time-varying BMI, time-varying physical activity level, time-varying total calories
McCullough et al. (2013)
Cancer Prevention Study II (CPSII) Nutrition Cohort / USA
2,315 m/w (40-93) 7.5 Red and processed meat, unprocessed red meat
Pre- and post-diagnosis
FFQ usual diet of the year before diagnosis and two times during follow-up
Overall mortality, cancer-specific mortality, death from other causes
Pre-diagnosis: Red and processed meat RR1= 1.29; 95% CI 1.05-1.59 RR2= 1.09; 95% CI 0.79-1.51 RR3= 1.39; 95% CI 1.00-1.92 Unprocessed red meat RR1= 1.12; 95% CI 0.92-1.38 RR2= 1.16; 95% CI 0.84-1.58 RR3= 1.19; 95% CI 0.87-1.64 Post-diagnosis: Red and processed meat RR1= 0.94; 95% CI 0.68-1.30 RR2= 1.10; 95% CI 0.61-1.98 RR3= 0.87; 95% CI 0.54-1.41 Unprocessed red meat: RR1= 0.75; 95% CI 0.55-1.03 RR2= 1.13; 95% CI 0.62-2.06 RR3= 0.64; 95% CI 0.40-1.03
pre-diagnosis model: age at diagnosis, sex, tumour stage at diagnosis, 1992 pre-diagnostic energy intake, BMI in 1992, history of diabetes, and history of myocardial infarction. Post-diagnosis model: age at diagnosis, sex, tumour stage at diagnosis, and post-diagnostic energy intake, weight change between 1992 pre-diagnostic, post-diagnostic questionnaires, and 1992 pre-diagnostic meat intake
Zhu et al. (2013)
Familial CRC registry in Newfoundland (FBCR) / Canada
529 m/w (20-75) 6.4 PCA: prudent vegetable pattern
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
Prudent vegetable pattern: HR1= 1.03; 95% CI 0.61-1.75 HR2= 1.12; 95% CI 0.69-1.84
total energy intake, sex, age at diagnosis, stage at diagnosis, marital status, family history, reported screening procedure, reported chemo-radiothr and microsatellite instability status
Dik et al. (2014)
The European Prospective Investigation into Cancer and Nutrition (EPIC) / pooled analysis data Denmark, Italy,
3,859 m/w (25-70) 4.1 Total dairy, milk, yoghurt, cheese
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
Total dairy: HR1= 1.16; 95% CI 0.98-1.36 HR2= 1.17; 95% CI 0.96-1.43 Milk:
age at diagnosis, sex, pre-diagnosis BMI, smoking status, energy intake, tumour subsite, disease stage,
Page 73 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Netherlands, Norway, Spain, Sweden, United Kingdom, France, Germany, Greece
HR1= 1.21; 95% CI 1.03-1.43 HR2= 1.21; 95% CI 0.99-1.48 Yoghurt: HR1= 1.08; 95% CI 0.92-1.28 HR2= 1.09; 95% CI 0.88-1.34 Cheese: HR1= 0.87; 95% CI 0.74-1.04 HR2= 0.93; 95% CI 0.76-1.14
differentiation grade; stratified by centre
Fung et al. (2014)
Nurses' Health Study (NHS) / USA
1,201 w (30-55) 11.2 Dietary indices: AHEI, DASH, AMED PCA: prudent diet, Western diet
Post-diagnosis
FFQ at least 6 months after diagnosis
Overall mortality, cancer-specific mortality
AHEI: HR1= 0.71; 95% CI 0.52-0.98 HR2= 0.72; 95% CI 0.43-1.21 DASH: HR1= 0.98; 95% CI 0.71-1.35 HR2= 0.87; 95% CI 0.52-1.45 AMED: HR1= 0.87; 95% CI 0.63-1.21 HR2= 0.84; 95% CI 0.50-1.42 Prudent diet: HR1= 0.93; 95% CI 0.65-1.34 HR2= 0.67; 95% CI 0.37-1.22 Western diet: HR1= 1.32; 95% CI 0.89-1.97 HR2= 1.66; 95% CI 0.85-3.23
age, physical activity, BMI, weight change, cancer grade, chemotherapy, smoking status, energy intake, colon or rectal cancer, stage of disease, and date of colorectal cancer diagnosis
Pelser et al. (2014)
NIH-AARP Diet and Health Study / USA
5,727 m/w (50-71) 5 Dietary indices: HEI
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
HEI: HR1= 0.95; 95% CI 0.78-1.16 HR2= 0.99; 95% CI 0.77-1.27
age, sex, lag time, education, family history cancer, stage, treatment, BMI, physical activity, alcohol, smoking
Skeie et al. (2014)
HELGA cohort including the Norwegian Women and Cancer Study, the Northern Sweden Health and Disease Study, and the Danish Diet Cancer and Health Study / Denmark, Norway, Sweden
1,119 m/w (30-64) 7 Total whole grains, whole grain wheat, whole grain rye, whole grain oats, whole grain products
Pre-diagnosis
FFQ usual diet before diagnosis
Overall mortality For men Total whole grains: HR1= 1.00; 95% CI 0.67-1.48 Whole grain wheat: HR1= 0.97; 95% CI 0.64-1.49 Whole grain rye: HR1= 0.90; 95% CI 0.60-1.36 Whole grain oats: HR1= 1.11; 95% CI 0.72-1.70 Whole grain products: HR1= 1.06; 95% CI 0.71-1.56 For women Total whole grains: HR1= 0.91; 95% CI 0.60-1.39 Whole grain wheat: HR1= 1.35; 95% CI 0.72-2.53 Whole grain rye: HR1= 0.93; 95% CI 0.60-1.46 Whole grain oats: HR1= 0.83; 95% CI 0.55-1.26 Whole grain products: HR1= 1.10; 95% CI 0.74-1.64
age at diagnosis, metastasis, smoking, folate, margarine, energy intake, stratified for country and cancer location. Wheat, rye and oats were also adjusted for the other grains
Page 74 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Yang et al. (2014)
Cancer Prevention Study II (CPSII) Nutrition Cohort / USA
2,284 m/w (40-92) 7.5 Total dairy, milk Pre- and post-diagnosis
FFQ usual diet of the year before diagnosis and two times during follow-up
Overall mortality, cancer-specific mortality
Pre-diagnosis: Total dairy RR1= 0.88; 95% CI 0.72-1.09 RR2= 0.89; 95% CI 0.65-1.22 Milk RR1= 0.95; 95% CI 0.79-1.15 RR2= 0.98; 95% CI 0.73-1.32 Post-diagnosis: Total dairy RR1= 0.75; 95% CI 0.56-1.01 RR2= 0.73; 95% CI 0.44-1.23 Milk RR1= 0.72; 95% CI 0.55-0.94 RR2= 0.93; 95% CI 0.59-1.49
pre-diagnosis: age at diagnosis, sex, tumour stage, pre-diagnosis total energy and total folate intakes. post-diagnosis: age at diagnosis, sex, tumour stage, post-diagnosis total energy and total folate intakes
Carr et al. (2016)
Darmkrebs: chancen der Verhutung durch Screening study (DACHS) / Germany
3,122 m/w (>30) 4.8 Red and processed meat
Pre-diagnosis
FFQ usual diet before diagnosis
Overall mortality, cancer-specific mortality, cancer recurrence
Red and processed meat: HR1= 0.85; 95% CI 0.67-1.09 HR2= 0.83; 95% CI 0.61-1.14 HR4= 1.03; 95% CI 0.80-1.33
age at diagnosis, sex, cancer stage, chemotherapy, surgery, BMI, physical activity, diabetes, stroke, heart failure, myocardial infarction, dairy intake, wholegrain intake, time between diagnosis and interview, time-dependent effect of chemotherapy
Romaguera et al. (2016)
The European Prospective Investigation into Cancer and Nutrition (EPIC) / pooled analysis data Denmark, Italy, Netherlands, Norway, Spain, Sweden, United Kingdom, France, Germany, Greece
3,292 m/w (25-70) 4.2 Dietary indices: WCRF/AICR score
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
WCRF/AICR score: HR1= 0.79; 95% CI 0.65-0.98 HR2= 0.70; 95% CI 0.56–0.89
age at diagnosis as entry time and age at death or censoring as exit time, year of diagnosis, tumour stage, tumour grade, tumour site, sex, educational level, and smoking status; stratified by country
Ward et al. (2016)
The European Prospective Investigation into Cancer and Nutrition (EPIC) / pooled analysis data Denmark, Italy, Netherlands, Norway, Spain, Sweden, United Kingdom, France, Germany, Greece
3,789 m/w (25-70) 4.1 Red meat, unprocessed meat, poultry
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
Red and processed meat: HR1= 1.00; 95% CI 0.83-1.20 HR2= 1.00; 95% CI 0.81-1.23 Unprocessed red meat: HR1= 0.95; 95% CI 0.78-1.14 HR2= 0.93; 95% CI 0.75-1.15 Poultry: HR1= 0.87; 95% CI 0.73-1.03 HR2= 0.91; 95% CI 0.75-1.10
adjusted for age at diagnosis, sex, BMI, smoking status, tumour grade, tumour stage, year of tumour diagnosis, energy intake, calcium intake, folate intake, alcohol intake, education; stratified by country
Ratjen et al. (2017)
Patients with histologically confirmed colorectal cancer recruited by the PopGen biobank / Germany
1,404 m/w (56-67) 7 Dietary indices: Modified Mediterranean Diet Score (MMDS), Healthy Nordic Food Index (HNFI)
Post-diagnosis
FFQ usual diet assessed 6 years (median) after diagnosis
Overall mortality MMDS: HR1= 0.48; 95% CI 0.32-0.74 HNFI: HR1= 0.63; 95% CI 0.39-1.04
sex, age at diet assessment, BMI, physical activity, survival time from CRC diagnosis until diet assessment, tumour location, occurrence of metastases, occurrence of other cancer, chemotherapy, smoking status, total energy intake, time 3 age, time 3 BMI, and time 3 metastases
Page 75 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Breast cancer
Table S3: Summary of studies breast cancer
Author (year)
Study / country Number of participants /
sex (age)
Follow-up
period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI)
multivariate adjusted
Adjustment
Hebert et al. (1998)
Memorial Sloan- Kettering Cancer Center Follow-up Study / USA
472 w (20-80) 8-10 Red meat, butter/margarine/lard
Post-diagnosis
FFQ usual diet before diagnosis
Cancer-specific mortality, cancer recurrence
Red Meat: RR2= 2.60; 95% CI 0.96-7.03 RR4= 1.12; 95% CI 0.66-1.89 Butter/margarine/lard: RR2= 1.03; 95% CI 0.61-1.76 RR4= 1.30; 95% CI 1.03-1.64
disease stage, oestrogen receptor status, age, BMI, menopausal status, energy intake
Holmes et al. (1999)
Nurses' Health Study (NHS) / USA
1,504 w pre-diagnosis (mean age 54) and 1,982 w post-diagnosis)
13.1 Vegetables, poultry, fish, dairy, red meat (processed and unprocessed combined)
Pre- and post-diagnosis
FFQ usual diet after diagnosis
Overall mortality, cancer-specific mortality (results for breast cancer-specific mortality are not shown in paper)
Pre-diagnosis: Vegetables RR1= 0.98; 95% CI 0.62-1.53 Poultry RR1= 0.60; 95% CI 0.39-0.92 Fish RR1= 0.94; 95% CI 0.62-1.43 Dairy RR1= 0.71; 95% CI 0.44-1.14 Red meat (not shown) Post-diagnosis: Vegetables RR1= 0.81; 95% CI 0.59–1.11 Poultry RR1= 0.70; 95% CI 0.50–0.97 Fish RR1= 0.80; 95% CI 0.60–1.07 Dairy RR1= 0.72; 95% CI 0.52–1.00 Red meat RR1= 1.06; 95% CI 0.76–1.49
pre-diagnosis: quantiles of nutrient or food intake prior to diagnosis, previous diet interval, age, diet interval, calendar year of diagnosis, body mass index, oral contraceptive use, menopausal status, postmenopausal hormone use, smoking, age at first birth and parity, number of metastatic lymph nodes, tumour size, and caloric intake post-diagnosis: age, diet interval, calendar year of diagnosis, body mass index, oral contraceptive use, menopausal status, postmenopausal hormone use, smoking, age at first birth and parity, number of metastatic lymph nodes, tumour size, caloric intake
Kroenke et al. (2005)
Nurses' Health Study (NHS) / USA
2,619 w (30-55) 9 PCA: prudent diet, Western diet
Pre- and post-diagnosis
FFQ usual intake 4 years before diagnosis and FFQ at least one year after diagnosis
Overall mortality, cancer-specific mortality, death from other causes
Pre-diagnosis: Prudent diet RR1= non-significant (not shown) RR2= non-significant (not shown) RR3= non-significant (not shown) Western diet RR1= 1.40; 95% CI 0.93-2.09 RR2= 1.01; 95% CI 0.59-1.72 RR3= 1.95; 95% CI 1.06-3.60
age, time since diagnosis, BMI, energy intake, smoking, physical activity, diet missing, age at menarche, oral contraceptive use, menopausal status and use of postmenopausal hormone therapy, age at menopause, tamoxifen, chemotherapy, tumour stage at diagnosis, time between dietary assessment and diagnosis (for pre-diagnosis diet)
Page 76 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Post-diagnosis: Prudent diet RR1= 0.78; 95% CI 0.54-1.12 RR2= 1.07; 95% CI 0.66-1.73 RR3= 0.54; 95% CI 0.31-0.95 Western diet RR1= 1.53; 95% CI 1.03-2.29 RR2= 1.01; 95% CI 0.60-1.70 RR3= 2.09; 95% CI 1.30-3.36
McEligot et al. (2006)
Cancer Surveillance Program of Orange County (CSPOC) / USA
516 w (age >50) 6.7 Fruit, vegetables Pre-diagnosis
FFQ usual diet one year before diagnosis
Overall mortality Total fruit: HR1= 0.63; 95% CI 0.38-1.05 Total vegetables: HR1= 0.57; 95% CI 0.35–0.94
tumour stage, age at diagnosis, BMI, parity, HRT, alcohol intake, multivitamins, energy intake
Chlebowski et al. (2006)
Women’s Intervention Nutrition Study (WINS) RCT / USA
2,437 w 5 Dietary indices: Low fat diet
Post-diagnosis
FFQ with interview on dietary intake after diagnosis
Overall mortality, relapse-free survival
Intervention versus control Low-fat diet: HR1= 0.89; 95% CI 0.65-1.21 HR4= 0.76; 95% CI 0.60-0.98
nodal status, systemic adjuvant therapy, ER status, tumour size, mastectomy
Pierce et al. (2007a)
Women's Healthy Eating and Living (WHEL) RCT / USA
3,088 w (18-70) 7.3 Dietary indices: Low fat diet
Post-diagnosis
FFQ with interview on dietary intake after diagnosis
Overall mortality Intervention versus control Low-fat diet: HR1= 0.91; 95% CI 0.72-1.15
anti-oestrogen use, bilateral oophorectomy, age, BMI, physical activity, energy intake, tumour characteristics (including hormone receptor status), years from diagnosis to study entry
Dal Maso et al. (2008)
Six Italian Regions Follow-up Study / Italy
1,453 w (23-74) 12.6 Fruit and vegetables Pre-diagnosis
FFQ usual diet year before diagnosis
Overall mortality, cancer-specific mortality
Fruit and vegetables: HR1= 1.27; 95% CI 1.00–1.61 HR2= 1.26; 95% CI 0.96–1.64 (low versus high intake!)
region, age at diagnosis, year of diagnosis, TNM stage, receptor status
Kwan et al. (2009)
Life After Cancer Epidemiology (LACE) study / USA
1,901 w (18-79) 4.2 PCA: prudent diet, Western diet
Post-diagnosis
FFQ usual diet 3 years after diagnosis
Overall mortality, cancer-specific mortality, death from other causes, cancer recurrence
Prudent diet: HR1= 0.57; 95% CI 0.36-0.90 HR2= 0.79; 95% CI 0.43-1.43 HR3= 0.35; 95% CI 0.17-0.73 HR4= 0.95; 95% CI 0.63-1.43 Western diet: HR1= 1.76; 95% CI 1.10-2.81 HR2= 1.20; 95% CI 0.62-2.32 HR3= 2.15; 95% CI 0.97-4.77 HR4= 0.98; 95% CI 0.62-1.54
age at diagnosis, total energy intake, ethnicity, BMI, weight change before diagnosis to baseline, smoking status, menopausal status at diagnosis, stage, hormone receptor status, treatment
Beasley et al. (2011)
Collaborative Women’s Longevity Study (CWLS) / USA
4,441 w (20-79) 5.5 Fruit, vegetables, dairy, meat (poultry, fish, beef, and processed)
Post-diagnosis
FFQ usual diet after diagnosis (1-16 years)
Overall mortality, cancer-specific mortality
Total fruit: HR1= 1.38; 95% CI 0.88-2.17 HR2= 1.39; 95% CI 0.64-2.99 Total vegetables: HR1= 1.44; 95% CI 0.91-2.27 HR2= 0.96; 95% CI 0.38-2.45 Cruciferous vegetables: HR1= 1.02; 95% CI 0.80-1.30 HR2= 0.95; 95% CI 0.59-1.54 Dairy: HR1= 1.18; 95% CI 0.90-1.54 HR2= 0.94; 95% CI 0.56-1.59
age, state of residence, menopausal status, smoking, breast cancer stage, alcohol, history of hormone replacement therapy at diagnosis, interval between diagnosis and diet assessment, and energy intake, breast cancer treatment, body mass at follow-up
Page 77 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Meat (poultry, fish, beef, and processed): HR1= 1.12; 95% CI 0.83-1.51 HR2= 0.89; 95% CI 0.50-1.60
Buck et al. (2011)
MARIE study / Germany
2,653 w (50-74) 6.4 Fruit, vegetables, bread, sunflower/pumpkin seeds, sesame/flaxseeds
Pre-diagnosis
FFQ usual diet year before diagnosis
Overall mortality, cancer-specific mortality
Fruit: HR1= 0.84; 95% CI 0.61-1.16 HR2= 0.86; 95% CI 0.59-1.25 Vegetables: HR1= 1.09; 95% CI 0.80-1.48 HR2= 1.01; 95% CI 0.70-1.46 Bread: HR1= 1.31; 95% CI 0.93-1.83 HR2= 1.10; 95% CI 0.74-1.63 Sunflower/pumpkinseeds: HR1= 0.87; 95% CI 0.66-1.15 HR2= 1.12; 95% CI 0.79-1.57 Sesame/flaxseeds: HR1= 0.90; 95% CI 0.68-1.19 HR2= 1.21; 95% CI 0.87-1.68
tumour size, nodal status, metastasis, grade, oestrogen and progesterone receptor status, breast cancer detection type, diabetes, HRT use at diagnosis, study centre, energy intake, age at diagnosis
George et al. (2011)
Health, Eating, Activity, and Lifestyle (HEAL) / USA
670 w (older than 18)
6 Dietary indices: HEI
Post-diagnosis
FFQ usual diet approx. 2,5 yrs after diagnosis
Overall mortality, cancer-specific mortality
HEI: HR1= 0.40; 95% CI 0.17-0.94 HR2= 0.12; 95% CI 0.02-0.99
energy intake, physical activity, race, tumour stage, tamoxifen use, BMI
Kim et al. (2011)
Nurses' Health Study (NHS) / USA
2,729 w (30-55) not stated
Dietary indices: AHEI, DQIR, RFS, AMED
Post-diagnosis
FFQ usual diet around 1 year after diagnosis
Overall mortality, cancer specific mortality, death from other causes
AHEI: RR1= 0.85; 95% CI 0.63-1.17 RR2= 1.53; 95% CI 0.98-2.39 RR3= 0.52; 95% CI 0.32-0.83 DQIR: RR1= 0.78; 95% CI 0.58-1.07 RR2= 0.81; 95% CI 0.53-1.24 RR3= 0.85; 95% CI 0.54-1.34 RFS: RR1= 1.03; 95% CI 0.74-1.42 RR2= 1.54; 95% CI 0.95-2.47 RR3= 0.86; 95% CI 0.54-1.37 AMED: RR1= 0.87; 95% CI 0.64-1.17 RR2= 1.15; 95% CI 0.74-1.77 RR3= 0.80; 95% CI 0.50-1.26
age, time since diagnosis, alcohol intake (only for RFS because alcohol is a component in the other 3 diet quality indices), energy, multivitamin use (except for AHEI because it is a component), BMI, weight change (BMI at time of diet minus BMI just prior to diagnosis), oral contraceptive use, smoking status, physical activity in METs, stage, categories of treatment, age at first birth and parity, menopausal status and postmenopausal hormone use
Izano et al. (2013)
Nurses' Health Study (NHS) / USA
4,103 w (30-55) 9.3 Dietary indices: AHEI, DASH
Post-diagnosis
FFQ usual diet around 1 year after diagnosis
Cancer specific mortality, death from other causes
AHEI: RR2= 1.07; 95% CI 0.77-1.49 RR3= 0.57; 95% CI 0.42-0.77 DASH: RR2= 0.85; 95% CI 0.61-1.19 RR3= 0.72; 95% CI 0.53-0.99
stratified by time since diagnosis, adjusted for age at diagnosis, quintiles of energy intake, BMI and BMI change, age at first birth and parity, oral contraceptive use, menopausal status and HRT use, smoking, stage of disease, radiation treatment, chemotherapy and hormonal treatment, and physical activity
Kroenke et al. (2013)
Life After Cancer Epidemiology (LACE) study / USA
1,893 w (18-70) 11.8 Dairy Post-diagnosis
FFQ diet at diagnosis when cancer recurs before 6 yrs -
Overall mortality, cancer recurrence
Total Dairy: HR1= 1.39; 95% CI 1.02-1.90 HR4= 1.13; 95% CI 0.83-1.54 Low-fat dairy: HR1= 1.05; 95% CI 0.80-1.36
age, time between diagnosis and dietary assessment, high- and low-fat dairy intake, race, education, cancer stage at diagnosis, tumour size, human epidermal growth receptor 2, nodal and oestrogen receptor
Page 78 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
otherwise 6 years after diagnosis
HR4= 1.01; 95% CI 0.78-1.32 High-fat dairy: HR1= 1.64; 95% CI 1.24-2.17 HR4= 1.22; 95% CI 0.92-1.55
status, chemotherapy, radiation, tamoxifen, comorbidity, menopausal status, BMI, physical activity, energy intake, alcohol intake, red meat intake, fibre intake, fruit intake
Nechuta et al. (2013)
After Breast Cancer Pooling Project (includes cohorts SBCSS, LACE, WHEL, NHS) / USA and China
11,390 w (20-83) 9.0 Cruciferous vegetables Post-diagnosis
FFQ approx. 2 yrs after diagnosis
Overall mortality, cancer recurrence
Cruciferous vegetables: HR1= 0.99; 95% CI 0.86-1.13 HR4= 1.10; 95% CI 0.95-1.28
age at diagnosis, ER/PR status, TNM stage, surgery, chemotherapy, radiotherapy, hormonal therapy, smoking, BMI, exercise, menopausal status, race/ethnicity, education
Vrieling et al. (2013)
Mammary carcinoma Risk factor Investigation (MARIE) study / Germany
2,522 w 5.5 PCA: healthy pattern, unhealthy pattern
Pre-diagnosis
FFQ usual diet the year before diagnosis
Overall mortality, cancer specific mortality, death from other causes, cancer recurrence
Healthy pattern: HR1= 0.87; 95% CI 0.61-1.23 HR2= 0.89; 95% CI 0.59-1.35 HR3= 0.81; 95% CI 0.40-1.61 HR4= 0.71; 95% CI 0.48-1.06 Unhealthy pattern: HR1= 1.34; 95% CI 0.93-1.94 HR2= 0.99; 95% CI 0.64-1.52 HR3= 3.69; 95% CI 1.66-8.17 HR4= 0.91; 95% CI 0.61-1.36
tumour size, nodal status, metastases, tumour grade, ERPR status, radiotherapy, HRT use at diagnosis, mode of detection, and total energy intake and stratified by age at diagnosis and study centre
George et al. (2014)
Women’s Health Initiative’s Dietary Modification Trial and Observational Study (WHI) / USA
2,317 w (50-79) 9.6 Dietary indices: HEI
Post-diagnosis
FFQ usual diet approx. 1.5 yrs after diagnosis
Overall mortality, cancer specific mortality, death from other causes
HEI: HR1= 0.74; 95% CI 0.55-0.99 HR2= 0.91; 95% CI 0.60-1.40 HR3= 0.58; 95% CI 0.38-0.87
age at screening visit, WHI component, ethnicity, income, education, stage, estrogen receptor status, progesterone receptor status, time since diagnosis, energy intake in kcals, physical activity in MET, servings of alcohol per week, use of postmenopausal hormone therapy
McCullough et al. (2016)
Cancer Prevention Study II (CPS-II) Nutrition Cohort / USA
4,452 w for pre-diagnosis and 2,152 w for post-diagnosis (mean age 70.7 yrs)
9.8-9.9 Dietary indices: ACS Fruit and vegetables, red and processed meat
Pre- and post-diagnosis
FFQ usual diet in 1992 (before diagnosis) and usual diet at least 1 year after diagnosis (after diagnosis)
Overall mortality, cancer-specific mortality, death from other causes
Pre-diagnosis: ACS diet score RR1= 1.00; 95% CI 0.84-1.18 RR2= 1.06; 95% CI 0.79-1.42 RR3= 1.02; 95% CI 0.79-1.31 Fruit and vegetables RR1= 1.06; 95% CI 0.85-1.33 RR2= 1.00; 95% CI 0.66-1.50 RR3= 1.11; 95% CI 0.81-1.52 Red and processed meat: RR1= 0.88; 95% CI 0.73-1.06 RR2= 1.10; 95% CI 0.80-1.52 RR3= 0.81; 95% CI 0.62-1.07 Post-diagnosis: ACS diet score RR1= 0.93; 95% CI 0.73-1.18 RR2= 1.44; 95% CI 0.90-2.30 RR3= 0.78; 95% CI 0.56-1.07 Fruit and vegetables RR1= 1.03; 95% CI 0.80-1.33 RR2= 1.31; 95% CI 0.83-2.06 RR3= 0.93; 95% CI 0.65-1.34 Red and processed meat RR1= 0.64; 95% CI 0.49-0.84
age at diagnosis, diagnosis year, tumour stage at diagnosis, tumour grade at diagnosis, estrogen receptor status, progesterone receptor status, initial treatment (surgery, chemotherapy, radiation, hormone therapy, aromatase inhibitor use and/or Herceptin use), and the following assessed at the time of FFQ completion: BMI, smoking status, physical activity and energy intake
Page 79 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
RR2= 0.88; 95% CI 0.54-1.43 RR3= 0.57; 95% CI 0.39-0.82
Page 80 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Laryngeal cancer
Table S4: Summary of studies laryngeal cancer
Author (year) Study / country Number of participants / sex
(age)
Follow-up period (yrs)
Exposure Exposure timeframe Exposure assessment Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Crosignani et al. (1996)
Lombardy Cancer Registry (LCR) / Italy
213 m (32-75) 8-10 yr Meat (beef, veal), poultry, fish, eggs, milk, cheese, bread, pasta, potatoes, vegetables, citrus fruits, other fruits, butter, olive oil
Pre-diagnosis Interview usual diet in year before diagnosis
Overall mortality
Citrus fruits: HR1= 0.76; 95% 0.49-1.19 Other fruits: HR1= 0.65; 95% CI 0.39-1.07 Vegetables: HR1= 0.57; 95% CI 0.35-0.94 Meat: HR1= 0.50; 95% CI 0.30-0.83 Poultry: HR1= 0.90; 95% CI 0.55-1.46 Fish: HR1= 0.91; 95% CI 0.59-1.39 Eggs: HR1= 1.22; 95% CI 0.74-2.00 Milk: HR1= 1.58; 95% CI 0.99-2.55 Cheese: HR1= 0.70; 95% CI 0.44-1.12 Bread: HR1= 0.54; 95% CI 0.32-0.90 Pasta: HR1= 1.25; 95% CI 0.76-2.04 Potatoes: HR1= 1.02; 95% CI 0.64-1.64 Butter: HR1= 1.11; 95% CI 0.69-1.80 Olive oil: HR1= 0.71; 95% CI 0.44-1.16
age at diagnosis, clinical stage, occurrence of new primary cancers
Page 81 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Non-Hodgkin lymphoma (NHL)
Table S5: Summary of studies non-Hodgkin lymphoma
Author (year)
Study / country Number of participants /
sex (age)
Follow-up period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Han et al. (2010)
Yale Connecticut Tumor Registry New York (CTR) / USA
568 w (21-84) 7.7 Fruit, vegetables Pre-diagnosis
FFQ usual diet the year before diagnosis
Overall mortality, cancer-specific mortality
Total fruit and vegetables: HR1= 0.68; 95% CI 0.49-0.95 HR2= 0.70; 95% CI 0.45-1.10 Total fruit: HR1= 0.91; 95% CI 0.70-1.18 HR2= 1.04; 95% CI 0.74-1.45 Total vegetables: HR1= 0.58; 95% CI 0.38-0.89 HR2= 0.58; 95% CI 0.33-1.03 Cruciferous vegetables: HR1= 0.91; 95% CI 0.67–1.24 HR2= 0.75; 95% CI 0.49–1.14 Bean vegetables: HR1= 1.14; 95% CI 0.85-1.54 HR2= 1.05; 95% CI 0.71-1.55 Green leafy vegetables: HR1= 0.71; 95% CI 0.51-0.98 HR2= 0.82; 95% CI 0.54-1.23 Red vegetables: HR1= 1.03; 95% CI 0.76-1.38 HR2= 1.11; 95% CI 0.76-1.62 Yellow vegetables: HR1= 0.93; 95% CI 0.69-1.25 HR2= 1.11; 95% CI 0.77-1.61 Citrus fruits: HR1= 0.73; 95% CI 0.54-0.99 HR2= 0.81; 95% CI 0.54-1.20
age, education, stage, B-symptom, initial treatment, total energy intake
Leo et al. (2015)
Multi-ethnic Cohort (MEC) / USA
2,339 m/w (45-75)
4.5 Fruit, vegetables, dairy, legumes, fish, red meat
Pre-diagnosis
FFQ usual diet the year before diagnosis
Overall mortality, cancer specific mortality
Vegetables: HR1= 0.98; 95% CI 0.85-1.12 HR2= 0.98; 95% CI 0.83-1.16 Fruits: HR1= 1.03; 95% CI 0.90-1.19 HR2= 1.04; 95% CI 0.88-1.24 Red meat: HR1= 1.00; 95% CI 0.87-1.15 HR2= 0.95; 95% CI 0.81-1.13 Fish: HR1= 0.90; 95% CI 0.78-1.03 HR2= 0.91; 95% CI 0.76-1.08 Legumes: HR1= 0.88; 95% CI 0.76-1.01 HR2= 0.86; 95% CI 0.72-1.02 Dairy products:
age at NHL diagnosis, BMI, sex, ethnicity, SEER summary stage, NHL subtype, chemo-, radio-, immuno-, and steroid-therapy, smoking status at baseline, alcohol use, education status, energy intake, number of comorbidities
Page 82 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
HR1= 1.14; 95% CI 1.00-1.31 HR2= 1.16; 95% CI 0.98-1.37
Page 83 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Prostate cancer
Table S6: Summary of studies prostate cancer
Author (year)
Study / country Number of participants /
sex (age)
Follow-up period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Chavarro et al. (2008)
Physician’s Health Study (PHS) / USA
2,161 m 19 Total fish Pre-diagnosis
FFQ usual diet before diagnosis
Cancer-specific mortality
Total fish: HR2= 0.52; 95% CI 0.30-0.91
age at prostate cancer diagnosis, BMI, physical activity, alcohol use, tomato and dairy products, smoking, ethnicity, multivitamin and vitamin E supplements, random assignment to aspirin or beta-carotene, tumour stage, grade at diagnosis, clinical presentation of case
Kenfield et al. (2014)
Health Professionals Follow-up Study (HPFS) / USA
4,538 m (40-75) 23.2 Dietary indices: Mediterranean diet score (MDS)
Post-diagnosis
FFQ usual diet after diagnosis
Overall mortality, cancer-specific mortality
MDS: HR1= 0.78; 95% CI 0.67-0.90 HR2= 1.01; 95% CI 0.75-1.38
age at diagnosis, time period, time diagnosis to FFQ, energy, BMI, vigorous physical activity, smoking status, clinical stage, Gleason score, treatment
Yang et al. (2015a)
Physician’s Health Study (PHS) / USA
926 m (40-84) 9.6 Total dairy, high-fat dairy, low-fat dairy
Post-diagnosis
FFQ usual diet after diagnosis
Overall mortality Total dairy: HR1 = 1.76; 95% CI 1.21-2.55 HR2 = 2.41; 95% CI 0.96-6.02 High-fat dairy: HR1= 1.22; 95% CI 1.08-1.38 HR2= 1.30; 95% CI 0.97-1.73 Low-fat dairy: HR1= 1.17; 95% CI 1.05-1.29 HR2= 1.16; 95% CI 0.88-1.53
age at diagnosis, total energy intake, BMI, smoking status, exercise, Gleason score, clinical stage, prostate-specific antigen level, time interval between diagnosis and FFQ completion, initial treatment after diagnosis, family history of prostate cancer, and indicators for prudent dietary pattern and Western dietary pattern after excluding dairy products
Yang et al. (2015b)
Physician’s Health Study (PHS) / USA
926 m (40-84) 8.7 PCA: prudent diet, Western diet
Post-diagnosis
FFQ usual after diagnosis
Overall mortality, cancer specific mortality
Prudent diet: RR1= 0.64; 95% CI 0.44-0.93 RR2= 0.46; 95% CI 0.17-1.24 Western diet: RR1= 1.67; 95% CI 1.16-2.42 RR2= 2.53; 95% CI 1.00-6.42
age at diagnosis, total energy intake, BMI, smoking status, vigorous physical activity, Gleason score, clinical stage, prostate-specific antigen level, time interval between diagnosis and FFQ completion, initial treatment, family history of prostate cancer
HR1/RR1= overall mortality
HR2/RR2= cancer-specific mortality
HR3/RR3= death from other causes
HR4/RR4= cancer recurrence
FFQ= food frequency questionnaire
BMI= body mass index
Page 84 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Online supporting data File S3 for manuscript ‘the impact of dietary patterns
and the main food groups on mortality and recurrence in cancer survivors:
systematic review of current epidemiological literature’
Jochems et al., 11-08-2017
Page 85 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
The Grading of Recommendations Assessment, Development and Evaluation (GRADE)
GRADE is a systematic and explicit approach to making judgements about quality of
evidence and strength of recommendations. The focus is on clinical outcomes that patients
themselves are aware of in relation to their condition – in this systematic review these
include overall mortality, cancer-specific mortality, death from other causes, and cancer
recurrence. With the use of GRADE, the evidence is not rated study by study but across
studies for each individual outcome. Individual study quality was assessed with the
Cochrane Collaboration risk of bias assessment tools; the RoB 2.0 tool for randomised
trials and the ROBINS-I tool for cohort studies. Even before assessing the individual
study quality, studies were excluded from the systematic review if the sample size for the
analysis was <200 (comparisons containing less than 200 participants in total are
described as sparse data), the follow-up period was <4 years (for most cancer types, the
risk of cancer recurrence is the greatest within the first three years), no adjustments in the
statistical analysis were made for age and disease stage and, where possible, for cancer
treatment (e.g. studies adjusting for age and energy intake only were excluded).
Additionally, outcomes combining cancer recurrence with cancer progression, or
confirmed cancer-specific mortality combined with a diagnosis of metastasis, or prostate
cancer recurrence is determined by a rising PSA level only, were excluded. Therefore,
methodological flaws within the component studies will not cause any problems in the
GRADE evaluation - inconsistency of results across different studies will.
1. Quality of evidence
Table 1: Quality of Evidence Grades
Grade Definition
High We are very confident that the true effect lies close to that of the estimate of the effect.
Moderate We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect.
Very Low We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
2. Included study design
Table 2: Judgements about the quality of evidence by study design
Study Consequence
Randomised trials without important limitations – high quality evidence (+ 4 points)
Cohort studies without strengths or important limitations – low quality evidence (+ 2 points)
3. Determining the quality of evidence
Table 3: Factors that can reduce or increase the quality of the evidence
Factor Consequence
Limitations in study design or execution ↓ 1 or 2 levels
Inconsistency of results ↓ 1 or 2 levels
Indirectness of evidence ↓ 1 or 2 levels
Imprecision ↓ 1 or 2 levels
Publication bias ↓ 1 or 2 levels
Large magnitude of effect ↑ 1 or 2 levels
Dose-response gradient ↑ 1 level
All plausible confounding would reduce the demonstrated effect or increase the effect if no effect was observed
↑ 1 level
Page 86 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4. Study limitations randomised trials
Table 4: Study limitations in randomised trials Factor Explanation
Lack of allocation concealment Those enrolling patients are aware of the group (or period in a crossover trial) to which the next enrolled patient will be allocated (a major problem in “pseudo” or “quasi” randomized trials with allocation by day of week, birth date, chart number, etc.)
Lack of blinding Patient, caregivers, those recording outcomes, those adjudicating outcomes, or data analysts are aware of the arm to which patients are allocated (or the medication currently being received in a crossover trial)
Incomplete accounting of patients and outcome events Loss to follow-up and failure to adhere to the intention-to-treat principle in superiority trials; or in non-inferiority trials, loss to follow-up, and failure to conduct both analyses considering only those who adhered to treatment, and all patients for whom outcome data are available. The significance of rates of loss to follow-up, however, varies widely and is dependent on the relation between loss to follow-up and number of events. The higher the proportion lost to follow-up in relation to intervention and control group event rates, and differences between intervention and control groups, the greater the threat of bias
Selective outcome reporting Incomplete or absent reporting of some outcomes and not others based on the results
Other limitations Stopping trial early for benefit. Substantial overestimates are likely in trials with fewer than 500 events and that large overestimates are likely in trials with fewer than 200 events. Empirical evidence suggests that formal stopping rules do not reduce this bias. Use of invalidated outcome measures (e.g. patient-reported outcomes). Carryover effects in crossover trial. Recruitment bias in cluster-randomized trials
5. Study limitations cohort studies
Table 5: Study limitations in observational studies Factor Explanation
Failure to develop and apply appropriate eligibility criteria (inclusion of control population)
Selection of exposed and unexposed in cohort studies from different populations
Flawed measurement of both exposure and outcome Differences in measurement of exposure
Differential surveillance for outcome in exposed and unexposed in cohort studies
Failure to adequately control confounding Failure of accurate measurement of all known prognostic factors
Failure to match for prognostic factors and/or adjustment in statistical analysis
Incomplete or inadequately short follow-up Especially within prospective cohort studies, both groups should be followed for the same amount of time
6. Grading assessors
A first assessor will grade the quality of evidence for each outcome (cancer recurrence or
overall mortality or cancer-specific mortality or death from other causes) for each cancer type
with data on pre- or post-diagnosis dietary patterns or foods as exposure (bladder cancer, bowel
cancer, breast cancer, laryngeal cancer, prostate cancer). The first assessor will summarize the
findings in summary of findings tables for all evidence obtained. A second assessor will check
the consistency of the ratings of the first assessor. Disagreement about evidence were resolved
through consensus or a third party.
Page 87 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7. Summary of findings tables
Bladder cancer
Table 1: Bladder cancer and pre-diagnosis fruit and vegetable intake
Outcomes Hazard ratio / Relative risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total fruit: HR= 0.91; 95% CI 0.62-1.33
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total vegetables: HR= 0.91; 95% CI 0.62-1.36
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cruciferous vegetables: HR= 0.87; 95% CI 0.60-1.26
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Raw cruciferous vegetables: HR= 0.73; 95% CI 0.50-1.06
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total fruit: HR= 1.09; 95% CI 0.66-1.81
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total vegetables: HR= 1.06; 95% CI 0.63-1.78
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cruciferous vegetables: HR= 0.89; 95% CI 0.53-1.48
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Raw cruciferous vegetables: HR= 0.73; 95% CI 0.44-1.21
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bowel cancer
Table 2: Bowel cancer and pre-diagnosis diet quality indices
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality
HEI-2005: HR= 0.95; 95% CI 0.78-1.16
5727 (1) + Cohort ++ and downgraded one level; data of only 1 study
WCRF/AICR score: HR= 0.79; 95% CI 0.65-0.98
3292 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
HEI-2005: HR= 0.99; 95% CI 0.77-1.27
5727 (1) + Cohort ++ and downgraded one level; data of only 1 study
WCRF/AICR score: HR= 0.70; 95% CI 0.56–0.89
3292 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis diet quality indices
Overall mortality
AHEI-2010: HR= 0.71; 95% CI 0.52-0.98
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: HR= 0.98; 95% CI 0.71-1.35
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: HR= 0.87; 95% CI 0.63-1.21
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
MMDS: HR= 0.48; 95% CI 0.32-0.74
1404 (1) + Cohort ++ and downgraded one level; data of only 1 study. Although the study has a large estimate HR<0.5, it is based on 1 study only and will therefore be downgraded
HNFI: HR= 0.63; 95% CI 0.39-1.04
1404 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
AHEI-2010L HR= 0.72; 95% CI 0.43-1.21
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: HR= 0.87; 95% CI 0.52-1.45
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: HR= 0.84; 95% CI 0.50-1.42
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 88 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 3: Bowel cancer and pre-diagnosis prudent/healthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality
Prudent vegetable pattern: HR= 1.03; 95% CI 0.61-1.75
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Prudent vegetable pattern: HR= 1.12; 95% CI 0.69-1.84
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis prudent/healthy diet
Cancer recurrence
Prudent diet: HR= 1.13; 95% CI 0.77-1.67
1009 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality
Prudent diet: HR= 1.32; 95% CI 0.86-2.04 HR= 0.93; 95% CI 0.65-1.34
2210 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Cancer-specific mortality
Prudent diet: HR= 0.67; 95% CI 0.37-1.22
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 4: Bowel cancer and pre-diagnosis Western/unhealthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Processed meat pattern: HR= 1.53; 95% CI 0.85-2.74
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
High sugar pattern: HR= 1.27; 95% CI 0.72-2.25
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Processed meat pattern: HR= 1.82; 95% CI 1.07-3.09
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
High sugar pattern: HR= 1.02; 95% CI 0.62-1.69
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis Western/unhealthy diet
Cancer recurrence Western diet: HR= 2.85; 95% CI 1.75-4.63
1009 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Western diet: HR= 2.32; 95% CI 1.36-3.96 HR= 1.32; 95% CI 0.89-1.97
2210 (2) + Cohort ++ and downgraded one level; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with an increased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as only one estimate is ‘statistically significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Cancer-specific mortality
Western diet: HR= 1.66; 95% CI 0.85-3.23
1201(1) + Cohort ++ and downgraded one level; data of only 1 study
Table 5: Bowel cancer and pre-diagnosis grain foods
Outcomes Hazard Ratio / Relative Risk (95% CI) No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Men Total whole grains: HR= 1.00; 95% CI 0.67-1.48 Women Total whole grains: HR= 0.91; 95% CI 0.60-1.39
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain wheat: HR= 0.97; 95% CI 0.64-1.49 Women Whole grain wheat: HR= 1.35; 95% CI 0.72-2.53
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain rye: HR= 0.90; 95% CI 0.60-1.36 Women Whole grain rye: HR= 0.93; 95% CI 0.60-1.46
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain oats: HR= 1.11; 95% CI 0.72-1.70 Women Whole grain oats: HR= 0.83; 95% CI 0.55-1.26
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain products: HR= 1.06; 95% CI 0.71-1.56 Women Whole grain products: HR= 1.10; 95% CI 0.74-1.64
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 89 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 6: Bowel cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies) Quality of the evidence (GRADE) Comments
Cancer recurrence Red and processed meat: HR= 1.03; 95% CI 0.80-1.33
3122 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Unprocessed red meat: RR= 1.12; 95% CI 0.92-1.38 HR= 0.95; 95% CI 0.78-1.14
6104 (2) + Cohort ++ and downgraded one level; both studies found no statistically significant association, the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Red and processed meat: RR= 1.29; 95% CI 1.05-1.59 HR= 1.00; 95% CI 0.83-1.20 HR= 0.85; 95% CI 0.67-1.09
9226 (3) + Cohort ++ and downgraded one level; both studies found no statistically significant association, the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade one level for consistency lack of agreement between studies)
Poultry: HR= 0.87; 95% CI 0.73-1.03
3789 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer- specific mortality
Unprocessed red meat: RR= 1.16; 95% CI 0.84-1.58 HR= 0.93; 95% CI 0.75-1.15
6104 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Red and processed meat: RR= 1.09; 95% CI 0.79-1.51 HR= 1.00; 95% CI 0.81-1.23 HR= 0.83; 95% CI 0.61-1.14
9226 (3) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade for consistency lack of agreement between studies)
Poultry: HR= 0.91; 95% CI 0.75-1.10
3789 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Unprocessed red meat: RR= 1.19; 95% CI 0.87-1.64
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 1.39; 95% CI 1.00-1.92
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bowel cancer and post-diagnosis protein foods
Overall mortality Unprocessed red meat: RR= 0.75; 95% CI 0.55-1.03
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.94; 95% CI 0.68-1.30
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Unprocessed red meat: RR= 1.13; 95% CI 0.62-2.06
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 1.10; 95% CI 0.61-1.98
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Unprocessed red meat: RR= 0.64; 95% CI 0.40-1.03
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.87; 95% CI 0.54-1.41
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 90 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 7: Bowel cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy: RR= 0.88; 95% CI 0.72-1.09 HR= 1.16; 95% CI 0.98-1.36
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Milk: HR= 1.21; 95% CI 1.03-1.43 RR= 0.95; 95% CI 0.79-1.15
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Yoghurt: HR= 1.08; 95% CI 0.92-1.28
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cheese: HR= 0.87; 95% CI 0.74-1.04
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: HR= 1.17; 95% CI 0.96-1.43 RR= 0.89; 95% CI 0.65-1.22
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Milk: HR= 1.21; 95% CI 0.99-1.48 RR= 0.98; 95% CI 0.73-1.32
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade one level for consistency lack of agreement between studies)
Yoghurt: HR= 1.09; 95% CI 0.88-1.34
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cheese: HR= 0.93; 95% CI 0.76-1.14
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis dairy and alternatives
Overall mortality Total dairy: RR= 0.75; 95% CI 0.56-1.01
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Milk: RR= 0.72; 95% CI 0.55-0.94
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: RR= 0.73; 95% CI 0.44-1.23
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Milk: RR= 0.93; 95% CI 0.59-1.49
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 91 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Breast cancer
Table 8: Breast cancer and pre-diagnosis diet quality indices
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality ACS: RR= 1.00; 95% CI 0.84-1.18
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
ACS: RR= 1.06; 95% CI 0.79-1.42
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
ACS: RR= 1.02; 95% CI 0.79-1.31
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis diet quality indices
Overall mortality HEI-2005: HR= 0.40; 95% CI 0.17-0.94 HR= 0.74; 95% CI 0.55-0.99
2987 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association and a decreased risk with overall mortality. Both studies have the same direction HR<1.0 (one study has a large estimate HR<0.5) (no downgrade nor upgrade of the evidence)
AHEI: RR= 0.85; 95% CI 0.63-1.17
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
DQIR: RR= 0.78; 95% CI 0.58-1.07
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
RFS: RR= 1.03; 95% CI 0.74-1.42
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: RR= 0.87; 95% CI 0.64-1.17
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
ACS: RR= 0.93; 95% CI 0.73-1.18
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
HEI-2005: HR= 0.12; 95% CI 0.02-0.99 HR= 0.91; 95% CI 0.60-1.40
2987 (2) + Cohort ++ and downgraded one level; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with a decreased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as only one estimate is ‘statistically significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant). In addition, there will be no upgrade of the evidence based on the other study that found an HR larger than <0.50 as it is only true for one of the studies
AHEI: RR= 1.53; 95% CI 0.98-2.39 RR= 1.07; 95% CI 0.77-1.49
6832 (2) + Cohort ++ and downgraded one level; both studies found a ‘statistically non-significant’ association and a RR> 1.0. However, as the estimate of one study is relatively high (RR=1.53) and the other study RR=1.07, which almost indicates no increased nor decreased risk of cancer-specific mortality, we made the decision to downgrade the evidence with one level for inconsistency of the results
DQIR: RR= 0.81; 95% CI 0.53-1.24
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
RFS: RR= 1.54; 95% CI 0.95-2.47
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: RR= 1.15; 95% CI 0.74-1.77
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: RR= 0.85; 95% CI 0.61-1.19
4103 (1) + Cohort ++ and downgraded one level; data of only 1 study
ACS: RR= 1.44; 95% CI 0.90-2.30
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
HEI-2005: HR= 0.58; 95% CI 0.38-0.87
2317 (1) + Cohort ++ and downgraded one level; data of only 1 study
AHEI: RR= 0.52; 95% CI 0.32-0.83 RR= 0.57; 95% CI 0.42-0.77
6832 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association in the same direction HR>1.0 (one study even has a large estimate HR<0.5). Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
DQIR: RR= 0.85; 95% CI 0.54-1.34
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
RFS: RR= 0.86; 95% CI 0.54-1.37
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: RR= 0.80; 95% CI 0.50-1.26
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 92 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
DASH: RR= 0.72; 95% CI 0.53-0.99
4103 (1) + Cohort ++ and downgraded one level; data of only 1 study
ACS: RR= 0.78; 95% CI 0.56-1.07
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 9: Breast cancer and post-diagnosis low-fat diet RCTs
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Intervention versus control HR= 0.76; 95% CI 0.60-0.98
2437 (1) +++ RCT ++++ and downgraded by one level; data of only 1 study (even though it is a RCT with a low risk of bias)
Overall mortality Intervention versus control HR= 0.89; 95% CI 0.65-1.21 HR= 0.91; 95% CI 0.72-1.15
5525 (2) +++ RCT ++++ and downgraded by one level; although both studies present a HR<1 for overall mortality, both estimates are statistically non-significant. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as both estimates are ‘statistically non-significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Table 10: Breast cancer and pre-diagnosis prudent/healthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Prudent diet: HR= 0.71; 95% CI 0.48-1.06
2522 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Prudent diet: Not shown HR= 0.87; 95% CI 0.61-1.23
5141 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Cancer-specific mortality
Prudent diet: Not shown HR= 0.89; 95% CI 0.59-1.35
5141 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Death from other causes
Prudent diet: Not shown HR= 0.81; 95% CI 0.40-1.61
5141 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Breast cancer and post-diagnosis prudent/healthy diet
Cancer recurrence Prudent diet: HR= 0.95; 95% CI 0.63-1.43
1901 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Prudent diet: RR= 0.78; 95% CI 0.54-1.12 HR= 0.57; 95% CI 0.36-0.90
4520 (2) + Cohort ++ and downgraded one level; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with a decreased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as only one estimate is ‘statistically significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Cancer-specific mortality
Prudent diet: RR= 1.07; 95% CI 0.66-1.73 HR= 0.79; 95% CI 0.43-1.43
4520 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade one level for consistency lack of agreement between studies)
Death from other causes
Prudent diet: RR= 0.54; 95% CI 0.31-0.95 HR= 0.35; 95% CI 0.17-0.73
4520 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association with a decreased risk of death from other cause in the same direction HR<1.0 (one study even has a large estimate HR<0.5). Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
Page 93 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 11: Breast cancer and pre-diagnosis Western/unhealthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Western diet: HR= 0.91; 95% CI 0.61-1.36
2522 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Western diet: RR= 1.40; 95% CI 0.93-2.09 HR= 1.34; 95% CI 0.93-1.94
5141 (2) + Cohort ++ and downgraded one level; both studies found a ‘statistically non-significant’ association. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as both estimates are ‘statistically non-significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Cancer-specific mortality
Western diet: RR= 1.01; 95% CI 0.59-1.72 HR= 0.99; 95% CI 0.64-1.52
5141 (2) + Cohort ++ and downgraded one level; although both studies found a ‘statistically non-significant’ association and indicate no increase nor decrease with cancer-specific mortality, there is no reason for downgrading the evidence
Death from other causes
Western diet: RR= 1.95; 95% CI 1.06-3.60 HR= 3.69; 95% CI 1.66-8.17
5141 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association with an increased risk of death from other cause in the same direction HR>1.0 (one study even has a large estimate HR>2.0). Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies, even though the confidence intervals are broad
Breast cancer and post-diagnosis Western/unhealthy diet
Cancer recurrence Western diet: HR= 0.98; 95% CI 0.62-1.54
1901 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Western diet: RR= 1.53; 95% CI 1.03-2.29 HR= 1.76; 95% CI 1.10-2.81
4520 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association with an increased risk of death from other cause in the same direction HR>1.0. Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
Cancer-specific mortality
Western diet: RR= 1.01; 95% CI 0.60-1.70 HR= 1.20; 95% CI 0.62-2.32
4520 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR=1.0 (downgrade for consistency lack of agreement between studies)
Death from other causes
Western diet: RR= 2.09; 95% CI 1.30-3.36 HR= 2.15; 95% CI 0.97-4.77
4520 (2) + Cohort ++ and downgraded one level; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with a decreased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as only one estimate is ‘statistically significant’ and only two studies have investigated the association, we decided to downgrade the evidence despite both estimates are in the same direction, are large (HR>2), and the overlap in confidence intervals of both studies – although the 95% CI are wide (downgrade for effect on whether estimates are significant)
Page 94 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 12: Breast cancer and pre-diagnosis fruit and vegetables
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total vegetables: RR= 0.98; 95% CI 0.62-1.53 HR= 0.57; 95% CI 0.35-0.94 HR= 1.09; 95% CI 0.80-1.48
4673 (3) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR= 1.0 (almost) downgrade for consistency lack of agreement between studies)
Total fruit: HR= 0.63; 95% CI 0.38-1.05 HR= 0.84; 95% CI 0.61-1.16
3169 (2) + Cohort ++ and downgraded one level; both studies found a ‘statistically non-significant’ association. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as both estimates are ‘statistically non-significant’ and only two studies have investigated the association, we decided to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Total fruit + vegetables: HR= 1.27; 95% CI 1.00–1.61 (low versus high intake!) RR= 1.06; 95% CI 0.85-1.33 (high versus low intake!)
5905 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Cancer-specific mortality
Total vegetables: Not shown HR= 1.01; 95% CI 0.70-1.46
4157 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Total fruit: HR= 0.86; 95% CI 0.59-1.25
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit + vegetables: HR= 1.26; 95% CI 0.96–1.64 (low versus high intake!) RR= 1.00; 95% CI 0.66-1.50
5905 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR=1 (downgrade for consistency lack of agreement between studies)
Death from other causes
Total fruit + vegetables: RR= 1.11; 95% CI 0.81-1.52
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis fruit and vegetables
Cancer recurrence Cruciferous vegetables: HR= 1.10; 95% CI 0.95-1.28
11390 (1) + Cohort ++ and downgraded one level; data of only 1 study although large sample because participants of 4 cohort studies combined
Overall mortality Total vegetables: RR= 0.81; 95% CI 0.59–1.11 HR= 1.44; 95% CI 0.91-2.27
6423 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Cruciferous vegetables: HR= 1.02; 95% CI 0.80-1.30 HR= 0.99; 95% CI 0.86-1.13
15831 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Total fruit: HR= 1.38; 95% CI 0.88-2.17
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit + vegetables: RR= 1.03; 95% CI 0.80-1.33
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total vegetables: Not shown HR= 0.96; 95% CI 0.38-2.45
6423 (2) + Cohort ++ and downgraded one level; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Cruciferous vegetables: HR= 0.95; 95% CI 0.59-1.54
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit: HR= 1.39; 95% CI 0.64-2.99
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit + vegetables: RR= 1.31; 95% CI 0.83-2.06
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Total fruit + vegetables: RR= 0.93; 95% CI 0.65-1.34
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 13: Breast cancer and pre-diagnosis grain foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Bread: HR= 1.31; 95% CI 0.93-1.83
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Bread: HR= 1.10; 95% CI 0.74-1.63
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 95 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 14: Breast cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Poultry RR= 0.60; 95% CI 0.39-0.92
1504 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish RR= 0.94; 95% CI 0.62-1.43
1504 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red meat: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Red and processed meat: RR= 0.88; 95% CI 0.73-1.06
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sunflower/pumpkinseeds: HR= 0.87; 95% CI 0.66-1.15
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sesame/flaxseeds: HR= 0.90; 95% CI 0.68-1.19
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Poultry: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Fish: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Red meat: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Red and processed meat: RR= 1.10; 95% CI 0.80-1.52
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sunflower/pumpkinseeds: HR= 1.12; 95% CI 0.79-1.57
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sesame/flaxseeds: HR= 1.21; 95% CI 0.87-1.68
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Red and processed meat: RR= 0.81; 95% CI 0.62-1.07
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis protein foods
Cancer recurrence Red meat: RR= 1.12; 95% CI 0.66-1.89
472 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Red meat: RR= 1.06; 95% CI 0.76–1.49
1982 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total meat (poultry, fish, beef, and processed meat): HR= 1.12; 95% CI 0.83-1.51
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.64; 95% CI 0.49-0.84
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Red meat: RR= 2.60; 95% CI 0.96-7.03 Not shown
2454 (2)
+ Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Poultry: Not show
1982 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Fish: Not shown
1982 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Total meat (poultry, fish, beef, and processed meat): HR= 0.89; 95% CI 0.50-1.60
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.88; 95% CI 0.54-1.43
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Red and processed meat: RR= 0.57; 95% CI 0.39-0.82
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 96 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 15: Breast cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy RR= 0.71; 95% CI 0.44-1.14
1504 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis dairy and alternatives
Cancer recurrence
Total dairy: HR= 1.13; 95% CI 0.83-1.54
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
Low fat dairy: HR= 1.01; 95% CI 0.78-1.32
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
High fat dairy: HR= 1.22; 95% CI 0.92-1.55
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Total dairy: RR= 0.72; 95% CI 0.52-1.00 HR= 1.39; 95% CI 1.02-1.90 HR= 1.18; 95% CI 0.90-1.54
3875 (3) + Cohort ++ and downgraded one level; 2 studies found no ‘statistically significant’ association whilst one study found a ‘statistically significant’ association with an increased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. Nevertheless, we will downgrade the evidence with two HRs in the same direction and the third with in an opposite direction (downgrade for consistency lack of agreement between studies)
Low fat dairy: HR= 1.05; 95% CI 0.80-1.36
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
High fat dairy: HR= 1.64; 95% CI 1.24-2.17
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: Not shown HR= 0.94; 95% CI 0.56-1.59
1982 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Table 16: Breast cancer and post-diagnosis oils and spreads
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Butter/margarine/lard: RR= 1.30; 95% CI 1.03-1.64
472 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Butter/margarine/lard: RR= 1.03; 95% CI 0.61-1.76
472 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 97 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Laryngeal cancer
Table 17: Laryngeal cancer and pre-diagnosis fruit and vegetables
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Citrus fruits: HR= 0.76; 95% 0.49-1.19
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Other fruits: HR= 0.65; 95% CI 0.39-1.07
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Vegetables: HR= 0.57; 95% CI 0.35-0.94
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 18: Laryngeal cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Meat: HR= 0.50; 95% CI 0.30-0.83
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Poultry: HR= 0.90; 95% CI 0.55-1.46
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish: HR= 0.91; 95% CI 0.59-1.39
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Eggs: HR= 1.22; 95% CI 0.74-2.00
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 19: Laryngeal cancer and pre-diagnosis grain foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Bread: HR= 0.54; 95% CI 0.32-0.90
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Pasta: HR= 1.25; 95% CI 0.76-2.04
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Potatoes: HR= 1.02; 95% CI 0.64-1.64
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 20: Laryngeal cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Milk: HR= 1.58; 95% CI 0.99-2.55
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cheese: HR= 0.70; 95% CI 0.44-1.12
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 21: Laryngeal cancer and pre-diagnosis oils and spreads
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Butter: HR= 1.11; 95% CI 0.69-1.80
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Olive oil: HR= 0.71; 95% CI 0.44-1.16
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 98 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Non-Hodgkin Lymphoma
Table 22: Non-Hodgkin Lymphoma and pre-diagnosis fruit and vegetables
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total fruit and vegetables: HR= 0.68; 95% CI 0.49-0.95
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit: HR= 0.91; 95% CI 0.70-1.18 HR= 1.03; 95% CI 0.90-1.19
2907 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies). Additionally, one study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Total vegetables: HR= 0.58; 95% CI 0.38-0.89 HR= 0.98; 95% CI 0.85-1.12
2907 (2) + Cohort ++ and downgraded one level; one study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Cruciferous vegetables: HR= 0.91; 95% CI 0.67–1.24
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bean vegetables: HR= 1.14; 95% CI 0.85-1.54
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Green leafy vegetables: HR= 0.71; 95% CI 0.51-0.98
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red vegetables: HR= 1.03; 95% CI 0.76-1.38
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Yellow vegetables: HR= 0.93; 95% CI 0.69-1.25
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Citrus fruits: HR= 0.73; 95% CI 0.54-0.99
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality Total fruit and vegetables: HR= 0.70; 95% CI 0.45-1.10
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit: HR= 1.04; 95% CI 0.74-1.45 HR= 1.04; 95% CI 0.88-1.24
2907 (2) + Cohort ++ and downgraded one level; one study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Total vegetables: HR= 0.58; 95% CI 0.33-1.03 HR= 0.98; 95% CI 0.83-1.16
2907 (2) + Cohort ++ and downgraded one level; one study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Cruciferous vegetables: HR= 0.75; 95% CI 0.49–1.14
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bean vegetables: HR= 1.05; 95% CI 0.71-1.55
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Green leafy vegetables: HR= 0.82; 95% CI 0.54-1.23
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red vegetables: HR= 1.11; 95% CI 0.76-1.62
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Yellow vegetables: HR= 1.11; 95% CI 0.77-1.61
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Citrus fruits: HR= 0.81; 95% CI 0.54-1.20
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 23: Non-Hodgkin Lymphoma and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Red meat: HR= 1.00; 95% CI 0.87-1.15
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish: HR= 0.90; 95% CI 0.78-1.03
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Legumes: HR= 0.88; 95% CI 0.76-1.01
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality Red meat: HR= 0.95; 95% CI 0.81-1.13
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish: HR= 0.91; 95% CI 0.76-1.08
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Legumes: HR= 0.86; 95% CI 0.72-1.02
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 99 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 24: Non-Hodgkin Lymphoma and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy: HR= 1.14; 95% CI 1.00-1.31
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality Total dairy: HR= 1.16; 95% CI 0.98-1.37
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 100 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Prostate cancer
Table 25: Prostate cancer and post-diagnosis diet quality indices
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Med diet score: HR= 0.78; 95% CI 0.67-0.90
4538 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Med diet score: HR= 1.01; 95% CI 0.75-1.38
4538 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 26: Prostate cancer and post-diagnosis prudent/healthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Prudent diet: RR= 0.64; 95% CI 0.44-0.93
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Prudent diet: RR= 0.46; 95% CI 0.17-1.24
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 27: Prostate cancer and post-diagnosis Western/unhealthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Western diet: RR= 1.67; 95% CI 1.16-2.42
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Western diet: RR= 2.53; 95% CI 1.00-6.42
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 28: Prostate cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer-specific mortality
Fish: HR= 0.52; 95% CI 0.30-0.91
2161 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 29: Prostate cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy: HR = 1.76; 95% CI 1.21-2.55
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
High-fat dairy: HR= 1.22; 95% CI 1.08-1.38
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Low-fat dairy: HR= 1.17; 95% CI 1.05-1.29
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: HR = 2.41; 95% CI 0.96-6.02
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
High-fat dairy: HR= 1.30; 95% CI 0.97-1.73
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Low-fat dairy: HR= 1.16; 95% CI 0.88-1.53
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 101 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
PRISMAPRISMAPRISMAPRISMA ChecklistChecklistChecklistChecklist
Section/topic # Checklist item Reported on page #
TITLE
Title 1 Identify the report as a systematic review, meta-analysis, or both. 1
ABSTRACT
Structured summary 2 Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number.
2
INTRODUCTION
Rationale 3 Describe the rationale for the review in the context of what is already known. 4
Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).
5
METHODS
Protocol and registration 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.
9
Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered,
language, publication status) used as criteria for eligibility, giving rationale. 5-8
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.
5
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.
37
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable,
included in the meta-analysis). 6
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.
5
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.
5-8
Risk of bias in individual studies
12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.
8-9
Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means). 10
Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency
(e.g., I2) for each meta-analysis.
9-10
Page 102 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
PRISMAPRISMAPRISMAPRISMA ChecklistChecklistChecklistChecklist
Page 1 of 2
Section/topic # Checklist item Reported on page #
Risk of bias across studies 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within studies).
8-9
Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-specified.
NA
RESULTS
Study selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.
9 and flowchart in supplemental file
Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations.
Supplemental file
Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12). Data on risk of bias of each study can be obtained on request
Results of individual studies 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.
Supplemental file
Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency. NA
Risk of bias across studies 22 Present results of any assessment of risk of bias across studies (see Item 15). Supplemental file
Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]).
NA
DISCUSSION
Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).
21-23
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).
24
Page 103 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
PRISMAPRISMAPRISMAPRISMA ChecklistChecklistChecklistChecklist
Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research.
24-25
FUNDING
Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for the systematic review.
25-26
From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097
For more information, visit: www.prisma-statement.org.
Page 2 of 2
Page 104 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
The impact of dietary patterns and the main food groups on mortality and recurrence in cancer survivors: a systematic
review of current epidemiological literature
Journal: BMJ Open
Manuscript ID bmjopen-2016-014530.R3
Article Type: Research
Date Submitted by the Author: 07-Sep-2017
Complete List of Authors: Jochems, Sylvia; Maastricht University, NUTRIM School of Nutrition, Metabolism and Toxicology; University of Birmingham, Cancer and Genomic Sciences
Van Osch, Frits; Maastricht University, NUTRIM School of Nutrition, Metabolism and Toxicology; University of Birmingham, Cancer and Genomic Sciences Bryan, Richard; University of Birmingham, Cancer and Genomic Sciences Wesselius, Anke; Maastricht University, NUTRIM School of Nutrition, Metabolism and Toxicology van Schooten, Frederik; Maastricht University, NUTRIM School of Nutrition, Metabolism and Toxicology Cheng, Kar Keung; University of Birmingham, Public Health and Epidemiology Zeegers, Maurice; University of Maastricht, NUTRIM School of Nutrition, Metabolism and Toxicology; Maastricht University, CAPHRI Care and Public
Health Research Institute
<b>Primary Subject Heading</b>:
Epidemiology
Secondary Subject Heading: Oncology, Public health, Epidemiology
Keywords: cancer survivors, mortality, cancer recurrence, food, dietary pattern, diet
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open on D
ecember 7, 2020 by guest. P
rotected by copyright.http://bm
jopen.bmj.com
/B
MJ O
pen: first published as 10.1136/bmjopen-2016-014530 on 19 F
ebruary 2018. Dow
nloaded from
For peer review only
The impact of dietary patterns and the main food groups on 1
mortality and recurrence in cancer survivors: 2
a systematic review of current epidemiological literature 3
4
Sylvia H.J. Jochems (1,2), Frits H.M. van Osch (1,2), Richard T. Bryan (1), Anke Wesselius 5
(2), Frederik J. van Schooten (2), K.K. Cheng (3), Maurice P. Zeegers (2,4) 6
7
(1) Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, 8
United Kingdom 9
(2) NUTRIM School of Nutrition and Translational Research in Metabolism, Maastricht 10
University, Maastricht, The Netherlands 11
(3) Institute of Applied Health Research, Public Health, Epidemiology and Biostatistics, 12
University of Birmingham, Birmingham, United Kingdom 13
(4) CAPHRI School for Public Health and Primary Care, Maastricht University, 14
Maastricht, The Netherlands 15
16
17
Corresponding author contact information: 18
Sylvia H.J. Jochems 19
Institute of Cancer and Genomic Sciences, University of Birmingham, 21
B15 2TT Birmingham, United Kingdom 22
23
Page 1 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
2
ABSTRACT 24
Objective: To determine whether there is an association between dietary patterns/indices and 25
foods from the main food groups (highest versus lowest intakes) prior to or after cancer 26
diagnosis and mortality and cancer recurrence in cancer survivors. 27
Participants: Survivors of common cancers with a 10-year survival rate of 50% or more: 28
bladder, bowel, breast, cervical, kidney, laryngeal, prostate, testicular, uterine cancer, 29
malignant melanoma, and (non-)Hodgkin lymphoma. 30
Outcome measures: Mortality (overall, cancer-specific, from other causes) and cancer 31
recurrence. 32
Information sources: PubMed, Embase and the Cochrane Library were searched from 33
inception to April 2017. Additional studies were identified by searching reference lists. Two 34
authors independently screened titles and abstracts, assessed study quality, and extracted the 35
data. 36
Results: A total of 38 studies were included. The risk of bias was rated low for the included 37
RCTs and moderate for the cohort studies. The quality of evidence was assessed with the 38
GRADE approach and was rated moderate (RCTs), and (very)low (cohort studies). Reducing 39
the amount of fat after diagnosis appears to decrease the risk of breast cancer recurrence. 40
Adherence to a high-quality diet and prudent diet after diagnosis appears to decrease the risk 41
of death from other causes (and overall mortality for high-quality diet) in breast cancer 42
survivors. Adherence to a Western diet, before and after diagnosis, appears to increase the 43
risk of overall mortality and death from other causes amongst breast cancer survivors. 44
Evidence from studies amongst other cancer survivors were too limited or could not be 45
identified. 46
Conclusion: For many cancer survivors, there is little evidence to date to indicate that 47
particular dietary behaviours influence outcomes with regard to recurrence and mortality. 48
Page 2 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
3
Notwithstanding, limited evidence suggests that a low-fat diet, a high-quality diet, and a 49
prudent diet are beneficial for breast cancer survivors, whilst a Western diet is detrimental for 50
breast cancer survivors. 51
52
Strengths and limitations 53
- Dietary patterns/indices and whole foods reflect the complexity of dietary intake and 54
capture synergistic relationships between various dietary constituents 55
- Most studies investigating dietary patterns/indices and foods before diagnosis do not 56
consider potential modifications in dietary intake after cancer diagnosis 57
- Cohort studies provide weaker empirical evidence than RCTs for examining 58
relationships between dietary exposure and mortality and cancer recurrence 59
60
Page 3 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4
INTRODUCTION 61
As cancer survival rates continue to improve, there is an increased need to identify 62
modifiable lifestyle factors amongst cancer survivors in order to improve long-term health. 63
Adherence to a diet rich in fruit and vegetables could decrease the risk of several types of 64
cancer and increase overall life expectancy[1,2]. The suggestion that epigenetic aberrations 65
occurring in cancer could be altered by nutrients makes it plausible that dietary changes after 66
successful cancer treatment could improve prognosis[3,4]. 67
Although cancer survivors are responsive to health promotion[5,6], a recent study has 68
indicated that survivors had poorer diets than individuals without cancer[7]. One possible 69
explanation could be the difficulty for cancer survivors in adopting a healthier diet without 70
clear evidence that it will improve their survival[8]. While guidelines have been well 71
documented for the prevention of cancer, many uncertainties remain for nutrition after cancer 72
treatment[9]. A systematic review, as part of the Continuous Update Project (CUP) of the 73
World Cancer Research Fund International, was published on diet, nutrition, physical activity 74
and survival in breast cancer survivors[10]. The independent panel of scientists concluded 75
that the evidence to date was not strong enough to make specific recommendations for breast 76
cancer survivors[11]. A recent meta-analysis investigating the role of diet on overall 77
mortality and recurrence among cancer survivors concluded that adherence to a Western diet 78
is positively associated with overall mortality, and a high-quality diet / healthy dietary pattern 79
is inversely associated with overall mortality amongst all cancer survivors[12]. 80
In the setting of survivors of cancers with a 10-year survival rate ≥50%, this 81
systematic review provides a structured overview of RCTs and cohort studies addressing the 82
relationship between adherence to dietary patterns/indices and intake of foods from the main 83
food groups, prior to or after cancer diagnosis, and health outcomes including cancer 84
recurrence, cancer-specific mortality, overall mortality, and death from other causes than 85
Page 4 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
5
cancer. Given that these survivors have the potential for long-term survival, they may be most 86
likely to benefit from dietary changes to prevent or delay cancer recurrence and improve 87
survival. Notwithstanding, many of these survivors will die from other causes such as 88
cardiovascular disease – even if the dietary exposures identified will not help the investigated 89
outcomes, it could be desirable to follow a diet that could help reduce other conditions. 90
91
METHODS 92
Search strategy 93
From inception up to April 2017, Pubmed, Embase and the Cochrane Library were 94
searched to find English language articles of original and published randomized trials and 95
observational studies to answer the following research question: does adherence to/intake of 96
dietary patterns/indices and foods (highest versus lowest adherence/intake) prior to or after 97
cancer diagnosis, increase or decrease the risk of mortality and cancer recurrence amongst 98
cancer survivors of common cancers with a 10-year survival rate of 50% or more? This 99
research question was developed using the PICO framework (supporting data review protocol 100
File S1). Search strategies included search terms related to dietary patterns, dietary indices, 101
diet quality, foods from the main food groups, and outcomes of interest, including overall 102
mortality, cancer-specific mortality, death from other causes, and recurrence of cancer. 103
Additionally, studies were identified by searching reference lists of relevant studies, literature 104
reviews and meta-analyses. After the search was completed, articles were screened and 105
selected independently based on the title and abstract by two of the authors (SJ and FvO). 106
The data extraction was performed independently by the same authors (SJ and FvO) and any 107
disagreements about study inclusion were resolved through consensus or a third party. 108
109
Inclusion and exclusion criteria 110
Page 5 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
6
Eligibility criteria included adult survivors of cancer (no sex or age restriction) who 111
were defined as individuals who had been diagnosed with a primary cancer, received cancer 112
therapy, and were in remission or had recovered completely from cancer. Considered cancer 113
types were the commonly-occurring cancers in the Western world with a 10-year net survival 114
of at least 50% (based on cancer diagnoses of men and women during 2010-2011 in England 115
and Wales)[13]. These include in decreasing order of net survival: testicular cancer (98%), 116
malignant melanoma (MM) (89%), prostate cancer (84%), Hodgkin lymphoma (HL) (80%), 117
breast cancer (78%), uterine cancer (77%), non-Hodgkin lymphoma (NHL) (63%), cervical 118
cancer (63%), laryngeal cancer (62%), bowel cancer (57% including both colon and rectal 119
cancer), bladder cancer (50%), and kidney cancer (50%). In the statistical analyses 120
adjustments had to be made for at least age and disease stage at baseline and, where possible, 121
for cancer treatment. Excluded papers did not state hazard ratios (HRs) or relative risks 122
(RRs), nor 95% confidence intervals (95% CI); neither did they provide information on 123
disease stage or tumour grade or therapy. Additionally, studies were excluded when outcomes 124
were combined, such as mortality and cancer progression, mortality and diagnosed 125
metastasis, or where prostate cancer recurrence was determined by a rising PSA level 126
only. 127
128
Dietary exposure 129
Dietary patterns/indices that were considered were assessed by index-based methods 130
and data-driven approaches, such as principal component analysis (factor analysis) and 131
cluster analysis[14]. The following diet scores were considered: the Healthy Eating Index 132
2005 (HEI-2005)[15,16], the alternate Healthy Eating Index 2010 (AHEI)[17], the World 133
Cancer Research Fund and the American Institute for Cancer Research (WCRF/AIRC) 134
dietary guidelines adherence score[18] and the American Cancer Society diet-specific 135
Page 6 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7
recommendations for cancer prevention (ACS)[19], the recommended food score (RFS)[20], 136
the Diet Quality Index-Revised (DQIR)[21], the Dietary Approaches to Stop Hypertension 137
diet (DASH) diet[22], the Healthy Nordic Food Index (HNFI)[23], and the alternate 138
Mediterranean diet (aMed)[24,25]; empirical patterns reviewed included a low-fat diet, a 139
prudent/healthy diet, and a Western/unhealthy diet. The HEI-2005 was developed by the US 140
Department of Agriculture and targets foods that could possibly reduce the risk of chronic 141
diseases and include fruits, vegetables, fibre, soy, nuts, ratio white and red meat, alcohol, 142
trans fat, saturated fat ratio, and multivitamin use[15]. Five years later, the AHEI was 143
introduced, which differs from the HEI-2005 by distinguishing quality within food groups 144
and recognizing health benefits of unsaturated oils[26]. The RFS includes the foods fruits, 145
vegetables, whole grains, dairy and protein foods low in fat. Diet diversity and moderation 146
was addressed by the DQIR and included fruits, vegetables, cholesterol, total fat, saturated 147
fat, iron, calcium, and fat/sugar moderation. The aMed is based on the original Mediterranean 148
diet score and includes fruits, vegetables, legumes, nuts, whole grains, red and processed 149
meat, moderate alcohol, and the ratio of monounsaturated and saturated fat[27,28]. In 150
addition, whole foods of the main food groups (UK Eatwell Guide)[29] were considered. The 151
composition of the investigated groups was as follows: (I) fruit and vegetables including 152
citrus fruits, stone fruits, soft fruits, fleshy fruits, vine fruits, flower vegetables, leafy 153
vegetables, stem vegetables, fruit vegetables, mushrooms, bulbs and roots; (II) grain foods 154
including potatoes, bread, rice, pasta and cereal; (III) protein foods including unprocessed 155
meat, red meat, poultry, fish, eggs, tofu, nuts, seeds, pulses, legumes and beans; (IV) dairy 156
and alternative products including yoghurt, milk, cheese; (V) oils and spreads including 157
vegetable oils, spreads. Although processed (red) meats are not included in the main food 158
groups recommended by the UK Eatwell Guide, lean red meats (rich in protein, iron, zinc, 159
selenium and B vitamins) can be part of a healthy diet. Studies that made no distinction 160
Page 7 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
8
between (lean) red meats and processed meats in their estimates, were still included in this 161
systematic review – they will however, be interpreted with caution. Information on intake of 162
food was obtained before or after cancer diagnosis with food records, food frequency 163
questionnaires (FFQ) (self-administered or via an interview), or twenty-four-hour recalls, and 164
expressed in servings or (milli)grams per day/week/month. No restrictions were made for 165
time of follow-up, and timing or frequency of dietary intake. 166
167
Mortality and cancer recurrence 168
Considered endpoints were overall mortality, cancer-specific mortality, death from 169
other causes, and cancer recurrence. The cause of death was confirmed via death certificates 170
or the National Death Index in each of the studies. Cancer recurrence was defined as a new 171
occurrence of cancer after a period of time during which the cancer could not be detected at 172
the same or at a different site to the initial primary tumour. Cancer recurrence had to be 173
confirmed by a biopsy, scan, medical record, cancer registry, or treating physician. 174
175
Assessment risk of bias and level of quality 176
The Cochrane Collaboration risk of bias assessment tools were used for appraisal of 177
RCTs [30]and cohort studies[31]. For the two RCTs the RoB 2.0 tool (a revised tool for risk 178
of bias in randomized trials) was used to evaluate the risk of bias. Cohort studies were 179
appraised with an adjusted version of the ROBINS-I tool[30,31]. Levels of quality were 180
determined with the GRADE approach[32]; evidence from RCTs or multiple double-181
upgraded observational studies were considered as high quality, downgraded RCTs or 182
upgraded observational studies were considered as moderate quality, double-downgraded 183
RCTs or observational studies were considered as low quality, and triple-downgraded RCTs, 184
downgraded observational studies or case series/case reports were considered as very low 185
Page 8 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
9
quality[32]. Factors reducing the quality of the evidence include limitations in study design, 186
inconsistency between study results, indirectness of evidence, imprecision, and publication 187
bias. Factors increasing the quality of the evidence include a large magnitude of effect, 188
correction for all plausible confounding that could reduce the demonstrated effect or increase 189
the effect if no effect was observed, and presence of a dose-response gradient. For 190
observational studies, this could intent controlling for key knows risk factors and 191
confounders. GRADE separates the process of assessing the quality of evidence from making 192
recommendations. To determine whether evidence for an association between dietary 193
patterns/indices or foods and mortality or cancer recurrence amongst cancer survivors was 194
conclusive, the risk of bias and levels of quality were considered. 195
196
RESULTS 197
The search resulted in 2883 citations after removal of duplicates. After screening the 198
titles and abstracts, 95 full-text articles were assessed for eligibility - a total of 2 RCTs and 36 199
cohort studies were included in this systematic review. No studies could be identified for 200
cervical, kidney, testicular, uterine cancer, HL or MM survivors. Dietary patterns/indices 201
could be identified for bowel, breast, prostate cancer, and NHL. Whole foods from the main 202
food groups could be identified for bladder, bowel, breast, laryngeal, prostate cancer and 203
NHL survivors. 204
The protocol used for this systematic review is available in the supporting data (File 205
S1). A detailed search strategy is provided in Table 1 and the search was adapted accordingly 206
for the individual cancers and databases (File S1). The review was written according to the 207
PRISMA guidelines[33]. A summary of the number of studies for pre-diagnosis dietary 208
patterns/indices (Table 2) and post-diagnosis dietary patterns/indices (Table 3) and mortality 209
and cancer recurrence is provided. Additionally, tables with the number of studies for pre-210
Page 9 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
10
diagnosis food intake (Table 4) and post-diagnosis food intake (Table 5) and mortality and 211
cancer recurrence is given. The study characteristics including the HRs/RRs with their 212
corresponding 95% CI are provided in the supporting data (File S2). 213
Templates of the RoB 2.0 and ROBINS-I tools can be found in the supporting data 214
(File S1). Results for the assessment of the risk of bias for each individual RCT (RoB 2.0) 215
and cohort study (ROBINS-I) will be provided on request. Briefly, the included RCTs 216
investigating a low-fat diet and mortality amongst breast cancer survivors indicated a low risk 217
of bias[34]; the included cohort studies all had a moderate risk of bias[35]. 218
An overview of the GRADE ratings with comments can be found in the supporting 219
data (File S3). As the risk of bias was rated ‘low’ and ‘moderate’, there was no reason to 220
downgrade the quality of evidence on this matter. The quality level of the body of evidence 221
of the studies was rated ‘very low’, ‘low’ and ‘moderate’ by two of the authors (SJ and FvO) 222
when applying the grading system developed by the GRADE collaboration[32]. Briefly, the 223
level of evidence for the association between a low-fat diet and bladder cancer recurrence and 224
mortality was downgraded from ‘high’ to ‘moderate’ due to the presence of potential 225
confounding factors in many studies. Evidence for associations between dietary factors and 226
bladder cancer recurrence and mortality from cohort studies could not score higher than ‘low’ 227
level of evidence and was downgraded to 'very low' if inconsistent, indirect or under 228
suspicion of publication bias. 229
230
Bladder cancer 231
A total of 1 cohort study could be identified for bladder cancer survivors regarding 232
fruit and vegetable consumption. The study of Tang et al. investigated pre-diagnosis fruit and 233
vegetable consumption with data from 239 male and female bladder cancer survivors from 234
the Roswell Park Cancer Institute (RPCI) Tumor Registry[36]. After an average of 8-year 235
Page 10 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
11
follow-up, no associations were observed between overall mortality or bladder cancer-236
specific mortality when comparing survivors with the highest intakes of total fruit, total 237
vegetables or other cruciferous vegetables (raw or cooked) with those in the lowest intake 238
group. An association was, however, observed for broccoli intake (≥1 versus <1 serving per 239
month) with overall mortality (broccoli raw HR=0.57; 95% CI 0.39-0.83, broccoli cooked 240
HR=0.67; 95% CI 0.49-0.91) and bladder cancer-specific mortality (broccoli raw HR=0.43; 241
95% CI 0.25-0.74). The intake of other raw and cooked vegetables including cabbage, 242
cauliflower, Brussels sprouts, kale, turnip, collard or mustard greens was not related with 243
mortality[36]. 244
In summary, no conclusive evidence for an association between vegetable and fruit 245
intake and mortality amongst bladder cancer survivors could be provided, as evidence for 246
each exposure and outcome was based on the results of one study only. 247
248
Bowel cancer 249
A total of 12 cohort studies could be identified for bowel cancer survivors. Three 250
observational cohort studies could be identified investigating the role of a pre- and post-251
diagnosis prudent diet on mortality in bowel cancer survivors. Results of the Cancer and 252
Leukemia Group B (CALGB) study indicated no associations between a prudent diet after 253
cancer diagnosis and decreased mortality[37]. However, there was a higher overall mortality 254
amongst these survivors with the highest post-diagnosis intakes of a Western diet in 255
comparison with those in the lowest category (HR=2.32; 95% CI 1.36-3.96)[37]. When 256
comparing participants in the Familial Bowel Cancer Registry (FBCR) with the highest and 257
lowest intakes of a prudent diet before cancer diagnosis, no associations were found with 258
mortality[38]. Besides a prudent diet, two other dietary patterns comparable with a Western 259
diet were identified in this study: a high processed meat pattern and a high sugar pattern diet. 260
Page 11 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
12
No associations were reported for the pattern high in sugar and mortality when comparing the 261
highest to the lowest intake group, whereas a high processed meat pattern was specifically 262
related to increased colon cancer mortality (HR=2.13; 95% CI 1.03-4.43). This relationship 263
between a processed meat pattern and bowel cancer survival was modified by sex[38]. In the 264
Nurses' Health Study (NHS), no associations were observed between adherence to the AHEI, 265
DASH, or AMED score, a prudent diet, or a Western diet after diagnosis and mortality in 266
these bowel cancer survivors[24]. It should be noted, however, that even though there was 267
‘no statistically significant’ result for the role of a post-diagnosis Western diet in this study, 268
the HR was >1 (HR= 1.31; 95% CI 0.89-1.97)[39] as observed in the earlier described study 269
of Meyerhardt et al. (HR= 2.32; 95% CI 1.36-3.96)[37]. Adherence to the HEI diet score was 270
investigated in a large study including 5,727 male and female survivors in the USA and 271
indicated no association between pre-diagnosis adherence to the HEI-2005 score with overall 272
mortality or cancer-specific mortality[40]. Recently, a German study examined adherence to 273
the Modified Mediterranean Diet Score (MMDS) and the Healthy Nordic Food Index (HNFI) 274
and found that post-diagnosis adherence to this MMDS was associated with a decreased risk 275
of overall mortality amongst bowel cancer survivors (HR=0.48; 95% CI 0.32-0.74)[41]. In 276
the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort, data from 277
participants of 10 European countries was analysed on adherence to WCRF/AICR diet scores 278
and intake of total dairy, milk, yoghurt, cheese, red meat, and poultry[42–44]. Pre-diagnosis 279
adherence to this high-quality diet score indicated a decreased risk of overall mortality 280
amongst bowel cancer survivors (HR=0.79; 95% CI 0.65-0.98)[43]. No evidence of an 281
association with mortality was observed for foods from the main food groups, including 282
fruits, vegetables, dairy, or protein foods amongst these bowel cancer survivors[42,44]. The 283
study by Yang et al. indicated a protective association with milk consumption and overall 284
mortality after a diagnosis of bowel cancer (RR=0.72; 95% CI 0.55-0.94)[45]. Additionally, 285
Page 12 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
13
no association could be found for total dairy intake and mortality in this study[45]. Whole 286
gains, another food group investigated in bowel cancer survivors, were not associated with 287
overall mortality amongst 1,119 Danish, Swedish and Norwegian bowel cancer survivors in 288
the HELGA cohort[46]. Carr et al. reported that red and processed meat consumption was not 289
associated with a poorer survival amongst stage I–III bowel cancer survivors in a follow-up 290
study of the Darmkrebs: Chancen der Verhutung durch Screening (DACHS) study[47]. 291
However, it should be noted that the authors investigated red and processed meat combined 292
and they suggest that major changes in the consumption of red meat measured at 5-year 293
follow-up could have influenced survival[47]. The study of McCullough et al. indicated an 294
association with mortality when comparing highest versus lowest pre- and post-diagnosis red 295
and processed meat consumption for overall mortality (RR=1.29; 95% CI 1.05-1.59) and 296
death from other causes than bowel cancer (RR=1.39; 95% CI 1.00-1.92)[48]. It should be 297
noted that the authors combined the consumption of red and processed meat in these 298
estimates, and that there were no associations found for ‘fresh’ meats and mortality[48]. 299
In summary, no conclusive evidence for an association between adherence to a high-300
quality diet, a prudent diet, a Western diet, and the consumption of fruits, vegetables, meats 301
or dairy and mortality in bowel cancer survivors could be provided, as evidence for each 302
exposure and outcome was based on the results of one study only or on inconsistent results. 303
304
Breast cancer 305
A total of 2 RCTs and 16 cohort studies could be identified for breast cancer 306
survivors. Two dietary intervention trials amongst breast cancer survivors met the inclusion 307
criteria[49,50]. The study of Chlebowski et al. aimed to reduce post-diagnosis dietary fat 308
intake to almost one sixth of total energy intake while maintaining nutritional adequacy in 309
Page 13 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
14
women participating in the Women’s Intervention Nutrition Study (WINS)[49]. Breast cancer 310
survivors in the intervention group were informed extensively on maintaining weight based 311
on energy intake, whilst minimum dietary advice on nutrient intake was provided to breast 312
cancer survivors in the control group. Women in the intervention group had a lower dietary 313
fat intake compared to those in the control group, whereas no differences could be observed 314
for a lower energy or higher dietary fibre intake. According to the authors of this RCT, there 315
was no association with overall mortality between women adhering to a low-fat diet and 316
women given minimum dietary advice (HR=0.89; 95% CI 0.65-1.21). However, for relapse 317
events (including local, regional, distant, or ipsilateral breast cancer recurrence or new 318
contralateral breast cancer) the HR of an event in the intervention group compared to the 319
control group was HR=0.76; 95% CI 0.60-0.98. This could indicate that a lifestyle 320
intervention reducing dietary fat intake, could improve relapse-free survival of breast cancer 321
survivors[49]. In the Women’s Healthy Eating and Living (WHEL) study breast cancer 322
survivors in the intervention group received telephone counselling with additional cooking 323
classes and brochures to support adherence to a post-diagnosis diet high in fruit (3 324
servings/day), high in vegetables (5 servings/day and 16oz of vegetable juice), high in fibre 325
(30g/day), and low in fat (15-20% of energy intake from fat)[50]. In the control group, breast 326
cancer survivors received written advice to eat at least 5 portions of fruit and vegetables each 327
day (5-a-day advice). Differences between the former and latter groups in mean consumption 328
of vegetables (+65%), fruit (+25%), fibre (+30%), and energy from fat (-13%) were observed 329
at 4 years. The authors of this trial reported that no associations were observed for overall 330
survival when comparing women in the intervention group with those in the control group 331
(HR=0.91; 95% CI 0.72-1.15)[50]. Although the results for overall mortality in the trials 332
were statistically non-significant, the HRs of both studies were <1 (HR=0.89; 95% CI 0.65-333
1.21 [49] and HR=0.91; 95% CI 0.72-1.15 [50]). 334
Page 14 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
15
Post-diagnosis dietary indices were examined in the Health, Eating, Activity, and Lifestyle 335
(HEAL) study[51], Women’s Health Initiative’s Dietary Modification Trial and 336
Observational Study (WHI)[52], Nurses’ Health Study (NHS)[25,53] and Cancer Prevention 337
Study II Nutrition Cohort (CPS-II)[54]. McCullough et al. demonstrated that pre- and post-338
diagnosis adherence to the ACS diet amongst breast cancer survivors in the CPS-II cohort 339
was not association with breast cancer-specific mortality[54]. It should be noted, however, 340
that an inverse association was observed for the continuous post-diagnosis diet scores and 341
other causes of death RR=0.88; 95% CI 0.79-0.99)[54]. Whilst no associations were found 342
between pre- and post-diagnosis fruit and vegetable intake and the intake of whole grains, 343
detrimental associations were found with post-diagnosis red and processed meat consumption 344
and overall mortality and death from other causes (respectively RR=0.64; 95% CI 0.49-0.84 345
and RR=0.57; 95% CI 0.39-0.82)[54]. In the NHS, post-diagnosis dietary DQIR, RFS, aMed, 346
AHEI, and DASH scores were not associated with overall mortality or breast cancer-specific 347
mortality[25,53]. Closer adherence to DASH and AHEI were, however, related to a lower 348
risk of death from other causes than breast cancer (respectively RR=0.72; 95% CI 0.53-0.99 349
and RR=0.57; 95% CI 0.42-0.77)[53]. George et al. examined post-diagnosis adherence to 350
the HEI-2005 scores and concluded that there was an association with a decreased risk of 351
mortality (overall mortality HR=0.40; 95% CI 0.17-0.94 and breast cancer-specific mortality 352
HR=0.12; 95% CI 0.02-0.99)[51]. In the WHI cohort, results of post-diagnosis adherence to 353
the HEI-2005 scores indicated that women who consumed better quality diets had a 26% 354
lower risk of overall mortality (HR=0.74; 95% CI 0.55-0.99) and a 42% lower risk of death 355
from non-breast cancer related death (HR=0.58; 95% CI 0.38-0.87)[52]. Even though the 356
result for breast cancer-specific mortality and adherence to the HEI-2005 score in this study 357
was statistically non-significant (HR= 0.91; 95% CI 0.60-1.40), the HR is <1, as observed for 358
women in the HEAL study regarding cancer-specific mortality [51]. Results of the NHS 359
Page 15 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
16
study indicated that a post-diagnosis prudent diet was not associated with overall or breast 360
cancer-specific mortality whilst death from other causes was associated with a prudent diet 361
after diagnosis when comparing breast cancer survivors of the highest and lowest intake 362
group (HR=0.54; 95% CI 0.31-0.95)[55]- adherence to a prudent diet before diagnosis was 363
not associated with mortality amongst breast cancer survivors in the NHS[55]. Both pre- and 364
post-diagnosis adherence to a Western diet was associated with death from other causes 365
(respectively RR=1.95; 95% CI 1.06-3.60 and RR=2.31; 95% CI 1.23-4.32)[55]. The study of 366
Kwan et al. concludes no associations between adherence to a pre- or post-diagnosis Western 367
diet and overall mortality, breast cancer-specific mortality or cancer recurrence[56]. The HR 368
for a Western diet and death from other causes was, however, >1 (HR=2.15; 95% CI 0.97-369
4.77)[56], and therefore in agreement with the HR for a Western diet and death from other 370
causes observed in the study of Kroenke et al. (RR= 2.09; 95% CI 1.30-3.36)[55]. In the Life 371
After Cancer Epidemiology (LACE) study, post-diagnosis adherence to a prudent diet in 372
women with early-stage breast cancer resulted in a decreased risk of death from other causes 373
(HR=0.35; 95% CI 0.17-0.73) and overall mortality (HR=0.57; 95% CI 0.36-0.90)[56]. The 374
study of Vrieling et al. investigated associations between a ‘healthy’ and ‘unhealthy’ pre-375
diagnosis dietary pattern and mortality in German breast cancer survivors in the Mammary 376
carcinoma Risk factor Investigation (MARIE) study[57]. The characteristics of the defined 377
healthy diet are comparable with a prudent diet; nevertheless, no associations between the 378
highest and lowest intake of this defined ‘healthy’ diet before cancer diagnosis and mortality 379
in breast cancer survivors were observed. However, the results did indicate that a higher 380
intake of an ‘unhealthy’ diet could increase the risk of death from other causes (HR=3.69; 381
95% CI 1.66-8.17) amongst breast cancer survivors compared to those with the lowest intake 382
of this diet[57]. 383
Page 16 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
17
The majority of studies investigating pre- or post-diagnosis fruit and vegetable intake 384
indicated no association with mortality in breast cancer survivors. However, one study found 385
that, when comparing postmenopausal breast cancer survivors in the highest tertile to the 386
lowest tertile group, pre-diagnosis total vegetable intake improved overall survival 387
(HR=0.57; 95% CI 0.35-0.94) - no association was found for total fruit intake and mortality 388
in this cohort of breast cancer survivors[58]. In addition, Dal Maso et al. found an association 389
with total fruit and vegetable consumption and overall mortality (HR=1.27; 95% CI 1.00-390
1.61) when comparing survivors of the lowest intake group to the highest intake group[59]. 391
Results from the After Breast Cancer Pooling Project, combining data from 4 cohort studies, 392
indicated no association between post-diagnosis intakes of cruciferous vegetables and 393
survival amongst 11,390 breast cancer survivors[60]. Holmes et al. reported an association 394
between the highest post-diagnosis poultry consumption and mortality in women once 395
diagnosed with breast cancer (HR=0.70; 95% CI 0.50–0.97)[61]. No associations were found 396
for fish or red meat consumption and mortality in this population. Additionally, a high dairy 397
intake before diagnosis amongst female registered nurses who participated in the NHS, was 398
related to overall survival (HR=0.72; 95% CI 0.52–1.00)[61]. Kroenke et al. found that post-399
diagnosis dairy intake amongst women diagnosed with early-stage invasive breast cancer in 400
the LACE study, was associated with an increased overall mortality (HR=1.39; 95% CI 1.02-401
1.90)[62]. More specifically, high fat dairy was related to overall mortality and breast cancer-402
specific mortality in these women (respectively HR=1.64; 95% CI 1.24-2.17 and HR=1.49; 403
95% CI 1.00-2.24) whilst low-fat dairy was not[62]. Beasley et al. examined both meat and 404
dairy intake after diagnosis and found no association with survival in the Collaborative 405
Woman’s Longevity Study (CWLS)[63]. Pre-diagnosis intakes of neither bread, 406
sunflower/pumpkin seeds nor sesame/flaxseeds reduced the risk of mortality in the MARIE 407
study[64]. Finally, post-diagnosis butter/margarine/lard consumption did increase the risk of 408
Page 17 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
18
breast cancer recurrence in a follow-up study amongst 472 breast cancer survivors enrolled 409
from the Memorial Sloan-Kettering Cancer Centre (RR=1.30; 95% CI 1.03-1.64)[65]. 410
In summary, no conclusive evidence could be provided for an association between 411
most foods of the main food groups, including fruits, vegetables, meat, or dairy, and cancer 412
recurrence or mortality - evidence for each exposure and outcome was based on the results of 413
one study only or on inconsistent results. However, limited evidence appears to indicate that 414
the reduction of dietary fat after breast cancer diagnosis could increase relapse-free survival 415
amongst breast cancer survivors, adherence to the HEI-2005 score after diagnosis is 416
associated with decreased overall mortality, adherence to the AHEI diet after diagnosis is 417
associated with decreased death from other causes, and that adherence to a prudent diet after 418
diagnosis is associated with decreased death from other causes amongst breast cancer 419
survivors. Adherence to a pre-diagnosis Western diet is associated with death from other 420
causes whilst post-diagnosis adherence to a Western diet is associated with an increased risk 421
of overall mortality in breast cancer survivors. 422
423
Laryngeal cancer 424
One cohort study could be identified for the association between several foods from 425
the main food groups and mortality amongst laryngeal cancer survivors[66]. Crosignani et.al 426
examined dietary habits and survival in of 215 Italian male laryngeal cancer survivors on pre-427
diagnosis dietary habits and survival. The consumption of total vegetables (HR=0.57; 95% CI 428
0.35-0.94), beef/veal (HR=0.50; 95% CI 0.30-0.83), and bread (HR=0.54; 95% CI 0.32-0.90) 429
were all associated with a decreased risk of overall mortality when comparing the highest 430
versus the lowest intake group. No associations were found for poultry, fish, eggs, milk, 431
cheese, pasta, potatoes, citrus fruits, other fruits, butter, or olive oil. The authors speculate 432
Page 18 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
19
that the association between the highest beef/veal intakes and mortality could tentatively be 433
interpreted as an indicator of a good nutritional status of those participants[66]. 434
In summary, no conclusive evidence for an association between fruits, vegetables, 435
protein foods, grain foods, dairy, or oils and spreads, and mortality amongst laryngeal cancer 436
survivors could be provided, as evidence for each exposure and outcome was based on the 437
results of one study only. 438
439
Non-Hodgkin Lymphoma (NHL) 440
A total of 2 cohort studies could be identified for NHL survivors regarding the intake 441
of food items. One study indicated that pre-diagnosis intakes of total fruit and vegetables and 442
vegetables only (highest versus lowest intake) were associated with decreased overall 443
mortality (respectively HR=0.68; 95% CI 0.49-0.95 and HR=0.58; 95% CI 0.38-0.89) 444
amongst female NHL survivors[67]. Additionally, the highest intakes of citrus fruits and 445
green leafy vegetables compared with the lowest intakes were related to overall mortality 446
amongst survivors with NHL (respectively HR=0.73; 95% CI 0.54-0.99 and HR=0.71; 95% 447
CI 0.51-0.98). No associations were observed for total fruit intake, yellow vegetables, red 448
vegetables or bean vegetables and mortality whilst sub-analysis investigating fruit and 449
vegetables separately for each NHL subtypes did; consumption of citrus fruits improved 450
survival in diffuse large B-cell lymphoma survivors (overall mortality HR=0.40; 95% CI 451
0.22–0.72, cancer-specific mortality HR=0.36; 95% CI 0.16–0.80), and the highest 452
consumption of green leafy vegetables favoured overall mortality in follicular lymphoma 453
survivors (HR=0.27; 95% CI 0.10–0.76)[67]. Although Leo et al. found no association 454
between pre-diagnosis intakes of fruit, vegetables, meat, fish, or legumes, and mortality in 455
2,339 NHL survivors[68], dairy intake did appear to be associated with a higher overall 456
Page 19 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
20
mortality (HR=1.14; 95% CI 1.00-1.31), yet not with NHL-specific mortality (HR=1.16; 95% 457
CI 0.98-1.37)[68]. 458
In summary, no conclusive evidence for an association between intakes of fruit, 459
vegetables, protein foods, or dairy and mortality in NHL survivors could be provided, as 460
evidence for each exposure and outcome was based on the results of one study only or on 461
inconsistent results. 462
463
Prostate cancer 464
For prostate cancer 4 cohort studies could be identified. Adherence to a Western diet 465
after prostate cancer diagnosis was associated with increased overall mortality (HR=1.67; 466
95% CI 1.16–2.42) and prostate cancer-specific mortality (HR=2.53; 95% CI 1.00-6.42) 467
amongst non-metastatic prostate cancer survivors in the Physician’s Health Study (PHS)[69]. 468
The derived Western dietary patterns appeared to be driven by the consumption of processed 469
meat[69]. A prudent diet was investigated (showing overlapping characteristics with the 470
Mediterranean diet examined in the Health Professionals Follow-up Study (HPFS)); 471
adherence to a prudent diet after prostate cancer diagnosis was inversely associated with 472
overall mortality (RR=0.64; 95% CI 0.44–0.93) and appeared to be driven by the use of oil 473
and vinegar dressings[70]. The HPFS reported on a Mediterranean diet and mortality in 474
prostate cancer survivors after diagnosis[71]. Kenfield et al. demonstrated that post-diagnosis 475
adherence to a Mediterranean diet was associated with decreased overall mortality (HR=0.78; 476
95% CI 0.67-0.90); no association was observed for prostate cancer-specific mortality and 477
adherence to the Mediterranean diet[71]. A pre-diagnosis high fish consumption in men who 478
were diagnosed with prostate cancer while participating in the PHS was related to prolonged 479
survival (HR=0.52; 95% CI 0.30-0.91) according to Chavarro et al.[72]. Another study of 480
Yang et al. investigated post-diagnosis dairy intake amongst prostate cancer survivors[73]. 481
Page 20 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
21
The consumption of total dairy was associated with increased overall mortality (HR=1.76; 482
95% CI 1.21-2.55). Both high-fat and low-fat dairy consumption contributed to this adverse 483
association and overall mortality (respectively HR=1.22; 95% CI 1.08-1.38 and HR=1.17; 484
95% CI 1.05-1.29)[73]. 485
In summary, no conclusive evidence for an association between a Mediterranean diet 486
score, adherence to a prudent or Western diet, fish, or dairy, and mortality in prostate cancer 487
survivors could be provided, as evidence for each exposure and outcome was based on the 488
results of one study only. 489
490
DISCUSSION 491
This systematic review summarizes current scientific literature regarding dietary 492
patterns/indices and foods from the main food groups and health outcomes amongst different 493
groups of cancer survivors. Limited evidence appears to indicate that the reduction of dietary 494
fat after breast cancer diagnosis could increase relapse-free survival amongst breast cancer 495
survivors, adherence to the HEI-2005 score after diagnosis is associated with decreased 496
overall mortality, adherence to the AHEI diet after diagnosis is associated with decreased 497
death from other causes, and that adherence to a prudent diet after diagnosis is associated 498
with decreased death from other causes amongst breast cancer survivors. Adherence to a pre-499
diagnosis Western diet is associated with death from other causes whilst post-diagnosis 500
adherence to a Western diet is associated with an increased risk of overall mortality in breast 501
cancer survivors. Although no conclusive evidence could be provided for other survivors than 502
of breast cancer, the results of available studies investigating dietary patterns/indices and 503
food in other cancer survivors were described in detail. 504
505
Dietary patterns/indices 506
Page 21 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
22
It could be speculated that the lack of effect in the two identified RCTs investigating a 507
low-fat diet in breast cancer survivors is a consequence of the relatively short follow-up 508
period when using mortality as the primary outcome[49,50]. It did appear, however, that a 509
reduction in dietary fat intake could increase relapse-free survival amongst these 510
survivors[49]. Nevertheless, the true beneficial effect of dietary intake in this trial remains 511
uncertain since increased exercise and weight loss during the intervention may also have 512
advantaged these breast cancer survivors[49]. Adherence to a high-quality diet or a prudent 513
diet and the increase in survival could be explained by the effects of fruit and vegetables on 514
health in general. This could also clarify the increase in mortality amongst survivors with 515
adherence to a Western diet, as it is characterised by low intakes of vegetables and fruits. It 516
remains difficult, however, to disentangle the beneficial effect of fruit and vegetables from 517
other foods in the diet – it could even be speculated that not the consumption of fruit and 518
vegetables in a high-quality and prudent diet decrease mortality, but eating less amounts of 519
sugars, salt, and saturated fats, could explain the associations found with mortality and 520
relapse-free survival. 521
Besides the evidence for a potential role of a low-fat diet in breast cancer recurrence, 522
most studies showed an association with overall mortality and death from other causes; not 523
with cancer-specific mortality or cancer recurrence. Even if the exposures identified cannot 524
help these cancer outcomes, given the survivors of the investigated cancers have potential for 525
long-term survival, it is desirable for them to follow a diet that could help reduce other 526
conditions such as cardiovascular disease and increase overall life expectancy. The limited 527
number of studies indicate that additional long-term prospective studies are urgently needed 528
to improve the strength of evidence on the influence of dietary pattern/indices adherence on 529
cancer survival. 530
531
Page 22 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
23
Foods from the main food groups 532
The investigated healthy dietary patterns/indices are characterised by foods of the 533
main food groups. Epidemiological research on fruit and vegetable intake and cancer risk 534
increased rapidly over the last few decades and it has been suggested that people with high 535
intakes of fruit and vegetables, compared to those with low intakes, have a reduced risk of 536
developing cancer[74]. The wide variety of nutrients including vitamins, minerals, 537
phytochemicals and fibre in fruit and vegetables could influence epigenetic processes and 538
potentially via this way improve cancer outcomes[75,76]. However, the exact mechanisms of 539
how diet can alter genetic and epigenetic changes in cancer cells has yet to be established. 540
The majority of the identified studies found statistically non-significant results, based on a p-541
value that indicates the degree to which the data conform to the pattern predicted by the test 542
hypothesis and all the other assumptions used in the test. Nonetheless, the HRs<1 of two 543
studies investigating pre-diagnosis fruit intake overall mortality [64,77], although statically 544
non-significant results, could strengthen the evidence that adherence to a high-quality diet, 545
characterised by high intakes of fruit and vegetables, could decrease overall mortality in 546
breast cancer survivors. The consumption of fruits could, therefore, be encouraged in breast 547
cancer survivors as they are an important part of a high-quality diet to increase overall life 548
expectancy. Studies investigating the role of fruit after diagnosis in cancer survivors are 549
urgently needed. 550
551
Study strengths and limitations 552
The strengths of this systematic review are the inclusion of dietary patterns/indices 553
and whole foods, and the large total number of cancer survivors investigated. By examining 554
the whole diet, the intake of nutrients in combination is considered which provides 555
translatable real-life scenarios for clinical recommendations. 556
Page 23 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
24
The limitations of this systematic review were the inclusion of only 2 RCTs, the few 557
studies investigating post-diagnosis intake, the use of FFQs to collect dietary information 558
from participants in most studies, and the considerable heterogeneity in study design and 559
participant characteristics (tumour characteristics (stage/grade), treatment, age, time of 560
follow-up, comorbidity, differences in countries and ethnicity). Due to potential bias, data 561
from observational studies generally provide a lower strength of evidence than from RCTs, 562
even if they were well conducted. Conducting RCTs to investigate dietary intake in cancer 563
survivors with mortality as an outcome can be challenging for cancers with a relatively long 564
survival necessitating adherence to a diet in the long-term. The majority of studies included 565
in this systematic review investigated foods before cancer diagnosis, with only a few studies 566
in the post-diagnosis setting. Information on food intake after diagnosis is valuable for 567
investigating the effect of dietary changes on health outcomes amongst cancer survivors – 568
even though it is too late to amend lifestyle factors from before diagnosis, patients are more 569
receptive to advice after diagnosis. Although the use of FFQs is an inexpensive approach to 570
capture data from hundreds or thousands of individuals, it may not represent the usual foods 571
or portion sizes chosen by participants, and intake data can be compromised when multiple 572
foods are grouped with single listings. Developments in the screening, diagnosis and 573
treatment of cancers differ greatly between countries and therefore could influence survival. 574
Although most studies are adjusted for tumour stage, age and treatment, often no adjustments 575
could be made for influential lifestyle factors including BMI, physical activity and smoking. 576
It remains a challenge to disentangle the impact of diet from other lifestyle factors, and this 577
should always be taken into consideration when interpreting study results. 578
579
CONCLUSION 580
Page 24 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
25
To conclude, the reduction of dietary fat after breast cancer diagnosis could increase 581
relapse-free survival amongst breast cancer survivors, adherence to a high-quality diet may 582
protect against overall mortality and death from other causes amongst breast cancer 583
survivors, and adherence to a prudent diet may protect against death from other causes 584
amongst breast cancer survivors. Adherence to a Western diet before diagnosis may be 585
detrimental for breast cancer survivors concerning death from other causes whilst a Western 586
diet after diagnosis may increase overall mortality amongst these survivors. Additional large 587
and well-conducted studies, preferably RCTs, are needed to clarify whether dietary 588
patterns/indices and food intake could influence health outcomes in other cancer survivors. 589
590
Authorship contribution statement 591
SJ drafted the manuscript and worked on the conception, design and interpretation of data. SJ 592
and FvO selected articles, screened titles and abstracts, assessed study quality and extracted 593
data. SJ, FvO, RB and MZ were involved in the interpretation and discussion of the results 594
and critically revised the systematic review for important intellectual content. All authors, SJ, 595
FvO, RB, AW, FJvS, KKC and MZ, approved the final version of the systematic review. SJ 596
is the guarantor. 597
598
Conflict of interest 599
None of the authors have any conflict of interest in connection with this systematic review. 600
601
Funding 602
This research received no grant from any funding agency in the public, commercial or not-603
for-profit sectors. This systematic review has not as yet been registered. 604
Page 25 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
26
605
Data sharing statement 606
No additional data available. 607
608
Page 26 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
27
LITERATURE 609
1 Schwingshackl L, Schwedhelm C, Hoffmann G, et al. Food groups and risk of all-610
cause mortality: a systematic review and meta-analysis of prospective studies. Am J 611
Clin Nutr 2017;:ajcn153148. doi:10.3945/ajcn.117.153148 612
2 Schwingshackl L, Hoffmann G. Adherence to Mediterranean diet and risk of cancer: 613
an updated systematic review and meta-analysis of observational studies. Cancer Med 614
2015;4:1933–47. doi:10.1002/cam4.539 615
3 Leenders M, Sluijs I, Ros MM, et al. Fruit and vegetable consumption and mortality: 616
European prospective investigation into cancer and nutrition. Am J Epidemiol 617
2013;178:590–602. doi:10.1093/aje/kwt006 618
4 Nguyen CT, Pham NM, Lee AH, et al. Prevalence of and Risk Factors for Type 2 619
Diabetes Mellitus in Vietnam: A Systematic Review. Asia Pac J Public Heal 620
2015;27:588–600. doi:10.1177/1010539515595860 621
5 Bours MJ, Beijer S, Winkels RM, et al. Dietary changes and dietary supplement use, 622
and underlying motives for these habits reported by colorectal cancer survivors of the 623
Patient Reported Outcomes Following Initial Treatment and Long-Term Evaluation of 624
Survivorship (PROFILES) registry. Br J Nutr 2015;114:286–96. 625
doi:10.1017/S0007114515001798 626
6 Humpel N, Magee C, Jones SC. The impact of a cancer diagnosis on the health 627
behaviors of cancer survivors and their family and friends. Support Care Cancer 628
2007;15:621–30. doi:10.1007/s00520-006-0207-6 629
7 Zhang FF, Liu S, John EM, et al. Diet quality of cancer survivors and noncancer 630
individuals: Results from a national survey. Cancer 2015;121:4212–21. 631
doi:10.1002/cncr.29488 632
8 D’Avanzo B, La Vecchia C, Negri E, et al. Attributable risks for bladder cancer in 633
northern Italy. Ann Epidemiol 1995;5:427–31. 634
9 Rock CL, Doyle C, Demark-Wahnefried W, et al. Nutrition and physical activity 635
guidelines for cancer survivors. CA Cancer J Clin 2012;62:242–74. 636
doi:10.3322/caac.21142 637
10 World Cancer Research Fund International / American Institute for Cancer Research. 638
Diet, Nutrition, Physical Activity, and Breast Cancer Survivors. 2014. 639
http://www.wcrf.org/sites/default/files/Breast-Cancer-Survivors-2014-Report.pdf 640
(accessed 2 Jun 2017). 641
11 Parsons JK, Pierce JP, Natarajan L, et al. NIH Public Access. 2014;6:1–13. 642
doi:10.1158/1940-6207.CAPR-13-0050.A 643
12 Schwedhelm C, Boeing H, Hoffmann G, et al. Effect of diet on mortality and cancer 644
recurrence among cancer survivors: a systematic review and meta-analysis of cohort 645
studies. Nutr Rev 2016;74:737–48. doi:10.1093/nutrit/nuw045 646
13 Cancer Research UK. Cancer survival statistics. 647
http://www.cancerresearchuk.org/health-professional/cancer-statistics/survival 648
(accessed 10 Jun 2017). 649
Page 27 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
28
14 Reedy J, Wirfält E, Flood A, et al. Comparing 3 dietary pattern methods--cluster 650
analysis, factor analysis, and index analysis--With colorectal cancer risk: The NIH-651
AARP Diet and Health Study. Am J Epidemiol 2010;171:479–87. 652
doi:10.1093/aje/kwp393 653
15 Guenther PM, Reedy J, Krebs-Smith SM. Development of the Healthy Eating Index-654
2005. J Am Diet Assoc 2008;108:1896–901. doi:10.1016/j.jada.2008.08.016 655
16 Guenther PM, Reedy J, Krebs-Smith SM, et al. Evaluation of the Healthy Eating 656
Index-2005. J Am Diet Assoc 2008;108:1854–64. doi:10.1016/j.jada.2008.08.011 657
17 Chiuve SE, Fung TT, Rimm EB, et al. Alternative dietary indices both strongly predict 658
risk of chronic disease. J Nutr 2012;142:1009–18. doi:10.3945/jn.111.157222 659
18 Inoue-Choi M, Lazovich D, Prizment AE, et al. Adherence to the World Cancer 660
Research Fund/American Institute for Cancer Research Recommendations for Cancer 661
Prevention Is Associated With Better Health-Related Quality of Life Among Elderly 662
Female Cancer Survivors. J Clin Oncol 2013;31:1758–66. 663
doi:10.1200/JCO.2012.45.4462 664
19 Catsburg C, Miller AB, Rohan TE. Adherence to cancer prevention guidelines and risk 665
of breast cancer. Int J Cancer 2014;135:2444–52. doi:10.1002/ijc.28887 666
20 Kant AK, Schatzkin A, Graubard BI, et al. A prospective study of diet quality and 667
mortality in women. JAMA 2000;283:2109–668
15.http://www.ncbi.nlm.nih.gov/pubmed/10791502 (accessed 10 Jun 2017). 669
21 Haines PS, Siega-Riz AM, Popkin BM. The Diet Quality Index revised: a 670
measurement instrument for populations. J Am Diet Assoc 1999;99:697–704. 671
doi:10.1016/S0002-8223(99)00168-6 672
22 Fung TT, Chiuve SE, McCullough ML, et al. Adherence to a DASH-Style Diet and 673
Risk of Coronary Heart Disease and Stroke in Women. Arch Intern Med 674
2008;168:713. doi:10.1001/archinte.168.7.713 675
23 Olsen A, Egeberg R, Halkjær J, et al. Healthy aspects of the Nordic diet are related to 676
lower total mortality. J Nutr 2011;141:639–44. doi:10.3945/jn.110.131375 677
24 Fung TT, Kashambwa R, Sato K, et al. Post diagnosis diet quality and colorectal 678
cancer survival in women. PLoS One 2014;9:1–13. doi:10.1371/journal.pone.0115377 679
25 Kim EHJ, Willett WC, Fung T, et al. Diet quality indices and postmenopausal breast 680
cancer survival. Nutr Cancer 2011;63:381–8. doi:10.1080/01635581.2011.535963 681
26 McCullough ML, Willett WC. Evaluating adherence to recommended diets in adults: 682
the Alternate Healthy Eating Index. Public Health Nutr 2006;9:152–683
7.http://www.ncbi.nlm.nih.gov/pubmed/16512963 (accessed 12 Aug 2017). 684
27 Sofi F, Macchi C, Abbate R, et al. Mediterranean diet and health status: an updated 685
meta-analysis and a proposal for a literature-based adherence score. Public Health 686
Nutr 2014;17:2769–82. doi:10.1017/S1368980013003169 687
28 Schwingshackl L, Hoffmann G. Diet Quality as Assessed by the Healthy Eating Index, 688
the Alternate Healthy Eating Index, the Dietary Approaches to Stop Hypertension 689
Page 28 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
29
Score, and Health Outcomes: A Systematic Review and Meta-Analysis of Cohort 690
Studies. J Acad Nutr Diet 2015;115:780–800.e5. doi:10.1016/j.jand.2014.12.009 691
29 Public Health England. The Eatwell Guide. 692
2017.https://www.gov.uk/government/publications/the-eatwell-guide (accessed 10 Jun 693
2017). 694
30 Sterne J, Julian H, Barney R. The Cochrane Collaboration’s tool for assessing risk of 695
bias Description. Handb CochraneTable 85.a 696
2014;:4.http://handbook.cochrane.org/chapter_8/table_8_5_a_the_cochrane_collaborat697
ions_tool_for_assessing.htm 698
31 Sterne JA, Hernán MA, Reeves BC, et al. The Risk Of Bias In Non-randomized 699
Studies – of Interventions (ROBINS-I) assessment tool ROBINS-I tool (Stage I): At 700
protocol stage. 2016;:1–22. doi:10.1136/bmj.i4919 701
32 Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating 702
quality of evidence and strength of recommendations. BMJ 2008;336:924–6. 703
doi:10.1136/bmj.39489.470347.AD 704
33 Moher D. PRISMA 2009 Checklist PRISMA 2009 Checklist. PLoS Med 705
2009;6:e1000097. doi:10.1371/journal.pmed1000097 706
34 Higgins JP, Savović J, Page MJ, et al. Revised Cochrane risk of bias tool for 707
randomized trials (RoB 2.0). 2016;:52. 708
35 Sterne J, Higgins J, Reeves B, et al. A tool for evaluating risk of bias in non- 709
randomized studies of interventions ( ROBINS-I ): development and applications. 710
2015. 711
36 Tang L, Zirpoli GR, Guru K, et al. Intake of cruciferous vegetables modifies bladder 712
cancer survival. Cancer Epidemiol Biomarkers Prev 2010;19:1806–11. 713
doi:10.1158/1055-9965.EPI-10-0008 714
37 Meyerhardt JA, Niedzwiecki D, Hollis D, et al. Association of Dietary Patterns With 715
Cancer Recurrence and Survival in Patients With Stage III Colon Cancer. Jama 716
2007;298:754. doi:10.1001/jama.298.7.754 717
38 Zhu Y, Wu H, Wang PP, et al. Dietary patterns and colorectal cancer recurrence and 718
survival: a cohort study. BMJ Open 2013;3:e002270. doi:10.1136/bmjopen-2012-719
002270 720
39 Fung TT, Kashambwa R, Sato K, et al. Post Diagnosis Diet Quality and Colorectal 721
Cancer Survival in Women. PLoS One 2014;9:e115377. 722
doi:10.1371/journal.pone.0115377 723
40 Pelser C, Arem H, Pfeiffer RM, et al. Prediagnostic lifestyle factors and survival after 724
colon and rectal cancer diagnosis in the National Institutes of Health (NIH)-AARP 725
Diet and Health Study. Cancer 2014;120:1540–7. doi:10.1002/cncr.28573 726
41 Ratjen I, Schafmayer C, di Giuseppe R, et al. Postdiagnostic Mediterranean and 727
Healthy Nordic Dietary Patterns Are Inversely Associated with All-Cause Mortality in 728
Long-Term Colorectal Cancer Survivors. J Nutr 2017;147:636–44. 729
doi:10.3945/jn.116.244129 730
Page 29 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
30
42 Dik VK, Murphy N, Siersema PD, et al. Prediagnostic Intake of Dairy Products and 731
Dietary Calcium and Colorectal Cancer Survival--Results from the EPIC Cohort 732
Study. Cancer Epidemiol Biomarkers Prev 2014;23:1813–23. doi:10.1158/1055-733
9965.EPI-14-0172 734
43 Romaguera D, Ward H, Wark PA, et al. Pre-diagnostic concordance with the 735
WCRF/AICR guidelines and survival in European colorectal cancer patients: a cohort 736
study. BMC Med 2015;13:107. doi:10.1186/s12916-015-0332-5 737
44 Ward HA, Norat T, Overvad K, et al. Pre-diagnostic meat and fibre intakes in relation 738
to colorectal cancer survival in the European Prospective Investigation into Cancer and 739
Nutrition. Br J Nutr 2016;116:316–25. doi:10.1017/S0007114516001859 740
45 Yang B, McCullough ML, Gapstur SM, et al. Calcium, vitamin D, dairy products, and 741
mortality among colorectal cancer survivors: The cancer prevention study-II nutrition 742
cohort. J Clin Oncol 2014;32:2335–43. doi:10.1200/JCO.2014.55.3024 743
46 Skeie G, Braaten T, Olsen A, et al. Whole grain intake and survival among 744
Scandinavian colorectal cancer patients. Nutr Cancer 2014;66:6–13. 745
doi:10.1080/01635581.2014.847472 746
47 Carr PR, Jansen L, Walter V, et al. Associations of red and processed meat with 747
survival after colorectal cancer and differences according to timing of dietary 748
assessment. Am J Clin Nutr 2016;103:192–200. doi:10.3945/ajcn.115.121145 749
48 McCullough ML, Gapstur SM, Shah R, et al. Association between red and processed 750
meat intake and mortality among colorectal cancer survivors. J Clin Oncol 751
2013;31:2773–82. doi:10.1200/JCO.2013.49.1126 752
49 Chlebowski RT, Blackburn GL, Thomson CA, et al. Dietary fat reduction and breast 753
cancer outcome: interim efficacy results from the Women’s Intervention Nutrition 754
Study. J Natl Cancer Inst 2006;98:1767–76. doi:10.1093/jnci/djj494 755
50 Pierce JP, Natarajan L, Caan BJ, et al. Influence of a Diet Very High in Vegetables, 756
Fruit, and Fiber and Low in Fat on Prognosis Following Treatment for Breast Cancer. 757
Jama 2007;298:289. doi:10.1001/jama.298.3.289 758
51 George SM, Irwin ML, Smith AW, et al. Postdiagnosis diet quality, the combination of 759
diet quality and recreational physical activity, and prognosis after early-stage breast 760
cancer. Cancer Causes Control 2011;22:589–98. doi:10.1007/s10552-011-9732-9 761
52 George SM, Ballard-Barbash R, Shikany JM, et al. Better postdiagnosis diet quality is 762
associated with reduced risk of death among postmenopausal women with invasive 763
breast cancer in the women’s health initiative. Cancer Epidemiol Biomarkers Prev 764
2014;23:575–83. doi:10.1158/1055-9965.EPI-13-1162 765
53 Izano MA, Fung TT, Chiuve SS, et al. Are Diet Quality Scores After Breast Cancer 766
Diagnosis Associated with Improved Breast Cancer Survival? Nutr Cancer 767
2013;65:820–6. doi:10.1080/01635581.2013.804939 768
54 McCullough ML, Gapstur SM, Shah R, et al. Pre- and postdiagnostic diet in relation to 769
mortality among breast cancer survivors in the CPS-II Nutrition Cohort. Cancer 770
Causes Control 2016;27:1303–14. doi:10.1007/s10552-016-0802-x 771
Page 30 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
31
55 Kroenke CH, Fung TT, Hu FB, et al. Dietary patterns and survival after breast cancer 772
diagnosis. J Clin Oncol 2005;23:9295–303. doi:10.1200/JCO.2005.02.0198 773
56 Kwan ML, Weltzien E, Kushi LH, et al. Dietary patterns and breast cancer recurrence 774
and survival among women with early-stage breast cancer. J Clin Oncol 2009;27:919–775
26. doi:10.1200/JCO.2008.19.4035 776
57 Vrieling A, Buck K, Seibold P, et al. Dietary patterns and survival in German 777
postmenopausal breast cancer survivors. Br J Cancer 2013;108:188–92. 778
doi:10.1038/bjc.2012.521 779
58 McEligot AJ, Largent J, Ziogas A, et al. Dietary fat, fiber, vegetable, and 780
micronutrients are associated with overall survival in postmenopausal women 781
diagnosed with breast cancer. Nutr Cancer 2006;55:132–40. 782
doi:10.1207/s15327914nc5502_3 783
59 Dal Maso L, Zucchetto A, Talamini R, et al. Effect of obesity and other lifestyle 784
factors on mortality in women with breast cancer. Int J cancer 2008;123:2188–94. 785
doi:10.1002/ijc.23747 786
60 Nechuta S, Caan BJ, Chen WY, et al. Postdiagnosis Cruciferous Vegetable 787
Consumption and Breast Cancer Outcomes: A Report from the After Breast Cancer 788
Pooling Project. Cancer Epidemiol Biomarkers Prev 2013;22:1451–6. 789
doi:10.1158/1055-9965.EPI-13-0446 790
61 Holmes MD, Stampfer MJ, Colditz GA, et al. Dietary factors and the survival of 791
women with breast carcinoma. Cancer 1999;86:826–35. doi:10.1002/(SICI)1097-792
0142(19990901)86:5<826::AID-CNCR19>3.0.CO;2-0 793
62 Kroenke CH, Kwan ML, Sweeney C, et al. High-and low-fat dairy intake, recurrence, 794
and mortality after breast cancer diagnosis. J Natl Cancer Inst 2013;105:616–23. 795
doi:10.1093/jnci/djt027 796
63 Beasley JM, Newcomb PA, Trentham-Dietz A, et al. Post-diagnosis dietary factors 797
and survival after invasive breast cancer. Breast Cancer Res Treat 2011;128:229–36. 798
doi:10.1007/s10549-010-1323-z 799
64 Buck K, Zaineddin AK, Vrieling A, et al. Estimated enterolignans, lignan-rich foods, 800
and fibre in relation to survival after postmenopausal breast cancer. Br J Cancer 801
2011;105:1151–7. doi:10.1038/bjc.2011.374 802
65 Hebert JR, Hurley TG, Ma Y. The effect of dietary exposures on recurrence and 803
mortality in early stage breast cancer. Breast Cancer Res. Treat. 1998;51:17–28. 804
doi:10.1023/A:1006056915001 805
66 Crosignani P, Russo A, Tagliabue G, et al. Tobacco and diet as determinants of 806
survival in male laryngeal cancer patients. Int J Cancer 1996;65:308–13. 807
doi:10.1002/(SICI)1097-0215(19960126)65:3<308::AID-IJC5>3.0.CO;2-3 808
67 Han X, Zheng T, Foss F, et al. Vegetable and fruit intake and non-Hodgkin lymphoma 809
survival in Connecticut women. Leuk Lymphoma 2010;51:1047–54. 810
doi:10.3109/10428191003690364 811
68 Leo QJN, Ollberding NJ, Wilkens LR, et al. Nutritional factors and non-Hodgkin 812
Page 31 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
32
lymphoma survival in an ethnically diverse population: the Multiethnic Cohort. Eur J 813
Clin Nutr 2016;70:41–6. doi:10.1038/ejcn.2015.139 814
69 Yang M, Kenfield SA, Van Blarigan EL, et al. Dietary patterns after prostate cancer 815
diagnosis in relation to disease-specific and total mortality. Cancer Prev Res (Phila) 816
2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 817
70 Yang M, Kenfield SA, Van Blarigan EL, et al. Dietary Patterns after Prostate Cancer 818
Diagnosis in Relation to Disease-Specific and Total Mortality. Cancer Prev Res 819
2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 820
71 Kenfield SA, DuPre N, Richman EL, et al. Mediterranean diet and prostate cancer risk 821
and mortality in the Health Professionals Follow-up Study. Eur Urol 2014;65:887–94. 822
doi:10.1016/j.eururo.2013.08.009 823
72 Chavarro JE, Stampfer MJ, Hall MN, et al. A 22-y prospective study of fish intake in 824
relation to prostate cancer incidence and mortality. Am J Clin Nutr 2008;88:1297–825
303.http://www.ncbi.nlm.nih.gov/pubmed/18996866 (accessed 10 Jun 2017). 826
73 Yang M, Kenfield SA, Van Blarigan EL, et al. Dairy intake after prostate cancer 827
diagnosis in relation to disease-specific and total mortality. Int J Cancer 828
2015;137:2462–9. doi:10.1002/ijc.29608 829
74 Wiseman M. The second World Cancer Research Fund/American Institute for Cancer 830
Research expert report. Food, nutrition, physical activity, and the prevention of cancer: 831
a global perspective. Proc Nutr Soc 2008;67:253–6. doi:10.1017/S002966510800712X 832
75 Hardy TM, Tollefsbol TO. Epigenetic diet: impact on the epigenome and cancer. 833
Epigenomics 2011;3:503–18. doi:10.2217/epi.11.71 834
76 Daniel M, Tollefsbol TO. Epigenetic linkage of aging, cancer and nutrition. J Exp Biol 835
2015;218:59–70. doi:10.1242/jeb.107110 836
77 Dal Maso L, Zucchetto A, Talamini R, et al. Effect of obesity and other lifestyle 837
factors on mortality in women with breast cancer. Int J Cancer 2008;123:2188–94. 838
doi:10.1002/ijc.23747 839
840
841
Page 32 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
33
842
Table 1: literature search for the Pubmed database addressing the relationship between diet and mortality among bladder cancer survivors
(“urinary bladder neoplasms”[Mesh] OR “urinary bladder neoplasm*” OR “bladder cancer*” OR “bladder tumor*” OR “bladder tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang]
843
844
Page 33 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
34
Table 2: Number of studies investigating the association between pre-diagnosis dietary patterns/indices and mortality/cancer recurrence in different populations of cancer survivors
Diet quality indices Prudent / healthy diet Western diet / unhealthy diet
Cancer site /type No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 0 - - - - 0 - - - - 0 - - - -
Bowel 2 - 2 1 - 1 - 1 1 - 1 - 2 2 -
Breast 1 - 1 1 1 2 1 4 4 4 2 1 4 4 4
Cervix 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - -
Larynx 0 - - - - 0 - - - - 0 - - - -
MM 0 - - - - 0 - - - - 0 - - - -
NHL 0 - - - - 0 - - - - 0 - - - -
Prostate 0 - - - - 0 - - - - 0 - - - -
Testes 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - -
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and several dietary patterns in the same
population
Page 34 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
35
Table 3: Number of studies investigating the association between post-diagnosis dietary patterns/indices and mortality/cancer recurrence in different populations of cancer survivors
Diet quality indices Prudent diet / healthy diet Western diet / unhealthy
Cancer site /type No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 0 - - - - 0 - - - - 0 - - -
Bowel 2 - 5 3 - 2 1 2 1 - 2 1 2 1 -
Breast 7 1 11 9 8 2 1 2 2 2 2 1 2 2 2
Cervix 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - -
Larynx 0 - - - - 0 - - - - 0 - - - -
MM 0 - - - - 0 - - - - 0 - - - -
NHL 0 - - - - 0 - - - - 0 - - - -
Prostate 1 - 1 1 - 1 - 1 1 - 1 - 1 1 -
Testes 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - -
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and several dietary patterns in the same
population
Page 35 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
36
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and several food items in the same population
Table 4: Number of studies investigating the association between pre-diagnosis foods and mortality/cancer recurrence in different populations of cancer survivors
Fruit and vegetables Grain foods Protein foods Dairy and alternatives Oils and spreads
Cancer site
/type
No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 1 - 4 4 - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Bowel 0 - - - - 1 - 5 - - 3 1 6 6 2 2 - 6 6 - 0 - - - -
Breast 5 - 7 5 1 1 - 1 1 - 3 - 6 3 1 1 - 1 1 - 0 - - - -
Cervix 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Larynx 1 - 3 - - 1 - 3 - - 1 - 4 - - 1 - 2 - - 1 - 2 - -
MM 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
NHL 2 - 11 11 - 0 - - - - 1 - 3 3 - 1 - 1 1 - 0 - - - -
Prostate 0 - - - - 0 - - - - 1 - - 1 - 0 - - - - 0 - - - -
Testes 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Page 36 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
37
HL= Hodgkin lymphoma; NHL= non-Hodgkin lymphoma; MM= malignant melanoma; CR= cancer recurrence; OM= overall mortality; CSM= cancer-specific mortality; DO= death from other
causes than cancer; the number of studies does not correspond with the number of outcomes as some studies investigate multiple outcomes and several food items in the same population
Table 5: Number of studies investigating the association between post-diagnosis foods and mortality/cancer recurrence in different populations of cancer survivors
Fruit and vegetables Grain foods Protein foods Dairy and alternatives Oils and spreads
Cancer site
/type
No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO No of
studies
CR OM CSM DO
Bladder 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Bowel 0 - - - - 0 - - - - 1 - 2 2 2 1 - 2 2 - 0 - - - -
Breast 4 1 6 4 1 0 - - - - 4 1 5 6 1 3 3 5 2 - 1 1 - 1 -
Cervix 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
HL 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Kidney 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Larynx 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
MM 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
NHL 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Prostate 0 - - - - 0 - - - - 0 - - - - 1 - 3 - - 0 - - - -
Testes 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Uterus 0 - - - - 0 - - - - 0 - - - - 0 - - - - 0 - - - -
Page 37 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
Online supporting data File S1 for manuscript ‘the impact of dietary patterns and the
main food groups on mortality and recurrence in cancer survivors: systematic review
of current epidemiological literature’
Jochems et al., 31-05-2017
Page 38 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
2
Protocol Systematic Literature Review
INTRODUCTION Title The impact of diet on mortality in cancer survivors: A systematic review of current epidemiological literature Authors Sylvia H.J. Jochems, Frits H.M. van Osch, Richard T. Bryan, Anke Wesselius, Frederik J. van Schooten, K.K. Cheng, Maurice P. Zeegers Article type Systematic literature review (SLR) Language article English
Actual start date 01 December 2014 Updated 07 Mei 2017
- inclusion one additional exclusion criteria: sample size had to be > 200 survivors - exclusion of beverages for more clear focus of systematic review - inclusion of dietary indices and grain products as an exposure - inclusion of cancer recurrence as an outcome (these adjustments had influence on the outcome, exposure and search terms of the systematic review and were therefore adjusted accordingly)
Funding sources No Conflicts of interest No
Research question The aim of this study was to conduct a structured summary and evaluation of randomised controlled trials and observational studies addressing the relationship between the highest versus the lowest intake of dietary patterns/indices and foods of the main food groups and mortality and cancer recurrence amongst groups of survivors of common cancers with a ten-year survival rate of at least 50%. PICO model Population: cancer survivors Intervention/exposure: dietary patterns/indices and foods from the main food groups Comparator/control: highest versus lowest intake Outcome: mortality (overall, cancer-specific, and death from other causes) and cancer recurrence
Page 39 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
3
METHODS Database search
Pubmed / Medline (1966 - May 2017)
Embase (1980 – May 2017)
Cochrane Library (1993 – May 2017) Additional search Reference tracking of included and related articles, systematic reviews and meta-analyses Study types
Inclusion criteria
Randomized Controlled Trials
Cross-sectional studies
Cohort studies - retrospective & longitudinal
Case-control (including follow-up of cases)
Exclusion criteria
Animal studies
In vitro studies
Gene-nutrient interaction studies Population / participants
Inclusion criteria
Adult population, at least 18 years of age (both men and women)
Survivors of common cancers with a ten-year survival of at least 50% including bladder, bowel, breast, cervical, kidney, laryngeal, prostate, testicular, uterine cancer, malignant melanoma, and (non-)Hodgkin lymphoma
Exclusion criteria
Pre-cancerous conditions of other cancer types
Combination of different types of cancers Study characteristics
Exclusion criteria
Sample size of at least 200 survivors in the analysis [http://www.tandfonline.com/toc/hsem20/current]
Follow-up period of at least 4 years (the risk of cancer recurrence is the greatest within the first three years for most included cancers) [http://www.cancerresearchuk.org/about-cancer/what-is-cancer/why-some-cancers-come-back]
Adjustments had to be made for at least age, tumour characteristics (stage/grade), and preferably initial treatment, in the statistical analysis
Page 40 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4
Exposure / intervention Dietary patterns/indices and foods from the main food groups including:
(I) Dietary patterns that were considered were assessed by index-based methods and data-driven approaches, such as principal component analysis (factor analysis) and cluster analysis
(II) fruit and vegetables including citrus fruits, stone fruits, soft fruits, fleshy fruits, vine fruits, flower vegetables, leafy vegetables, stem vegetables, fruit vegetables, mushrooms, bulbs and roots;
(III) grain foods including potatoes, bread, rice, pasta and cereal; (IV) protein foods including meat (processed meat, unprocessed meat, red meat,
poultry), fish, eggs, tofu, nuts, seeds, pulses, legumes and beans; (V) dairy and alternative products including yoghurt, milk, cheese; (VI) oils and spreads including vegetable oils, spreads
Comparators / control Highest compared to lowest intake category of dietary patterns/indices and foods from the main food groups described under exposure Outcomes Primary outcomes Overall mortality, cancer-specific mortality, death from other causes, and cancer recurrence Secondary outcomes None
Analysis (consideration of a meta-analysis besides a systematic review) We expect diversity in pre- and post-diagnosis dietary patterns/indice and foods, as well as the different cancers, and therefore decided to only consider comparable studies (same timeframe pre- or post-diagnosis and same cancer type) for meta-analysis. If more than 75% of the review will not have 3 or more studies that can be pooled under these conditions, only a systematic review and no meta-analysis will be conducted.
Article selection Inclusion criteria
Investigate the associations between dietary patterns/indices and foods, and mortality and cancer recurrence in survivors of primary cancer
Report a measure of the effect/association of the exposure on the outcomes
The duration of the exposure/intervention had to be recorded as well as the time between exposure assessment / intervention and outcome assessment
Diet and lifestyle modifications/changes consequent on the disease or its treatment will not be included
Present results of primary and secondary analysis
Present results for any of the following outcomes: • Overall mortality • Cancer-specific mortality • Death from other causes • Cancer recurrence
Page 41 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
5
Original articles published in peer-reviewed journals
Articles written in the English language Exclusion criteria
Systematic reviews, meta-analysis Reviews, comments, letters, conference abstracts
Data extraction The data extraction will be performed independently by two of the authors (Sylvia Jochems and Frits van Osch) and any disagreements about study inclusion will be resolved through consensus or, if necessary, a third party (Rik Bryan). Information to extract from studies: Author, Study, Country, Number of participants, sex, age, follow-up period, exposure, exposure timeframe, exposure assessment, outcome, results (HR/RR and 95% CI), adjustments in the statistical analysis. Statistical significance tests used in individual studies The authors of this review would like to note that results from the individual studies described as ‘not associated with mortality or cancer recurrence’, are mostly based on statistical significance tests with the focus on traditional definitions of p-values and statistical significance on null hypotheses. Notwithstanding, it is our believe that a correct interpretation of statistical tests demands critical examining the sizes of effect estimates and confidence limits, p-values, and the assumptions and conventions used for the statistical analyses.
Tools for assessing Risk of Bias and Level of Quality The Cochrane Collaboration risk of bias assessment tools were used for appraisal of RCTs and cohort studies. For RCTs the RoB 2.0 tool (a revised tool for risk of bias in randomized trials) will be used to evaluate the risk of bias. Cohort studies will be evaluated with an adjusted version of the ROBINS-I tool [http://methods.cochrane.org/bias/risk-bias-non-randomized-studies-interventions]. Levels of quality were determined with the GRADE system [http://handbook.cochrane.org/chapter_12/12_2_1_the_grade_approach.htm].
General search terms in Pubmed 1. Searching for all studies relating to cancer and survival:
Neoplasms, neoplasm staging, neoplasm recurrence local, neoplasia, tumours, cancer, survivors, survival analysis, recurrence, mortality, survival rate, disease management
2. Searching for all studies relating to dietary modification: Foods from the main food groups, dietary indices, dietary patterns
3. Selecting randomised control trials: Randomized controlled trial, random allocation, double blind method, single-blind method, clinical trial
4. Selecting cohort studies: epidemiologic studies, cohort studies, follow-up studies, longitudinal studies, prospective studies, retrospective studies
5. Additional filters: English language, human, full text These search terms will be adapted for use in the Ovid database (EMBASE and the Cochrane library).
Page 42 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
6
Specific search terms per cancer site in Pubmed The search strategy will be adapted for use in EMBASE and the Cochrane Library
1. Searching for all studies relating to bladder cancer: (“urinary bladder neoplasms”[Mesh] OR “urinary bladder neoplasm*” OR “bladder cancer*” OR “bladder tumor*” OR “bladder tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 43 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7
2. Searching for all studies relating to breast cancer: (“breast neoplasms/diet therapy”[Mesh] OR “breast neoplasms/mortality”[Mesh] OR “breast neoplasms/prevention and control”[Mesh] OR “mammary neoplasm*” OR “breast neoplasm*” OR “mammary cancer*” OR “breast cancer*” OR “breast carcinoma*” OR “human mammary carcinoma*” OR “breast tumor*” OR “breast tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 44 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
8
3. Searching for all studies relating to cervical cancer: (“uterine cervical neoplasms/diet therapy”[Mesh] OR “uterine cervical neoplasms/mortality”[Mesh] OR “uterine cervical neoplasms/prevention and control”[Mesh] OR “uterine cervical neoplasm*” OR “cervical cancer*” OR “cervical neoplasm*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 45 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
9
4. Searching for all studies relating to bowel cancer:
(“colorectal neoplasms/diet therapy”[Mesh] OR “colorectal neoplasms/mortality”[Mesh] OR “colorectal neoplasms/prevention and control”[Mesh] OR “colorectal neoplasm*” OR “colorectal cancer*” OR “colorectal tumor*” OR “colorectal tumour*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 46 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
10
5. Searching for all studies relating to Hodgkin lymphoma: (“Hodgkin disease/diet therapy”[Mesh] OR “Hodgkin disease/mortality”[Mesh] OR “Hodgkin disease/prevention and control”[Mesh] OR “Hodgkin disease” OR “Hodgkin Lymphoma” OR “Malignant Lymphogranuloma*” OR “Hodgkin lymphoma” OR “nodular lymphocyte predominant Hodgkin's lymphoma” OR “nodular sclerosing Hodgkin's lymphoma” OR “lymphocyte rich classical Hodgkin's lymphoma” OR “mixed cellularity Hodgkin's lymphoma”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 47 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
11
6. Searching for all studies relating to non-Hodgkin lymphoma:
(“lymphoma, non-Hodgkin/diet therapy”[Mesh] OR “lymphoma, non-Hodgkin/mortality”[Mesh] OR “lymphoma, non-Hodgkin/prevention and control”[Mesh] OR “lymphoma, non-Hodgkin” OR “nonhodgkins lymphoma” OR “non-Hodgkins lymphoma”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 48 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
12
7. Searching for all studies relating to kidney cancer: (“Kidney Neoplasms/diet therapy”[Mesh] OR “Kidney Neoplasms/mortality”[Mesh] OR “Kidney Neoplasms/prevention and control”[Mesh] OR “Kidney Neoplasm*” OR “Renal Neoplasm*” OR “Kidney Cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 49 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
13
8. Searching for all studies relating to larynx cancer:
(“laryngeal neoplasms/diet therapy”[Mesh] OR “laryngeal neoplasms/mortality”[Mesh] OR “laryngeal neoplasms/prevention and control”[Mesh] OR “laryngeal neoplasm*” OR “larynx cancer*” OR “larynx neoplasm*” OR “laryngeal cancer”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 50 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
14
9. Searching for all studies relating to multiple myeloma:
(“multiple myeloma/diet therapy”[Mesh] OR “multiple myeloma/mortality”[Mesh] OR “multiple myeloma/prevention and control”[Mesh] OR “multiple myeloma*” OR “plasma cell myeloma*” OR “myelomatosis” OR “myelomatoses” OR “Kahler Disease” OR “myeloma multiple”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 51 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
15
10. Searching for all studies relating to malignant melanoma: ("Melanoma/ diet therapy”[Mesh] OR “Melanoma/mortality”[Mesh] OR “Melanoma/prevention and control”[Mesh] OR “Malignant Melanoma*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 52 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
16
11. Searching for all studies relating to prostate cancer: (“prostatic neoplasms/diet therapy”[Mesh] OR “prostatic neoplasms/mortality”[Mesh] OR “prostatic neoplasms/prevention and control”[Mesh] OR “prostate neoplasm*” OR “prostatic neoplasm*” OR “prostate cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 53 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
17
12. Searching for all studies relating to testicular cancer:
(“testicular neoplasms/diet therapy”[Mesh] OR “testicular neoplasms/mortality”[Mesh] OR “testicular neoplasms/prevention and control”[Mesh] OR “testicular neoplasm*” OR “testicular tumor*” OR “testicular tumour*” OR “testis neoplasm*” OR “testis cancer*” OR “testicular cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 54 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
18
13. Searching for all studies relating to uterus cancer: (“uterine neoplasms/diet therapy”[Mesh] OR “uterine neoplasms/mortality”[Mesh] OR “uterine neoplasms/prevention and control”[Mesh] OR “uterus neoplasm*” OR “uterine neoplasm*” OR “uterus cancer*” OR “uterine cancer*”) AND (“mortalit*” OR “mortality”[Mesh] OR “death”[Mesh] OR “recurrence*” OR “recurrence”[Mesh] OR “surviv*” OR “survival”[Mesh] OR “disease-free survival”[Mesh] OR “survival analysis”[Mesh] OR “survival rate”[Mesh] OR “proportional hazards models”[Mesh] OR “kaplan-meier” OR “cox” OR “survivors”[Mesh]) AND (“diet*” OR “diet”[Mesh] OR “dietary pattern*” OR “diet, Mediterranean”[Mesh] OR “diet, vegetarian”[Mesh] OR “diet, Western”[Mesh] OR “energy intake”[Mesh] OR “caloric restriction”[Mesh] OR “low calorie diet” OR “low fat diet” OR “dietary fat*” OR “dietary carbohydrate*” OR “dietary fiber” OR “dietary protein*” OR “nutrition*” OR “food*” OR “fruit”[Mesh] OR “fruit*” OR “citrus fruit*” OR “vegetables”[Mesh] OR “vegetable*” OR “brassica” OR “cruciferous vegetable*” OR “meat”[Mesh] OR “red meat”[Mesh] OR “meat*” OR “beef” OR “pork” OR “lamb” OR “poultry” OR “chicken” OR “turkey” OR “duck” OR “fish products”[Mesh] OR “fish” OR “shellfish” OR “seafood” OR “dairy” OR “milk” OR “ghee” OR “cheese” OR “ice cream” OR “egg*” OR “nut*” OR “edible grain”[Mesh] OR “whole grains”[Mesh] OR “potato*” OR “bread” OR “cereal*” OR “rice*”) NOT (“DNA-Binding Proteins”[Mesh] OR “Peptides”[Mesh] OR “Intercellular Signaling Peptides and Proteins”[Mesh] OR "Chromosomes, Human"[Mesh] OR "Immunohistochemistry"[Mesh] OR "In Situ Hybridization, Fluorescence"[Mesh]) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms])
Page 55 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
19
The Risk Of Bias In Non-randomized Studies – of Interventions (ROBINS-I) assessment tool (version for cohort-type studies) Version 19 September 2016
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License.
ROBINS-I tool (Stage I): At protocol stage
Specify the review question
Participants
Experimental
intervention
Comparator
Outcomes
List the confounding domains relevant to all or most studies
List co-interventions that could be different between intervention groups and that could
impact on outcomes
Page 56 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
20
ROBINS-I tool (Stage II): For each study
Specify a target randomized trial specific to the study
Design Individually randomized / Cluster randomized / Matched (e.g. cross-
over)
Participants
Experimental
intervention
Comparator
Is your aim for this study…? to assess the effect of assignment to intervention to assess the effect of starting and adhering to intervention
Specify the outcome
Specify which outcome is being assessed for risk of bias (typically from among those earmarked for
the Summary of Findings table). Specify whether this is a proposed benefit or harm of intervention.
Specify the numerical result being assessed
In case of multiple alternative analyses being presented, specify the numeric result (e.g. RR = 1.52
(95% CI 0.83 to 2.77) and/or a reference (e.g. to a table, figure or paragraph) that uniquely defines
the result being assessed.
Page 57 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
21
Preliminary consideration of confounders
Complete a row for each important confounding domain (i) listed in the review protocol; and (ii)
relevant to the setting of this particular study, or which the study authors identified as potentially
important. “Important” confounding domains are those for which, in the context of this study, adjustment is expected to lead to a clinically important change in the estimated effect of the intervention. “Validity” refers to whether the confounding variable or variables fully measure the domain, while “reliability” refers to the precision of the measurement (more measurement error means less reliability).
(i) Confounding domains listed in the review protocol
Confounding
domain
Measured
variable(s)
Is there evidence
that controlling for
this variable was
unnecessary?*
Is the confounding
domain measured
validly and
reliably by this
variable (or these
variables)?
OPTIONAL: Is
failure to adjust for
this variable
(alone) expected to
favour the
experimental
intervention or the
comparator?
Yes / No / No
information
Favour
experimental /
Favour comparator
/ No information
(ii) Additional confounding domains relevant to the setting of this particular study, or which
the study authors identified as important
Confounding
domain
Measured
variable(s)
Is there evidence
that controlling for
this variable was
unnecessary?*
Is the confounding
domain measured
validly and
reliably by this
variable (or these
variables)?
OPTIONAL: Is
failure to adjust for
this variable
(alone) expected to
favour the
experimental
intervention or the
Page 58 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
22
comparator?
Yes / No / No
information
Favour
experimental /
Favour comparator
/ No information
* In the context of a particular study, variables can be demonstrated not to be confounders and so not included in the analysis: (a) if they are not predictive of the outcome; (b) if they are not predictive of intervention; or (c) because adjustment makes no or minimal difference to the estimated
effect of the primary parameter. Note that “no statistically significant association” is not the same as “not predictive”.
Page 59 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
23
Preliminary consideration of co-interventions
Complete a row for each important co-intervention (i) listed in the review protocol; and (ii) relevant
to the setting of this particular study, or which the study authors identified as important. “Important” co-interventions are those for which, in the context of this study, adjustment is expected to lead to a clinically important change in the estimated effect of the intervention.
(i) Co-interventions listed in the review protocol
Co-intervention Is there evidence that controlling
for this co-intervention was
unnecessary (e.g. because it was
not administered)?
Is presence of this co-
intervention likely to favour
outcomes in the experimental
intervention or the comparator
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
(ii) Additional co-interventions relevant to the setting of this particular study, or which the
study authors identified as important
Co-intervention Is there evidence that controlling
for this co-intervention was
unnecessary (e.g. because it was
not administered)?
Is presence of this co-
intervention likely to favour
outcomes in the experimental
intervention or the comparator
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Favour experimental / Favour
comparator / No information
Page 60 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
24
Risk of bias assessment
Responses underlined in green are potential markers for low risk of bias, and responses in red are
potential markers for a risk of bias. Where questions relate only to sign posts to other questions, no
formatting is used.
Signalling questions Description Response
options
Bias due to confounding
1.1 Is there potential for confounding of the effect of
intervention in this study?
If N/PN to 1.1: the study can be considered to be at low risk of
bias due to confounding and no further signalling questions need
be considered
Y / PY / PN / N
If Y/PY to 1.1: determine whether there is a need to assess time-
varying confounding:
1.2. Was the analysis based on splitting participants’ follow
up time according to intervention received?
If N/PN, answer questions relating to baseline
confounding (1.4 to 1.6)
If Y/PY, go to question 1.3.
NA / Y / PY /
PN / N / NI
1.3. Were intervention discontinuations or switches likely to
be related to factors that are prognostic for the outcome?
If N/PN, answer questions relating to baseline
confounding (1.4 to 1.6)
If Y/PY, answer questions relating to both baseline and
time-varying confounding (1.7 and 1.8)
NA / Y / PY /
PN / N / NI
Questions relating to baseline confounding only
1.4. Did the authors use an appropriate analysis method that
controlled for all the important confounding domains?
NA / Y / PY / PN / N / NI
1.5. If Y/PY to 1.4: Were confounding domains that were
controlled for measured validly and reliably by the
variables available in this study?
NA / Y / PY / PN / N / NI
1.6. Did the authors control for any post-intervention
variables that could have been affected by the intervention?
NA / Y / PY / PN / N / NI
Questions relating to baseline and time-varying confounding
1.7. Did the authors use an appropriate analysis method that
controlled for all the important confounding domains and
for time-varying confounding?
NA / Y / PY / PN / N / NI
1.8. If Y/PY to 1.7: Were confounding domains that were
controlled for measured validly and reliably by the
variables available in this study?
NA / Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious /
Critical / NI
Optional: What is the predicted direction of bias due to
confounding?
Favours experimental /
Favours comparator /
Unpredictable
Page 61 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
25
Bias in selection of participants into the study
2.1. Was selection of participants into the study (or into
the analysis) based on participant characteristics
observed after the start of intervention?
If N/PN to 2.1: go to 2.4
Y / PY / PN / N / NI
2.2. If Y/PY to 2.1: Were the post-intervention
variables that influenced selection likely to be
associated with intervention?
2.3 If Y/PY to 2.2: Were the post-intervention
variables that influenced selection likely to be
influenced by the outcome or a cause of the
outcome?
NA / Y / PY / PN / N / NI
NA / Y / PY / PN / N / NI
2.4. Do start of follow-up and start of intervention
coincide for most participants?
Y / PY / PN / N / NI
2.5. If Y/PY to 2.2 and 2.3, or N/PN to 2.4: Were
adjustment techniques used that are likely to correct for
the presence of selection biases?
NA / Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical /
NI
Optional: What is the predicted direction of bias due to
selection of participants into the study?
Favours experimental / Favours
comparator / Towards null /Away
from null / Unpredictable
Bias in classification of interventions
3.1 Were intervention groups clearly defined? Y / PY / PN / N / NI
3.2 Was the information used to define
intervention groups recorded at the start of the
intervention?
Y / PY / PN / N / NI
3.3 Could classification of intervention status have
been affected by knowledge of the outcome or risk
of the outcome?
Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of bias
due to classification of interventions?
Favours experimental / Favours
comparator / Towards null /Away from
null / Unpredictable
Page 62 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
26
Bias due to deviations from intended interventions
If your aim for this study is to assess the effect of assignment to
intervention, answer questions 4.1 and 4.2
4.1. Were there deviations from the intended intervention beyond what would
be expected in usual practice?
Y / PY / PN /
N / NI
4.2. If Y/PY to 4.1: Were these deviations from intended intervention
unbalanced between groups and likely to have affected the outcome?
NA / Y / PY /
PN / N / NI
If your aim for this study is to assess the effect of starting and adhering to
intervention, answer questions 4.3 to 4.6
4.3. Were important co-interventions balanced across intervention groups? Y / PY / PN /
N / NI
4.4. Was the intervention implemented successfully for most participants? Y / PY / PN /
N / NI
4.5. Did study participants adhere to the assigned intervention regimen? Y / PY / PN /
N / NI
4.6. If N/PN to 4.3, 4.4 or 4.5: Was an appropriate analysis used to estimate
the effect of starting and adhering to the intervention?
NA / Y / PY /
PN / N / NI
Risk of bias judgement
Optional: What is the predicted direction of bias due to deviations from the
intended interventions?
Bias due to missing data
5.1 Were outcome data available for all, or nearly all,
participants?
Y / PY / PN / N / NI
5.2 Were participants excluded due to missing data
on intervention status?
Y / PY / PN / N / NI
5.3 Were participants excluded due to missing data
on other variables needed for the analysis?
Y / PY / PN / N / NI
5.4 If PN/N to 5.1, or Y/PY to 5.2 or 5.3: Are the
proportion of participants and reasons for missing
data similar across interventions?
NA / Y / PY / PN / N / NI
5.5 If PN/N to 5.1, or Y/PY to 5.2 or 5.3: Is there
evidence that results were robust to the presence of
missing data?
NA / Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of bias due
to missing data?
Favours experimental / Favours
comparator / Towards null /Away from
null / Unpredictable
Page 63 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
27
Bias in measurement of outcomes
6.1 Could the outcome measure have been
influenced by knowledge of the intervention
received?
Y / PY / PN / N / NI
6.2 Were outcome assessors aware of the
intervention received by study participants?
Y / PY / PN / N / NI
6.3 Were the methods of outcome assessment
comparable across intervention groups?
Y / PY / PN / N / NI
6.4 Were any systematic errors in measurement
of the outcome related to intervention
received?
Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of
bias due to measurement of outcomes?
Favours experimental / Favours comparator /
Towards null /Away from null /
Unpredictable
Bias in selection of the reported result
Is the reported effect estimate likely to be
selected, on the basis of the results, from...
7.1. ... multiple outcome measurements
within the outcome domain?
Y / PY / PN / N / NI
7.2 ... multiple analyses of the intervention-
outcome relationship?
Y / PY / PN / N / NI
7.3 ... different subgroups? Y / PY / PN / N / NI
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the predicted direction of
bias due to selection of the reported result?
Favours experimental / Favours comparator /
Towards null /Away from null / Unpredictable
Overall bias
Risk of bias judgement Low / Moderate / Serious / Critical / NI
Optional: What is the overall predicted
direction of bias for this outcome?
Favours experimental / Favours comparator /
Towards null /Away from null / Unpredictable
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License.
Page 64 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
28
The RoB 2.0 tool (individually randomized, parallel group trials)
Assessor name/initials
Study ID and/or reference(s)
Study design
Randomized parallel group trial Cluster-randomized trial Randomized cross-over or other matched design
Specify which outcome is being assessed for risk
of bias
Specify the numerical result being assessed. In
case of multiple alternative analyses being presented,
specify the numeric result (e.g. RR = 1.52 (95% CI
0.83 to 2.77) and/or a reference (e.g. to a table, figure
or paragraph) that uniquely defines the result being
assessed.
Is your aim for this study…? to assess the effect of assignment to intervention to assess the effect of starting and adhering to intervention
Which of the following sources have you obtained to help inform your risk of bias judgements
(tick as many as apply)?
Journal article(s) with results of the trial
Trial protocol
Statistical analysis plan (SAP)
Non-commercial trial registry record (e.g. ClinicalTrials.gov record)
Company-owned trial registry record (e.g. GSK Clinical Study Register record)
“Grey literature” (e.g. unpublished thesis)
Conference abstract(s) about the trial
Regulatory document (e.g. Clinical Study Report, Drug Approval Package)
Research ethics application
Grant database summary (e.g. NIH RePORTER, Research Councils UK Gateway to
Research)
Personal communication with trialist
Personal communication with the sponsor
Page 65 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
29
Page 66 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
30
Domain Signalling questions Response options
Description/Support
for judgement
Bias arising from
the
randomization
process
1.1 Was the allocation
sequence random?
Y / PY / PN / N / NI
1.2 Was the allocation
sequence concealed until
participants were
recruited and assigned to
interventions?
Y / PY / PN / N / NI
1.3 Were there baseline
imbalances that suggest a
problem with the
randomization process?
Y / PY / PN / N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias arising from the
randomization process?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias due to
deviations from
intended
interventions
2.1. Were participants
aware of their assigned
intervention during the
trial?
Y / PY / PN / N / NI
2.2. Were carers and trial
personnel aware of
participants' assigned
intervention during the
trial?
Y / PY / PN / N / NI
2.3. If Y/PY/NI to 2.1 or
2.2: Were important co-
interventions balanced
across intervention
groups?
NA / Y / PY / PN /
N / NI
2.4. Was the intervention
implemented
successfully?
Y / PY / PN / N / NI
2.5. Did study
participants adhere to the
assigned intervention
regimen?
Y / PY / PN / N / NI
Page 67 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
31
Domain Signalling questions Response options
Description/Support
for judgement
2.6. If N/PN/NI to 2.3,
2.4 or 2.5: Was an
appropriate analysis used
to estimate the effect of
starting and adhering to
the intervention?
NA / Y / PY / PN /
N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to deviations
from intended
interventions?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias due to
missing outcome
data
3.1 Were outcome data
available for all, or
nearly all, participants
randomized?
Y / PY / PN / N / NI
3.2 If N/PN/NI to 3.1:
Are the proportions of
missing outcome data
and reasons for missing
outcome data similar
across intervention
groups?
NA / Y / PY / PN /
N / NI
3.3 If N/PN/NI to 3.1: Is
there evidence that
results were robust to the
presence of missing
outcome data?
NA / Y / PY / PN /
N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to missing
outcome data?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias in
measurement of
the outcome
4.1 Were outcome assessors aware of the intervention received by study participants?
Y / PY / PN / N / NI
Page 68 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
32
Domain Signalling questions Response options
Description/Support
for judgement
4.2 If Y/PY/NI to 4.1: Was the assessment of the outcome likely to be influenced by knowledge of intervention received?
NA / Y / PY / PN /
N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to measurement
of the outcome?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Bias in selection
of the reported
result
Are the reported outcome data likely to have been selected, on the basis of the results, from...
5.1. ... multiple outcome measurements (e.g. scales, definitions, time points) within the outcome domain?
Y / PY / PN / N / NI
5.2 ... multiple analyses of the data?
Y / PY / PN / N / NI
Risk of bias judgement Low / High / Some
concerns
Optional: What is the
predicted direction of
bias due to selection of
the reported result?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Overall bias Risk of bias judgement Low / High / Some
concerns
Optional:
What is the overall
predicted direction of
bias for this outcome?
Favours
experimental /
Favours comparator
/ Towards null
/Away from null /
Unpredictable
Page 69 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
33
RESULT SEARCH: PRISMA FLOWCHART
PRISMA Flow Diagram – dietary patterns/indices and foods from
the main food groups and mortality/cancer recurrence amongst
different groups of cancer survivors
Records identified through database searching: Bladder cancer n= 317 Breast cancer n= 2044 Cervical cancer n= 123 Colorectal cancer n= 1157 (Non-)Hodgkin lymphoma n= 245 Kidney cancer n= 134 Larynx cancer n= 94 Malignant melanoma n= 103 Multiple myeloma n= 161 Prostate cancer n= 902 Testicular cancer n= 35 Uterus cancer n= 145
Scre
enin
g In
clu
ded
El
igib
ility
Id
enti
fica
tio
n
Additional records identified through other sources:
(n= 8)
Records after duplicates removed (n= 2883)
Records screened (n= 2883)
Records excluded: on title and abstract (n= 2788)
Full-text articles assessed for eligibility
(n= 95) Full-text articles excluded:
- other outcome than mortality or recurrence (n= 28)
- other exposure than dietary patterns/indices or our selected food items (n= 11)
- Not only cancer survivors (n= 10)
- Excluded because of our own exclusion criteria (sample size, follow-up length) (n= 8)
Studies included in qualitative synthesis
(n= 38)
Page 70 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Online supporting data File S2 for manuscript ‘the impact of dietary patterns and the main food groups on mortality and
recurrence in cancer survivors: systematic review of current epidemiological literature’
Jochems et al., 11-08-2017
Page 71 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Bladder cancer
Table S1: Summary of studies bladder cancer
Author (year)
Study / country Number of participants / sex (age range)
Follow-up period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI)
multivariate adjusted
Adjustment
Tang et al. (2010)
Roswell Park Cancer Institute (RPCI) / United States
239 m/w (not specified) 8.0 Fruit and vegetables
Pre-diagnosis FFQ usual diet in the few years before diagnosis
Overall mortality, cancer-specific mortality
Total fruit: HR1= 0.91; 95% CI 0.62-1.33 HR2= 1.09; 95% CI 0.66-1.81 Total vegetables: HR1= 0.91; 95% CI 0.62-1.36 HR2= 1.06; 95% CI 0.63-1.78 Cruciferous vegetables: HR1= 0.87; 95% CI 0.60-1.26 HR2= 0.89; 95% CI 0.53-1.48 Raw cruciferous vegetables: HR1= 0.73; 95% CI 0.50-1.06 HR2= 0.73; 95% CI 0.44-1.21
age at diagnosis, total meat intake, pack-years of smoking, tumour stage, radiation therapy
Page 72 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Bowel cancer
Table S2: Summary of studies bowel cancer
Author (year) Study / country Number of participants / sex (age at baseline)
Follow-up
period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Meyerhardt et al. (2007)
Cancer and Leukemia Group B (CALGB 89803) adjuvant therapy trial for stage III colon cancer / USA
1,009 m/w (21-85) 5.3 PCA: prudent diet, Western diet
Post-diagnosis
FFQ during and 6 months after adjuvant chemotherapy
Overall mortality, cancer recurrence
Prudent diet: HR1= 1.32; 95% CI 0.86-2.04 HR4= 1.13; 95% CI 0.77-1.67 Western diet: HR1= 2.32; 95% CI 1.36-3.96 HR4= 2.85; 95% CI 1.75-4.63
sex, age, depth of invasion through bowel wall, number of positive lymph nodes, presence of clinical perforation at time of surgery, presence of bowel obstruction at time of surgery, baseline performance status, treatment group, weight change between first and second questionnaire, time-varying BMI, time-varying physical activity level, time-varying total calories
McCullough et al. (2013)
Cancer Prevention Study II (CPSII) Nutrition Cohort / USA
2,315 m/w (40-93) 7.5 Red and processed meat, unprocessed red meat
Pre- and post-diagnosis
FFQ usual diet of the year before diagnosis and two times during follow-up
Overall mortality, cancer-specific mortality, death from other causes
Pre-diagnosis: Red and processed meat RR1= 1.29; 95% CI 1.05-1.59 RR2= 1.09; 95% CI 0.79-1.51 RR3= 1.39; 95% CI 1.00-1.92 Unprocessed red meat RR1= 1.12; 95% CI 0.92-1.38 RR2= 1.16; 95% CI 0.84-1.58 RR3= 1.19; 95% CI 0.87-1.64 Post-diagnosis: Red and processed meat RR1= 0.94; 95% CI 0.68-1.30 RR2= 1.10; 95% CI 0.61-1.98 RR3= 0.87; 95% CI 0.54-1.41 Unprocessed red meat: RR1= 0.75; 95% CI 0.55-1.03 RR2= 1.13; 95% CI 0.62-2.06 RR3= 0.64; 95% CI 0.40-1.03
pre-diagnosis model: age at diagnosis, sex, tumour stage at diagnosis, 1992 pre-diagnostic energy intake, BMI in 1992, history of diabetes, and history of myocardial infarction. Post-diagnosis model: age at diagnosis, sex, tumour stage at diagnosis, and post-diagnostic energy intake, weight change between 1992 pre-diagnostic, post-diagnostic questionnaires, and 1992 pre-diagnostic meat intake
Zhu et al. (2013)
Familial CRC registry in Newfoundland (FBCR) / Canada
529 m/w (20-75) 6.4 PCA: prudent vegetable pattern
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
Prudent vegetable pattern: HR1= 1.03; 95% CI 0.61-1.75 HR2= 1.12; 95% CI 0.69-1.84
total energy intake, sex, age at diagnosis, stage at diagnosis, marital status, family history, reported screening procedure, reported chemo-radiothr and microsatellite instability status
Dik et al. (2014)
The European Prospective Investigation into Cancer and Nutrition (EPIC) / pooled analysis data Denmark, Italy,
3,859 m/w (25-70) 4.1 Total dairy, milk, yoghurt, cheese
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
Total dairy: HR1= 1.16; 95% CI 0.98-1.36 HR2= 1.17; 95% CI 0.96-1.43 Milk:
age at diagnosis, sex, pre-diagnosis BMI, smoking status, energy intake, tumour subsite, disease stage,
Page 73 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Netherlands, Norway, Spain, Sweden, United Kingdom, France, Germany, Greece
HR1= 1.21; 95% CI 1.03-1.43 HR2= 1.21; 95% CI 0.99-1.48 Yoghurt: HR1= 1.08; 95% CI 0.92-1.28 HR2= 1.09; 95% CI 0.88-1.34 Cheese: HR1= 0.87; 95% CI 0.74-1.04 HR2= 0.93; 95% CI 0.76-1.14
differentiation grade; stratified by centre
Fung et al. (2014)
Nurses' Health Study (NHS) / USA
1,201 w (30-55) 11.2 Dietary indices: AHEI, DASH, AMED PCA: prudent diet, Western diet
Post-diagnosis
FFQ at least 6 months after diagnosis
Overall mortality, cancer-specific mortality
AHEI: HR1= 0.71; 95% CI 0.52-0.98 HR2= 0.72; 95% CI 0.43-1.21 DASH: HR1= 0.98; 95% CI 0.71-1.35 HR2= 0.87; 95% CI 0.52-1.45 AMED: HR1= 0.87; 95% CI 0.63-1.21 HR2= 0.84; 95% CI 0.50-1.42 Prudent diet: HR1= 0.93; 95% CI 0.65-1.34 HR2= 0.67; 95% CI 0.37-1.22 Western diet: HR1= 1.32; 95% CI 0.89-1.97 HR2= 1.66; 95% CI 0.85-3.23
age, physical activity, BMI, weight change, cancer grade, chemotherapy, smoking status, energy intake, colon or rectal cancer, stage of disease, and date of colorectal cancer diagnosis
Pelser et al. (2014)
NIH-AARP Diet and Health Study / USA
5,727 m/w (50-71) 5 Dietary indices: HEI
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
HEI: HR1= 0.95; 95% CI 0.78-1.16 HR2= 0.99; 95% CI 0.77-1.27
age, sex, lag time, education, family history cancer, stage, treatment, BMI, physical activity, alcohol, smoking
Skeie et al. (2014)
HELGA cohort including the Norwegian Women and Cancer Study, the Northern Sweden Health and Disease Study, and the Danish Diet Cancer and Health Study / Denmark, Norway, Sweden
1,119 m/w (30-64) 7 Total whole grains, whole grain wheat, whole grain rye, whole grain oats, whole grain products
Pre-diagnosis
FFQ usual diet before diagnosis
Overall mortality For men Total whole grains: HR1= 1.00; 95% CI 0.67-1.48 Whole grain wheat: HR1= 0.97; 95% CI 0.64-1.49 Whole grain rye: HR1= 0.90; 95% CI 0.60-1.36 Whole grain oats: HR1= 1.11; 95% CI 0.72-1.70 Whole grain products: HR1= 1.06; 95% CI 0.71-1.56 For women Total whole grains: HR1= 0.91; 95% CI 0.60-1.39 Whole grain wheat: HR1= 1.35; 95% CI 0.72-2.53 Whole grain rye: HR1= 0.93; 95% CI 0.60-1.46 Whole grain oats: HR1= 0.83; 95% CI 0.55-1.26 Whole grain products: HR1= 1.10; 95% CI 0.74-1.64
age at diagnosis, metastasis, smoking, folate, margarine, energy intake, stratified for country and cancer location. Wheat, rye and oats were also adjusted for the other grains
Page 74 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Yang et al. (2014)
Cancer Prevention Study II (CPSII) Nutrition Cohort / USA
2,284 m/w (40-92) 7.5 Total dairy, milk Pre- and post-diagnosis
FFQ usual diet of the year before diagnosis and two times during follow-up
Overall mortality, cancer-specific mortality
Pre-diagnosis: Total dairy RR1= 0.88; 95% CI 0.72-1.09 RR2= 0.89; 95% CI 0.65-1.22 Milk RR1= 0.95; 95% CI 0.79-1.15 RR2= 0.98; 95% CI 0.73-1.32 Post-diagnosis: Total dairy RR1= 0.75; 95% CI 0.56-1.01 RR2= 0.73; 95% CI 0.44-1.23 Milk RR1= 0.72; 95% CI 0.55-0.94 RR2= 0.93; 95% CI 0.59-1.49
pre-diagnosis: age at diagnosis, sex, tumour stage, pre-diagnosis total energy and total folate intakes. post-diagnosis: age at diagnosis, sex, tumour stage, post-diagnosis total energy and total folate intakes
Carr et al. (2016)
Darmkrebs: chancen der Verhutung durch Screening study (DACHS) / Germany
3,122 m/w (>30) 4.8 Red and processed meat
Pre-diagnosis
FFQ usual diet before diagnosis
Overall mortality, cancer-specific mortality, cancer recurrence
Red and processed meat: HR1= 0.85; 95% CI 0.67-1.09 HR2= 0.83; 95% CI 0.61-1.14 HR4= 1.03; 95% CI 0.80-1.33
age at diagnosis, sex, cancer stage, chemotherapy, surgery, BMI, physical activity, diabetes, stroke, heart failure, myocardial infarction, dairy intake, wholegrain intake, time between diagnosis and interview, time-dependent effect of chemotherapy
Romaguera et al. (2016)
The European Prospective Investigation into Cancer and Nutrition (EPIC) / pooled analysis data Denmark, Italy, Netherlands, Norway, Spain, Sweden, United Kingdom, France, Germany, Greece
3,292 m/w (25-70) 4.2 Dietary indices: WCRF/AICR score
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
WCRF/AICR score: HR1= 0.79; 95% CI 0.65-0.98 HR2= 0.70; 95% CI 0.56–0.89
age at diagnosis as entry time and age at death or censoring as exit time, year of diagnosis, tumour stage, tumour grade, tumour site, sex, educational level, and smoking status; stratified by country
Ward et al. (2016)
The European Prospective Investigation into Cancer and Nutrition (EPIC) / pooled analysis data Denmark, Italy, Netherlands, Norway, Spain, Sweden, United Kingdom, France, Germany, Greece
3,789 m/w (25-70) 4.1 Red meat, unprocessed meat, poultry
Pre-diagnosis
FFQ usual diet of the year before diagnosis
Overall mortality, cancer-specific mortality
Red and processed meat: HR1= 1.00; 95% CI 0.83-1.20 HR2= 1.00; 95% CI 0.81-1.23 Unprocessed red meat: HR1= 0.95; 95% CI 0.78-1.14 HR2= 0.93; 95% CI 0.75-1.15 Poultry: HR1= 0.87; 95% CI 0.73-1.03 HR2= 0.91; 95% CI 0.75-1.10
adjusted for age at diagnosis, sex, BMI, smoking status, tumour grade, tumour stage, year of tumour diagnosis, energy intake, calcium intake, folate intake, alcohol intake, education; stratified by country
Ratjen et al. (2017)
Patients with histologically confirmed colorectal cancer recruited by the PopGen biobank / Germany
1,404 m/w (56-67) 7 Dietary indices: Modified Mediterranean Diet Score (MMDS), Healthy Nordic Food Index (HNFI)
Post-diagnosis
FFQ usual diet assessed 6 years (median) after diagnosis
Overall mortality MMDS: HR1= 0.48; 95% CI 0.32-0.74 HNFI: HR1= 0.63; 95% CI 0.39-1.04
sex, age at diet assessment, BMI, physical activity, survival time from CRC diagnosis until diet assessment, tumour location, occurrence of metastases, occurrence of other cancer, chemotherapy, smoking status, total energy intake, time 3 age, time 3 BMI, and time 3 metastases
Page 75 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Breast cancer
Table S3: Summary of studies breast cancer
Author (year)
Study / country Number of participants /
sex (age)
Follow-up
period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI)
multivariate adjusted
Adjustment
Hebert et al. (1998)
Memorial Sloan- Kettering Cancer Center Follow-up Study / USA
472 w (20-80) 8-10 Red meat, butter/margarine/lard
Post-diagnosis
FFQ usual diet before diagnosis
Cancer-specific mortality, cancer recurrence
Red Meat: RR2= 2.60; 95% CI 0.96-7.03 RR4= 1.12; 95% CI 0.66-1.89 Butter/margarine/lard: RR2= 1.03; 95% CI 0.61-1.76 RR4= 1.30; 95% CI 1.03-1.64
disease stage, oestrogen receptor status, age, BMI, menopausal status, energy intake
Holmes et al. (1999)
Nurses' Health Study (NHS) / USA
1,504 w pre-diagnosis (mean age 54) and 1,982 w post-diagnosis)
13.1 Vegetables, poultry, fish, dairy, red meat (processed and unprocessed combined)
Pre- and post-diagnosis
FFQ usual diet after diagnosis
Overall mortality, cancer-specific mortality (results for breast cancer-specific mortality are not shown in paper)
Pre-diagnosis: Vegetables RR1= 0.98; 95% CI 0.62-1.53 Poultry RR1= 0.60; 95% CI 0.39-0.92 Fish RR1= 0.94; 95% CI 0.62-1.43 Dairy RR1= 0.71; 95% CI 0.44-1.14 Red meat (not shown) Post-diagnosis: Vegetables RR1= 0.81; 95% CI 0.59–1.11 Poultry RR1= 0.70; 95% CI 0.50–0.97 Fish RR1= 0.80; 95% CI 0.60–1.07 Dairy RR1= 0.72; 95% CI 0.52–1.00 Red meat RR1= 1.06; 95% CI 0.76–1.49
pre-diagnosis: quantiles of nutrient or food intake prior to diagnosis, previous diet interval, age, diet interval, calendar year of diagnosis, body mass index, oral contraceptive use, menopausal status, postmenopausal hormone use, smoking, age at first birth and parity, number of metastatic lymph nodes, tumour size, and caloric intake post-diagnosis: age, diet interval, calendar year of diagnosis, body mass index, oral contraceptive use, menopausal status, postmenopausal hormone use, smoking, age at first birth and parity, number of metastatic lymph nodes, tumour size, caloric intake
Kroenke et al. (2005)
Nurses' Health Study (NHS) / USA
2,619 w (30-55) 9 PCA: prudent diet, Western diet
Pre- and post-diagnosis
FFQ usual intake 4 years before diagnosis and FFQ at least one year after diagnosis
Overall mortality, cancer-specific mortality, death from other causes
Pre-diagnosis: Prudent diet RR1= non-significant (not shown) RR2= non-significant (not shown) RR3= non-significant (not shown) Western diet RR1= 1.40; 95% CI 0.93-2.09 RR2= 1.01; 95% CI 0.59-1.72 RR3= 1.95; 95% CI 1.06-3.60
age, time since diagnosis, BMI, energy intake, smoking, physical activity, diet missing, age at menarche, oral contraceptive use, menopausal status and use of postmenopausal hormone therapy, age at menopause, tamoxifen, chemotherapy, tumour stage at diagnosis, time between dietary assessment and diagnosis (for pre-diagnosis diet)
Page 76 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Post-diagnosis: Prudent diet RR1= 0.78; 95% CI 0.54-1.12 RR2= 1.07; 95% CI 0.66-1.73 RR3= 0.54; 95% CI 0.31-0.95 Western diet RR1= 1.53; 95% CI 1.03-2.29 RR2= 1.01; 95% CI 0.60-1.70 RR3= 2.09; 95% CI 1.30-3.36
McEligot et al. (2006)
Cancer Surveillance Program of Orange County (CSPOC) / USA
516 w (age >50) 6.7 Fruit, vegetables Pre-diagnosis
FFQ usual diet one year before diagnosis
Overall mortality Total fruit: HR1= 0.63; 95% CI 0.38-1.05 Total vegetables: HR1= 0.57; 95% CI 0.35–0.94
tumour stage, age at diagnosis, BMI, parity, HRT, alcohol intake, multivitamins, energy intake
Chlebowski et al. (2006)
Women’s Intervention Nutrition Study (WINS) RCT / USA
2,437 w 5 Dietary indices: Low fat diet
Post-diagnosis
FFQ with interview on dietary intake after diagnosis
Overall mortality, relapse-free survival
Intervention versus control Low-fat diet: HR1= 0.89; 95% CI 0.65-1.21 HR4= 0.76; 95% CI 0.60-0.98
nodal status, systemic adjuvant therapy, ER status, tumour size, mastectomy
Pierce et al. (2007a)
Women's Healthy Eating and Living (WHEL) RCT / USA
3,088 w (18-70) 7.3 Dietary indices: Low fat diet
Post-diagnosis
FFQ with interview on dietary intake after diagnosis
Overall mortality Intervention versus control Low-fat diet: HR1= 0.91; 95% CI 0.72-1.15
anti-oestrogen use, bilateral oophorectomy, age, BMI, physical activity, energy intake, tumour characteristics (including hormone receptor status), years from diagnosis to study entry
Dal Maso et al. (2008)
Six Italian Regions Follow-up Study / Italy
1,453 w (23-74) 12.6 Fruit and vegetables Pre-diagnosis
FFQ usual diet year before diagnosis
Overall mortality, cancer-specific mortality
Fruit and vegetables: HR1= 1.27; 95% CI 1.00–1.61 HR2= 1.26; 95% CI 0.96–1.64 (low versus high intake!)
region, age at diagnosis, year of diagnosis, TNM stage, receptor status
Kwan et al. (2009)
Life After Cancer Epidemiology (LACE) study / USA
1,901 w (18-79) 4.2 PCA: prudent diet, Western diet
Post-diagnosis
FFQ usual diet 3 years after diagnosis
Overall mortality, cancer-specific mortality, death from other causes, cancer recurrence
Prudent diet: HR1= 0.57; 95% CI 0.36-0.90 HR2= 0.79; 95% CI 0.43-1.43 HR3= 0.35; 95% CI 0.17-0.73 HR4= 0.95; 95% CI 0.63-1.43 Western diet: HR1= 1.76; 95% CI 1.10-2.81 HR2= 1.20; 95% CI 0.62-2.32 HR3= 2.15; 95% CI 0.97-4.77 HR4= 0.98; 95% CI 0.62-1.54
age at diagnosis, total energy intake, ethnicity, BMI, weight change before diagnosis to baseline, smoking status, menopausal status at diagnosis, stage, hormone receptor status, treatment
Beasley et al. (2011)
Collaborative Women’s Longevity Study (CWLS) / USA
4,441 w (20-79) 5.5 Fruit, vegetables, dairy, meat (poultry, fish, beef, and processed)
Post-diagnosis
FFQ usual diet after diagnosis (1-16 years)
Overall mortality, cancer-specific mortality
Total fruit: HR1= 1.38; 95% CI 0.88-2.17 HR2= 1.39; 95% CI 0.64-2.99 Total vegetables: HR1= 1.44; 95% CI 0.91-2.27 HR2= 0.96; 95% CI 0.38-2.45 Cruciferous vegetables: HR1= 1.02; 95% CI 0.80-1.30 HR2= 0.95; 95% CI 0.59-1.54 Dairy: HR1= 1.18; 95% CI 0.90-1.54 HR2= 0.94; 95% CI 0.56-1.59
age, state of residence, menopausal status, smoking, breast cancer stage, alcohol, history of hormone replacement therapy at diagnosis, interval between diagnosis and diet assessment, and energy intake, breast cancer treatment, body mass at follow-up
Page 77 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Meat (poultry, fish, beef, and processed): HR1= 1.12; 95% CI 0.83-1.51 HR2= 0.89; 95% CI 0.50-1.60
Buck et al. (2011)
MARIE study / Germany
2,653 w (50-74) 6.4 Fruit, vegetables, bread, sunflower/pumpkin seeds, sesame/flaxseeds
Pre-diagnosis
FFQ usual diet year before diagnosis
Overall mortality, cancer-specific mortality
Fruit: HR1= 0.84; 95% CI 0.61-1.16 HR2= 0.86; 95% CI 0.59-1.25 Vegetables: HR1= 1.09; 95% CI 0.80-1.48 HR2= 1.01; 95% CI 0.70-1.46 Bread: HR1= 1.31; 95% CI 0.93-1.83 HR2= 1.10; 95% CI 0.74-1.63 Sunflower/pumpkinseeds: HR1= 0.87; 95% CI 0.66-1.15 HR2= 1.12; 95% CI 0.79-1.57 Sesame/flaxseeds: HR1= 0.90; 95% CI 0.68-1.19 HR2= 1.21; 95% CI 0.87-1.68
tumour size, nodal status, metastasis, grade, oestrogen and progesterone receptor status, breast cancer detection type, diabetes, HRT use at diagnosis, study centre, energy intake, age at diagnosis
George et al. (2011)
Health, Eating, Activity, and Lifestyle (HEAL) / USA
670 w (older than 18)
6 Dietary indices: HEI
Post-diagnosis
FFQ usual diet approx. 2,5 yrs after diagnosis
Overall mortality, cancer-specific mortality
HEI: HR1= 0.40; 95% CI 0.17-0.94 HR2= 0.12; 95% CI 0.02-0.99
energy intake, physical activity, race, tumour stage, tamoxifen use, BMI
Kim et al. (2011)
Nurses' Health Study (NHS) / USA
2,729 w (30-55) not stated
Dietary indices: AHEI, DQIR, RFS, AMED
Post-diagnosis
FFQ usual diet around 1 year after diagnosis
Overall mortality, cancer specific mortality, death from other causes
AHEI: RR1= 0.85; 95% CI 0.63-1.17 RR2= 1.53; 95% CI 0.98-2.39 RR3= 0.52; 95% CI 0.32-0.83 DQIR: RR1= 0.78; 95% CI 0.58-1.07 RR2= 0.81; 95% CI 0.53-1.24 RR3= 0.85; 95% CI 0.54-1.34 RFS: RR1= 1.03; 95% CI 0.74-1.42 RR2= 1.54; 95% CI 0.95-2.47 RR3= 0.86; 95% CI 0.54-1.37 AMED: RR1= 0.87; 95% CI 0.64-1.17 RR2= 1.15; 95% CI 0.74-1.77 RR3= 0.80; 95% CI 0.50-1.26
age, time since diagnosis, alcohol intake (only for RFS because alcohol is a component in the other 3 diet quality indices), energy, multivitamin use (except for AHEI because it is a component), BMI, weight change (BMI at time of diet minus BMI just prior to diagnosis), oral contraceptive use, smoking status, physical activity in METs, stage, categories of treatment, age at first birth and parity, menopausal status and postmenopausal hormone use
Izano et al. (2013)
Nurses' Health Study (NHS) / USA
4,103 w (30-55) 9.3 Dietary indices: AHEI, DASH
Post-diagnosis
FFQ usual diet around 1 year after diagnosis
Cancer specific mortality, death from other causes
AHEI: RR2= 1.07; 95% CI 0.77-1.49 RR3= 0.57; 95% CI 0.42-0.77 DASH: RR2= 0.85; 95% CI 0.61-1.19 RR3= 0.72; 95% CI 0.53-0.99
stratified by time since diagnosis, adjusted for age at diagnosis, quintiles of energy intake, BMI and BMI change, age at first birth and parity, oral contraceptive use, menopausal status and HRT use, smoking, stage of disease, radiation treatment, chemotherapy and hormonal treatment, and physical activity
Kroenke et al. (2013)
Life After Cancer Epidemiology (LACE) study / USA
1,893 w (18-70) 11.8 Dairy Post-diagnosis
FFQ diet at diagnosis when cancer recurs before 6 yrs -
Overall mortality, cancer recurrence
Total Dairy: HR1= 1.39; 95% CI 1.02-1.90 HR4= 1.13; 95% CI 0.83-1.54 Low-fat dairy: HR1= 1.05; 95% CI 0.80-1.36
age, time between diagnosis and dietary assessment, high- and low-fat dairy intake, race, education, cancer stage at diagnosis, tumour size, human epidermal growth receptor 2, nodal and oestrogen receptor
Page 78 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
otherwise 6 years after diagnosis
HR4= 1.01; 95% CI 0.78-1.32 High-fat dairy: HR1= 1.64; 95% CI 1.24-2.17 HR4= 1.22; 95% CI 0.92-1.55
status, chemotherapy, radiation, tamoxifen, comorbidity, menopausal status, BMI, physical activity, energy intake, alcohol intake, red meat intake, fibre intake, fruit intake
Nechuta et al. (2013)
After Breast Cancer Pooling Project (includes cohorts SBCSS, LACE, WHEL, NHS) / USA and China
11,390 w (20-83) 9.0 Cruciferous vegetables Post-diagnosis
FFQ approx. 2 yrs after diagnosis
Overall mortality, cancer recurrence
Cruciferous vegetables: HR1= 0.99; 95% CI 0.86-1.13 HR4= 1.10; 95% CI 0.95-1.28
age at diagnosis, ER/PR status, TNM stage, surgery, chemotherapy, radiotherapy, hormonal therapy, smoking, BMI, exercise, menopausal status, race/ethnicity, education
Vrieling et al. (2013)
Mammary carcinoma Risk factor Investigation (MARIE) study / Germany
2,522 w 5.5 PCA: healthy pattern, unhealthy pattern
Pre-diagnosis
FFQ usual diet the year before diagnosis
Overall mortality, cancer specific mortality, death from other causes, cancer recurrence
Healthy pattern: HR1= 0.87; 95% CI 0.61-1.23 HR2= 0.89; 95% CI 0.59-1.35 HR3= 0.81; 95% CI 0.40-1.61 HR4= 0.71; 95% CI 0.48-1.06 Unhealthy pattern: HR1= 1.34; 95% CI 0.93-1.94 HR2= 0.99; 95% CI 0.64-1.52 HR3= 3.69; 95% CI 1.66-8.17 HR4= 0.91; 95% CI 0.61-1.36
tumour size, nodal status, metastases, tumour grade, ERPR status, radiotherapy, HRT use at diagnosis, mode of detection, and total energy intake and stratified by age at diagnosis and study centre
George et al. (2014)
Women’s Health Initiative’s Dietary Modification Trial and Observational Study (WHI) / USA
2,317 w (50-79) 9.6 Dietary indices: HEI
Post-diagnosis
FFQ usual diet approx. 1.5 yrs after diagnosis
Overall mortality, cancer specific mortality, death from other causes
HEI: HR1= 0.74; 95% CI 0.55-0.99 HR2= 0.91; 95% CI 0.60-1.40 HR3= 0.58; 95% CI 0.38-0.87
age at screening visit, WHI component, ethnicity, income, education, stage, estrogen receptor status, progesterone receptor status, time since diagnosis, energy intake in kcals, physical activity in MET, servings of alcohol per week, use of postmenopausal hormone therapy
McCullough et al. (2016)
Cancer Prevention Study II (CPS-II) Nutrition Cohort / USA
4,452 w for pre-diagnosis and 2,152 w for post-diagnosis (mean age 70.7 yrs)
9.8-9.9 Dietary indices: ACS Fruit and vegetables, red and processed meat
Pre- and post-diagnosis
FFQ usual diet in 1992 (before diagnosis) and usual diet at least 1 year after diagnosis (after diagnosis)
Overall mortality, cancer-specific mortality, death from other causes
Pre-diagnosis: ACS diet score RR1= 1.00; 95% CI 0.84-1.18 RR2= 1.06; 95% CI 0.79-1.42 RR3= 1.02; 95% CI 0.79-1.31 Fruit and vegetables RR1= 1.06; 95% CI 0.85-1.33 RR2= 1.00; 95% CI 0.66-1.50 RR3= 1.11; 95% CI 0.81-1.52 Red and processed meat: RR1= 0.88; 95% CI 0.73-1.06 RR2= 1.10; 95% CI 0.80-1.52 RR3= 0.81; 95% CI 0.62-1.07 Post-diagnosis: ACS diet score RR1= 0.93; 95% CI 0.73-1.18 RR2= 1.44; 95% CI 0.90-2.30 RR3= 0.78; 95% CI 0.56-1.07 Fruit and vegetables RR1= 1.03; 95% CI 0.80-1.33 RR2= 1.31; 95% CI 0.83-2.06 RR3= 0.93; 95% CI 0.65-1.34 Red and processed meat RR1= 0.64; 95% CI 0.49-0.84
age at diagnosis, diagnosis year, tumour stage at diagnosis, tumour grade at diagnosis, estrogen receptor status, progesterone receptor status, initial treatment (surgery, chemotherapy, radiation, hormone therapy, aromatase inhibitor use and/or Herceptin use), and the following assessed at the time of FFQ completion: BMI, smoking status, physical activity and energy intake
Page 79 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
RR2= 0.88; 95% CI 0.54-1.43 RR3= 0.57; 95% CI 0.39-0.82
Page 80 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Laryngeal cancer
Table S4: Summary of studies laryngeal cancer
Author (year) Study / country Number of participants / sex
(age)
Follow-up period (yrs)
Exposure Exposure timeframe Exposure assessment Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Crosignani et al. (1996)
Lombardy Cancer Registry (LCR) / Italy
213 m (32-75) 8-10 yr Meat (beef, veal), poultry, fish, eggs, milk, cheese, bread, pasta, potatoes, vegetables, citrus fruits, other fruits, butter, olive oil
Pre-diagnosis Interview usual diet in year before diagnosis
Overall mortality
Citrus fruits: HR1= 0.76; 95% 0.49-1.19 Other fruits: HR1= 0.65; 95% CI 0.39-1.07 Vegetables: HR1= 0.57; 95% CI 0.35-0.94 Meat: HR1= 0.50; 95% CI 0.30-0.83 Poultry: HR1= 0.90; 95% CI 0.55-1.46 Fish: HR1= 0.91; 95% CI 0.59-1.39 Eggs: HR1= 1.22; 95% CI 0.74-2.00 Milk: HR1= 1.58; 95% CI 0.99-2.55 Cheese: HR1= 0.70; 95% CI 0.44-1.12 Bread: HR1= 0.54; 95% CI 0.32-0.90 Pasta: HR1= 1.25; 95% CI 0.76-2.04 Potatoes: HR1= 1.02; 95% CI 0.64-1.64 Butter: HR1= 1.11; 95% CI 0.69-1.80 Olive oil: HR1= 0.71; 95% CI 0.44-1.16
age at diagnosis, clinical stage, occurrence of new primary cancers
Page 81 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Non-Hodgkin lymphoma (NHL)
Table S5: Summary of studies non-Hodgkin lymphoma
Author (year)
Study / country Number of participants /
sex (age)
Follow-up period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Han et al. (2010)
Yale Connecticut Tumor Registry New York (CTR) / USA
568 w (21-84) 7.7 Fruit, vegetables Pre-diagnosis
FFQ usual diet the year before diagnosis
Overall mortality, cancer-specific mortality
Total fruit and vegetables: HR1= 0.68; 95% CI 0.49-0.95 HR2= 0.70; 95% CI 0.45-1.10 Total fruit: HR1= 0.91; 95% CI 0.70-1.18 HR2= 1.04; 95% CI 0.74-1.45 Total vegetables: HR1= 0.58; 95% CI 0.38-0.89 HR2= 0.58; 95% CI 0.33-1.03 Cruciferous vegetables: HR1= 0.91; 95% CI 0.67–1.24 HR2= 0.75; 95% CI 0.49–1.14 Bean vegetables: HR1= 1.14; 95% CI 0.85-1.54 HR2= 1.05; 95% CI 0.71-1.55 Green leafy vegetables: HR1= 0.71; 95% CI 0.51-0.98 HR2= 0.82; 95% CI 0.54-1.23 Red vegetables: HR1= 1.03; 95% CI 0.76-1.38 HR2= 1.11; 95% CI 0.76-1.62 Yellow vegetables: HR1= 0.93; 95% CI 0.69-1.25 HR2= 1.11; 95% CI 0.77-1.61 Citrus fruits: HR1= 0.73; 95% CI 0.54-0.99 HR2= 0.81; 95% CI 0.54-1.20
age, education, stage, B-symptom, initial treatment, total energy intake
Leo et al. (2015)
Multi-ethnic Cohort (MEC) / USA
2,339 m/w (45-75)
4.5 Fruit, vegetables, dairy, legumes, fish, red meat
Pre-diagnosis
FFQ usual diet the year before diagnosis
Overall mortality, cancer specific mortality
Vegetables: HR1= 0.98; 95% CI 0.85-1.12 HR2= 0.98; 95% CI 0.83-1.16 Fruits: HR1= 1.03; 95% CI 0.90-1.19 HR2= 1.04; 95% CI 0.88-1.24 Red meat: HR1= 1.00; 95% CI 0.87-1.15 HR2= 0.95; 95% CI 0.81-1.13 Fish: HR1= 0.90; 95% CI 0.78-1.03 HR2= 0.91; 95% CI 0.76-1.08 Legumes: HR1= 0.88; 95% CI 0.76-1.01 HR2= 0.86; 95% CI 0.72-1.02 Dairy products:
age at NHL diagnosis, BMI, sex, ethnicity, SEER summary stage, NHL subtype, chemo-, radio-, immuno-, and steroid-therapy, smoking status at baseline, alcohol use, education status, energy intake, number of comorbidities
Page 82 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
HR1= 1.14; 95% CI 1.00-1.31 HR2= 1.16; 95% CI 0.98-1.37
Page 83 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Prostate cancer
Table S6: Summary of studies prostate cancer
Author (year)
Study / country Number of participants /
sex (age)
Follow-up period (yrs)
Exposure Exposure timeframe
Exposure assessment
Outcome Results (HR/RR and 95% CI) multivariate adjusted
Adjustment
Chavarro et al. (2008)
Physician’s Health Study (PHS) / USA
2,161 m 19 Total fish Pre-diagnosis
FFQ usual diet before diagnosis
Cancer-specific mortality
Total fish: HR2= 0.52; 95% CI 0.30-0.91
age at prostate cancer diagnosis, BMI, physical activity, alcohol use, tomato and dairy products, smoking, ethnicity, multivitamin and vitamin E supplements, random assignment to aspirin or beta-carotene, tumour stage, grade at diagnosis, clinical presentation of case
Kenfield et al. (2014)
Health Professionals Follow-up Study (HPFS) / USA
4,538 m (40-75) 23.2 Dietary indices: Mediterranean diet score (MDS)
Post-diagnosis
FFQ usual diet after diagnosis
Overall mortality, cancer-specific mortality
MDS: HR1= 0.78; 95% CI 0.67-0.90 HR2= 1.01; 95% CI 0.75-1.38
age at diagnosis, time period, time diagnosis to FFQ, energy, BMI, vigorous physical activity, smoking status, clinical stage, Gleason score, treatment
Yang et al. (2015a)
Physician’s Health Study (PHS) / USA
926 m (40-84) 9.6 Total dairy, high-fat dairy, low-fat dairy
Post-diagnosis
FFQ usual diet after diagnosis
Overall mortality Total dairy: HR1 = 1.76; 95% CI 1.21-2.55 HR2 = 2.41; 95% CI 0.96-6.02 High-fat dairy: HR1= 1.22; 95% CI 1.08-1.38 HR2= 1.30; 95% CI 0.97-1.73 Low-fat dairy: HR1= 1.17; 95% CI 1.05-1.29 HR2= 1.16; 95% CI 0.88-1.53
age at diagnosis, total energy intake, BMI, smoking status, exercise, Gleason score, clinical stage, prostate-specific antigen level, time interval between diagnosis and FFQ completion, initial treatment after diagnosis, family history of prostate cancer, and indicators for prudent dietary pattern and Western dietary pattern after excluding dairy products
Yang et al. (2015b)
Physician’s Health Study (PHS) / USA
926 m (40-84) 8.7 PCA: prudent diet, Western diet
Post-diagnosis
FFQ usual after diagnosis
Overall mortality, cancer specific mortality
Prudent diet: RR1= 0.64; 95% CI 0.44-0.93 RR2= 0.46; 95% CI 0.17-1.24 Western diet: RR1= 1.67; 95% CI 1.16-2.42 RR2= 2.53; 95% CI 1.00-6.42
age at diagnosis, total energy intake, BMI, smoking status, vigorous physical activity, Gleason score, clinical stage, prostate-specific antigen level, time interval between diagnosis and FFQ completion, initial treatment, family history of prostate cancer
HR1/RR1= overall mortality
HR2/RR2= cancer-specific mortality
HR3/RR3= death from other causes
HR4/RR4= cancer recurrence
FFQ= food frequency questionnaire
BMI= body mass index
Page 84 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Online supporting data File S3 for manuscript ‘the impact of dietary patterns
and the main food groups on mortality and recurrence in cancer survivors:
systematic review of current epidemiological literature’
Jochems et al., 11-08-2017
Page 85 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
The Grading of Recommendations Assessment, Development and Evaluation (GRADE)
GRADE is a systematic and explicit approach to making judgements about quality of
evidence and strength of recommendations. The focus is on clinical outcomes that patients
themselves are aware of in relation to their condition – in this systematic review these
include overall mortality, cancer-specific mortality, death from other causes, and cancer
recurrence. With the use of GRADE, the evidence is not rated study by study but across
studies for each individual outcome. Individual study quality was assessed with the
Cochrane Collaboration risk of bias assessment tools; the RoB 2.0 tool for randomised
trials and the ROBINS-I tool for cohort studies. Even before assessing the individual
study quality, studies were excluded from the systematic review if the sample size for the
analysis was <200 (comparisons containing less than 200 participants in total are
described as sparse data), the follow-up period was <4 years (for most cancer types, the
risk of cancer recurrence is the greatest within the first three years), no adjustments in the
statistical analysis were made for age and disease stage and, where possible, for cancer
treatment (e.g. studies adjusting for age and energy intake only were excluded).
Additionally, outcomes combining cancer recurrence with cancer progression, or
confirmed cancer-specific mortality combined with a diagnosis of metastasis, or prostate
cancer recurrence is determined by a rising PSA level only, were excluded. Therefore,
methodological flaws within the component studies will not cause any problems in the
GRADE evaluation - inconsistency of results across different studies will.
1. Quality of evidence
Table 1: Quality of Evidence Grades
Grade Definition
High We are very confident that the true effect lies close to that of the estimate of the effect.
Moderate We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect.
Very Low We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
2. Included study design
Table 2: Judgements about the quality of evidence by study design
Study Consequence
Randomised trials without important limitations – high quality evidence (+ 4 points)
Cohort studies without strengths or important limitations – low quality evidence (+ 2 points)
3. Determining the quality of evidence
Table 3: Factors that can reduce or increase the quality of the evidence
Factor Consequence
Limitations in study design or execution ↓ 1 or 2 levels
Inconsistency of results ↓ 1 or 2 levels
Indirectness of evidence ↓ 1 or 2 levels
Imprecision ↓ 1 or 2 levels
Publication bias ↓ 1 or 2 levels
Large magnitude of effect ↑ 1 or 2 levels
Dose-response gradient ↑ 1 level
All plausible confounding would reduce the demonstrated effect or increase the effect if no effect was observed
↑ 1 level
Page 86 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
4. Study limitations randomised trials
Table 4: Study limitations in randomised trials Factor Explanation
Lack of allocation concealment Those enrolling patients are aware of the group (or period in a crossover trial) to which the next enrolled patient will be allocated (a major problem in “pseudo” or “quasi” randomized trials with allocation by day of week, birth date, chart number, etc.)
Lack of blinding Patient, caregivers, those recording outcomes, those adjudicating outcomes, or data analysts are aware of the arm to which patients are allocated (or the medication currently being received in a crossover trial)
Incomplete accounting of patients and outcome events Loss to follow-up and failure to adhere to the intention-to-treat principle in superiority trials; or in non-inferiority trials, loss to follow-up, and failure to conduct both analyses considering only those who adhered to treatment, and all patients for whom outcome data are available. The significance of rates of loss to follow-up, however, varies widely and is dependent on the relation between loss to follow-up and number of events. The higher the proportion lost to follow-up in relation to intervention and control group event rates, and differences between intervention and control groups, the greater the threat of bias
Selective outcome reporting Incomplete or absent reporting of some outcomes and not others based on the results
Other limitations Stopping trial early for benefit. Substantial overestimates are likely in trials with fewer than 500 events and that large overestimates are likely in trials with fewer than 200 events. Empirical evidence suggests that formal stopping rules do not reduce this bias. Use of invalidated outcome measures (e.g. patient-reported outcomes). Carryover effects in crossover trial. Recruitment bias in cluster-randomized trials
5. Study limitations cohort studies
Table 5: Study limitations in observational studies Factor Explanation
Failure to develop and apply appropriate eligibility criteria (inclusion of control population)
Selection of exposed and unexposed in cohort studies from different populations
Flawed measurement of both exposure and outcome Differences in measurement of exposure
Differential surveillance for outcome in exposed and unexposed in cohort studies
Failure to adequately control confounding Failure of accurate measurement of all known prognostic factors
Failure to match for prognostic factors and/or adjustment in statistical analysis
Incomplete or inadequately short follow-up Especially within prospective cohort studies, both groups should be followed for the same amount of time
6. Grading assessors
A first assessor will grade the quality of evidence for each outcome (cancer recurrence or
overall mortality or cancer-specific mortality or death from other causes) for each cancer type
with data on pre- or post-diagnosis dietary patterns or foods as exposure (bladder cancer, bowel
cancer, breast cancer, laryngeal cancer, prostate cancer). The first assessor will summarize the
findings in summary of findings tables for all evidence obtained. A second assessor will check
the consistency of the ratings of the first assessor. Disagreement about evidence were resolved
through consensus or a third party.
Page 87 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
7. Summary of findings tables
Bladder cancer
Table 1: Bladder cancer and pre-diagnosis fruit and vegetable intake
Outcomes Hazard ratio / Relative risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total fruit: HR= 0.91; 95% CI 0.62-1.33
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total vegetables: HR= 0.91; 95% CI 0.62-1.36
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cruciferous vegetables: HR= 0.87; 95% CI 0.60-1.26
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Raw cruciferous vegetables: HR= 0.73; 95% CI 0.50-1.06
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total fruit: HR= 1.09; 95% CI 0.66-1.81
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total vegetables: HR= 1.06; 95% CI 0.63-1.78
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cruciferous vegetables: HR= 0.89; 95% CI 0.53-1.48
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Raw cruciferous vegetables: HR= 0.73; 95% CI 0.44-1.21
239 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bowel cancer
Table 2: Bowel cancer and pre-diagnosis diet quality indices
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality
HEI-2005: HR= 0.95; 95% CI 0.78-1.16
5727 (1) + Cohort ++ and downgraded one level; data of only 1 study
WCRF/AICR score: HR= 0.79; 95% CI 0.65-0.98
3292 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
HEI-2005: HR= 0.99; 95% CI 0.77-1.27
5727 (1) + Cohort ++ and downgraded one level; data of only 1 study
WCRF/AICR score: HR= 0.70; 95% CI 0.56–0.89
3292 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis diet quality indices
Overall mortality
AHEI-2010: HR= 0.71; 95% CI 0.52-0.98
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: HR= 0.98; 95% CI 0.71-1.35
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: HR= 0.87; 95% CI 0.63-1.21
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
MMDS: HR= 0.48; 95% CI 0.32-0.74
1404 (1) + Cohort ++ and downgraded one level; data of only 1 study. Although the study has a large estimate HR<0.5, it is based on 1 study only and will therefore be downgraded
HNFI: HR= 0.63; 95% CI 0.39-1.04
1404 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
AHEI-2010L HR= 0.72; 95% CI 0.43-1.21
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: HR= 0.87; 95% CI 0.52-1.45
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: HR= 0.84; 95% CI 0.50-1.42
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 88 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 3: Bowel cancer and pre-diagnosis prudent/healthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality
Prudent vegetable pattern: HR= 1.03; 95% CI 0.61-1.75
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Prudent vegetable pattern: HR= 1.12; 95% CI 0.69-1.84
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis prudent/healthy diet
Cancer recurrence
Prudent diet: HR= 1.13; 95% CI 0.77-1.67
1009 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality
Prudent diet: HR= 1.32; 95% CI 0.86-2.04 HR= 0.93; 95% CI 0.65-1.34
2210 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Cancer-specific mortality
Prudent diet: HR= 0.67; 95% CI 0.37-1.22
1201 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 4: Bowel cancer and pre-diagnosis Western/unhealthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Processed meat pattern: HR= 1.53; 95% CI 0.85-2.74
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
High sugar pattern: HR= 1.27; 95% CI 0.72-2.25
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Processed meat pattern: HR= 1.82; 95% CI 1.07-3.09
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
High sugar pattern: HR= 1.02; 95% CI 0.62-1.69
529 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis Western/unhealthy diet
Cancer recurrence Western diet: HR= 2.85; 95% CI 1.75-4.63
1009 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Western diet: HR= 2.32; 95% CI 1.36-3.96 HR= 1.32; 95% CI 0.89-1.97
2210 (2) + Cohort ++ and downgraded one level; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with an increased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as only one estimate is ‘statistically significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Cancer-specific mortality
Western diet: HR= 1.66; 95% CI 0.85-3.23
1201(1) + Cohort ++ and downgraded one level; data of only 1 study
Table 5: Bowel cancer and pre-diagnosis grain foods
Outcomes Hazard Ratio / Relative Risk (95% CI) No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Men Total whole grains: HR= 1.00; 95% CI 0.67-1.48 Women Total whole grains: HR= 0.91; 95% CI 0.60-1.39
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain wheat: HR= 0.97; 95% CI 0.64-1.49 Women Whole grain wheat: HR= 1.35; 95% CI 0.72-2.53
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain rye: HR= 0.90; 95% CI 0.60-1.36 Women Whole grain rye: HR= 0.93; 95% CI 0.60-1.46
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain oats: HR= 1.11; 95% CI 0.72-1.70 Women Whole grain oats: HR= 0.83; 95% CI 0.55-1.26
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Men Whole grain products: HR= 1.06; 95% CI 0.71-1.56 Women Whole grain products: HR= 1.10; 95% CI 0.74-1.64
1119 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 89 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 6: Bowel cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies) Quality of the evidence (GRADE) Comments
Cancer recurrence Red and processed meat: HR= 1.03; 95% CI 0.80-1.33
3122 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Unprocessed red meat: RR= 1.12; 95% CI 0.92-1.38 HR= 0.95; 95% CI 0.78-1.14
6104 (2) + Cohort ++ and downgraded one level; both studies found no statistically significant association, the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Red and processed meat: RR= 1.29; 95% CI 1.05-1.59 HR= 1.00; 95% CI 0.83-1.20 HR= 0.85; 95% CI 0.67-1.09
9226 (3) + Cohort ++ and downgraded one level; both studies found no statistically significant association, the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade one level for consistency lack of agreement between studies)
Poultry: HR= 0.87; 95% CI 0.73-1.03
3789 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer- specific mortality
Unprocessed red meat: RR= 1.16; 95% CI 0.84-1.58 HR= 0.93; 95% CI 0.75-1.15
6104 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Red and processed meat: RR= 1.09; 95% CI 0.79-1.51 HR= 1.00; 95% CI 0.81-1.23 HR= 0.83; 95% CI 0.61-1.14
9226 (3) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade for consistency lack of agreement between studies)
Poultry: HR= 0.91; 95% CI 0.75-1.10
3789 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Unprocessed red meat: RR= 1.19; 95% CI 0.87-1.64
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 1.39; 95% CI 1.00-1.92
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bowel cancer and post-diagnosis protein foods
Overall mortality Unprocessed red meat: RR= 0.75; 95% CI 0.55-1.03
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.94; 95% CI 0.68-1.30
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Unprocessed red meat: RR= 1.13; 95% CI 0.62-2.06
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 1.10; 95% CI 0.61-1.98
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Unprocessed red meat: RR= 0.64; 95% CI 0.40-1.03
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.87; 95% CI 0.54-1.41
2315 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 90 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 7: Bowel cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy: RR= 0.88; 95% CI 0.72-1.09 HR= 1.16; 95% CI 0.98-1.36
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Milk: HR= 1.21; 95% CI 1.03-1.43 RR= 0.95; 95% CI 0.79-1.15
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Yoghurt: HR= 1.08; 95% CI 0.92-1.28
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cheese: HR= 0.87; 95% CI 0.74-1.04
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: HR= 1.17; 95% CI 0.96-1.43 RR= 0.89; 95% CI 0.65-1.22
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade one level for consistency lack of agreement between studies)
Milk: HR= 1.21; 95% CI 0.99-1.48 RR= 0.98; 95% CI 0.73-1.32
6143 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade one level for consistency lack of agreement between studies)
Yoghurt: HR= 1.09; 95% CI 0.88-1.34
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cheese: HR= 0.93; 95% CI 0.76-1.14
3859 (1) + Cohort ++ and downgraded one level; data of only 1 study
Post-diagnosis dairy and alternatives
Overall mortality Total dairy: RR= 0.75; 95% CI 0.56-1.01
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Milk: RR= 0.72; 95% CI 0.55-0.94
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: RR= 0.73; 95% CI 0.44-1.23
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Milk: RR= 0.93; 95% CI 0.59-1.49
2284 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 91 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Breast cancer
Table 8: Breast cancer and pre-diagnosis diet quality indices
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality ACS: RR= 1.00; 95% CI 0.84-1.18
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
ACS: RR= 1.06; 95% CI 0.79-1.42
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
ACS: RR= 1.02; 95% CI 0.79-1.31
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis diet quality indices
Overall mortality HEI-2005: HR= 0.40; 95% CI 0.17-0.94 HR= 0.74; 95% CI 0.55-0.99
2987 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association and a decreased risk with overall mortality. Both studies have the same direction HR<1.0 (one study has a large estimate HR<0.5) (no downgrade nor upgrade of the evidence)
AHEI: RR= 0.85; 95% CI 0.63-1.17
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
DQIR: RR= 0.78; 95% CI 0.58-1.07
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
RFS: RR= 1.03; 95% CI 0.74-1.42
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: RR= 0.87; 95% CI 0.64-1.17
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
ACS: RR= 0.93; 95% CI 0.73-1.18
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
HEI-2005: HR= 0.12; 95% CI 0.02-0.99 HR= 0.91; 95% CI 0.60-1.40
2987 (2) + Cohort ++ and downgraded one level; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with a decreased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as only one estimate is ‘statistically significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant). In addition, there will be no upgrade of the evidence based on the other study that found an HR larger than <0.50 as it is only true for one of the studies
AHEI: RR= 1.53; 95% CI 0.98-2.39 RR= 1.07; 95% CI 0.77-1.49
6832 (2) + Cohort ++ and downgraded one level; both studies found a ‘statistically non-significant’ association and a RR> 1.0. However, as the estimate of one study is relatively high (RR=1.53) and the other study RR=1.07, which almost indicates no increased nor decreased risk of cancer-specific mortality, we made the decision to downgrade the evidence with one level for inconsistency of the results
DQIR: RR= 0.81; 95% CI 0.53-1.24
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
RFS: RR= 1.54; 95% CI 0.95-2.47
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: RR= 1.15; 95% CI 0.74-1.77
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
DASH: RR= 0.85; 95% CI 0.61-1.19
4103 (1) + Cohort ++ and downgraded one level; data of only 1 study
ACS: RR= 1.44; 95% CI 0.90-2.30
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
HEI-2005: HR= 0.58; 95% CI 0.38-0.87
2317 (1) + Cohort ++ and downgraded one level; data of only 1 study
AHEI: RR= 0.52; 95% CI 0.32-0.83 RR= 0.57; 95% CI 0.42-0.77
6832 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association in the same direction HR>1.0 (one study even has a large estimate HR<0.5). Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
DQIR: RR= 0.85; 95% CI 0.54-1.34
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
RFS: RR= 0.86; 95% CI 0.54-1.37
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
AMED: RR= 0.80; 95% CI 0.50-1.26
2729 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 92 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
DASH: RR= 0.72; 95% CI 0.53-0.99
4103 (1) + Cohort ++ and downgraded one level; data of only 1 study
ACS: RR= 0.78; 95% CI 0.56-1.07
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 9: Breast cancer and post-diagnosis low-fat diet RCTs
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Intervention versus control HR= 0.76; 95% CI 0.60-0.98
2437 (1) +++ RCT ++++ and downgraded by one level; data of only 1 study (even though it is a RCT with a low risk of bias)
Overall mortality Intervention versus control HR= 0.89; 95% CI 0.65-1.21 HR= 0.91; 95% CI 0.72-1.15
5525 (2) +++ RCT ++++ and downgraded by one level; although both studies present a HR<1 for overall mortality, both estimates are statistically non-significant. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as both estimates are ‘statistically non-significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Table 10: Breast cancer and pre-diagnosis prudent/healthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Prudent diet: HR= 0.71; 95% CI 0.48-1.06
2522 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Prudent diet: Not shown HR= 0.87; 95% CI 0.61-1.23
5141 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Cancer-specific mortality
Prudent diet: Not shown HR= 0.89; 95% CI 0.59-1.35
5141 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Death from other causes
Prudent diet: Not shown HR= 0.81; 95% CI 0.40-1.61
5141 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Breast cancer and post-diagnosis prudent/healthy diet
Cancer recurrence Prudent diet: HR= 0.95; 95% CI 0.63-1.43
1901 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Prudent diet: RR= 0.78; 95% CI 0.54-1.12 HR= 0.57; 95% CI 0.36-0.90
4520 (2) + Cohort ++ and downgraded one level; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with a decreased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as only one estimate is ‘statistically significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Cancer-specific mortality
Prudent diet: RR= 1.07; 95% CI 0.66-1.73 HR= 0.79; 95% CI 0.43-1.43
4520 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR=1.0 (downgrade one level for consistency lack of agreement between studies)
Death from other causes
Prudent diet: RR= 0.54; 95% CI 0.31-0.95 HR= 0.35; 95% CI 0.17-0.73
4520 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association with a decreased risk of death from other cause in the same direction HR<1.0 (one study even has a large estimate HR<0.5). Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
Page 93 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 11: Breast cancer and pre-diagnosis Western/unhealthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Western diet: HR= 0.91; 95% CI 0.61-1.36
2522 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Western diet: RR= 1.40; 95% CI 0.93-2.09 HR= 1.34; 95% CI 0.93-1.94
5141 (2) + Cohort ++ and downgraded one level; both studies found a ‘statistically non-significant’ association. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as both estimates are ‘statistically non-significant’ and only two studies have investigated the association, we did decide to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Cancer-specific mortality
Western diet: RR= 1.01; 95% CI 0.59-1.72 HR= 0.99; 95% CI 0.64-1.52
5141 (2) + Cohort ++ and downgraded one level; although both studies found a ‘statistically non-significant’ association and indicate no increase nor decrease with cancer-specific mortality, there is no reason for downgrading the evidence
Death from other causes
Western diet: RR= 1.95; 95% CI 1.06-3.60 HR= 3.69; 95% CI 1.66-8.17
5141 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association with an increased risk of death from other cause in the same direction HR>1.0 (one study even has a large estimate HR>2.0). Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies, even though the confidence intervals are broad
Breast cancer and post-diagnosis Western/unhealthy diet
Cancer recurrence Western diet: HR= 0.98; 95% CI 0.62-1.54
1901 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Western diet: RR= 1.53; 95% CI 1.03-2.29 HR= 1.76; 95% CI 1.10-2.81
4520 (2) ++ Cohort ++; both studies found a ‘statistically significant’ association with an increased risk of death from other cause in the same direction HR>1.0. Therefore, we will not downgrade nor upgrade the evidence as both HRs are in the same direction and the overlap in confidence intervals of both studies
Cancer-specific mortality
Western diet: RR= 1.01; 95% CI 0.60-1.70 HR= 1.20; 95% CI 0.62-2.32
4520 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR=1.0 (downgrade for consistency lack of agreement between studies)
Death from other causes
Western diet: RR= 2.09; 95% CI 1.30-3.36 HR= 2.15; 95% CI 0.97-4.77
4520 (2) + Cohort ++ and downgraded one level; 1 study found no ‘statistically significant’ association whilst the other found a ‘statistically significant’ association with a decreased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as only one estimate is ‘statistically significant’ and only two studies have investigated the association, we decided to downgrade the evidence despite both estimates are in the same direction, are large (HR>2), and the overlap in confidence intervals of both studies – although the 95% CI are wide (downgrade for effect on whether estimates are significant)
Page 94 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 12: Breast cancer and pre-diagnosis fruit and vegetables
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total vegetables: RR= 0.98; 95% CI 0.62-1.53 HR= 0.57; 95% CI 0.35-0.94 HR= 1.09; 95% CI 0.80-1.48
4673 (3) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 and HR= 1.0 (almost) downgrade for consistency lack of agreement between studies)
Total fruit: HR= 0.63; 95% CI 0.38-1.05 HR= 0.84; 95% CI 0.61-1.16
3169 (2) + Cohort ++ and downgraded one level; both studies found a ‘statistically non-significant’ association. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. The GRADE guidelines do not recommend downgrading for statistically non-significant results - however, as both estimates are ‘statistically non-significant’ and only two studies have investigated the association, we decided to downgrade the evidence despite both estimates are in the same direction and that there is an overlap in confidence intervals of both studies (downgrade for effect on whether estimates are significant)
Total fruit + vegetables: HR= 1.27; 95% CI 1.00–1.61 (low versus high intake!) RR= 1.06; 95% CI 0.85-1.33 (high versus low intake!)
5905 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Cancer-specific mortality
Total vegetables: Not shown HR= 1.01; 95% CI 0.70-1.46
4157 (2) + Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Total fruit: HR= 0.86; 95% CI 0.59-1.25
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit + vegetables: HR= 1.26; 95% CI 0.96–1.64 (low versus high intake!) RR= 1.00; 95% CI 0.66-1.50
5905 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR=1 (downgrade for consistency lack of agreement between studies)
Death from other causes
Total fruit + vegetables: RR= 1.11; 95% CI 0.81-1.52
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis fruit and vegetables
Cancer recurrence Cruciferous vegetables: HR= 1.10; 95% CI 0.95-1.28
11390 (1) + Cohort ++ and downgraded one level; data of only 1 study although large sample because participants of 4 cohort studies combined
Overall mortality Total vegetables: RR= 0.81; 95% CI 0.59–1.11 HR= 1.44; 95% CI 0.91-2.27
6423 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Cruciferous vegetables: HR= 1.02; 95% CI 0.80-1.30 HR= 0.99; 95% CI 0.86-1.13
15831 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies)
Total fruit: HR= 1.38; 95% CI 0.88-2.17
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit + vegetables: RR= 1.03; 95% CI 0.80-1.33
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total vegetables: Not shown HR= 0.96; 95% CI 0.38-2.45
6423 (2) + Cohort ++ and downgraded one level; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Cruciferous vegetables: HR= 0.95; 95% CI 0.59-1.54
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit: HR= 1.39; 95% CI 0.64-2.99
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit + vegetables: RR= 1.31; 95% CI 0.83-2.06
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Total fruit + vegetables: RR= 0.93; 95% CI 0.65-1.34
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 13: Breast cancer and pre-diagnosis grain foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Bread: HR= 1.31; 95% CI 0.93-1.83
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Bread: HR= 1.10; 95% CI 0.74-1.63
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 95 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 14: Breast cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Poultry RR= 0.60; 95% CI 0.39-0.92
1504 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish RR= 0.94; 95% CI 0.62-1.43
1504 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red meat: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Red and processed meat: RR= 0.88; 95% CI 0.73-1.06
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sunflower/pumpkinseeds: HR= 0.87; 95% CI 0.66-1.15
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sesame/flaxseeds: HR= 0.90; 95% CI 0.68-1.19
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Poultry: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Fish: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Red meat: Not shown
1504 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Red and processed meat: RR= 1.10; 95% CI 0.80-1.52
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sunflower/pumpkinseeds: HR= 1.12; 95% CI 0.79-1.57
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Sesame/flaxseeds: HR= 1.21; 95% CI 0.87-1.68
2653 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Red and processed meat: RR= 0.81; 95% CI 0.62-1.07
4452 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis protein foods
Cancer recurrence Red meat: RR= 1.12; 95% CI 0.66-1.89
472 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Red meat: RR= 1.06; 95% CI 0.76–1.49
1982 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total meat (poultry, fish, beef, and processed meat): HR= 1.12; 95% CI 0.83-1.51
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.64; 95% CI 0.49-0.84
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Red meat: RR= 2.60; 95% CI 0.96-7.03 Not shown
2454 (2)
+ Cohort ++ and downgraded; data available of only 1 study as the results of the other study are not shown in the article (downgrade one level for publication bias)
Poultry: Not show
1982 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Fish: Not shown
1982 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Total meat (poultry, fish, beef, and processed meat): HR= 0.89; 95% CI 0.50-1.60
4441 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red and processed meat: RR= 0.88; 95% CI 0.54-1.43
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Death from other causes
Red and processed meat: RR= 0.57; 95% CI 0.39-0.82
2152 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 96 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 15: Breast cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy RR= 0.71; 95% CI 0.44-1.14
1504 (1) + Cohort ++ and downgraded one level; data of only 1 study
Breast cancer and post-diagnosis dairy and alternatives
Cancer recurrence
Total dairy: HR= 1.13; 95% CI 0.83-1.54
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
Low fat dairy: HR= 1.01; 95% CI 0.78-1.32
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
High fat dairy: HR= 1.22; 95% CI 0.92-1.55
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
Overall mortality Total dairy: RR= 0.72; 95% CI 0.52-1.00 HR= 1.39; 95% CI 1.02-1.90 HR= 1.18; 95% CI 0.90-1.54
3875 (3) + Cohort ++ and downgraded one level; 2 studies found no ‘statistically significant’ association whilst one study found a ‘statistically significant’ association with an increased risk of overall mortality. We strongly believe that ‘statistically non-significant’ (p-value>0.05) results should always be interpreted together with the sample size and number of events, size of the RR/HR, and the confidence intervals around the estimate to make the judgement on whether results have an impact or not. Nevertheless, we will downgrade the evidence with two HRs in the same direction and the third with in an opposite direction (downgrade for consistency lack of agreement between studies)
Low fat dairy: HR= 1.05; 95% CI 0.80-1.36
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
High fat dairy: HR= 1.64; 95% CI 1.24-2.17
1893 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: Not shown HR= 0.94; 95% CI 0.56-1.59
1982 (1) + Cohort ++ and downgraded; results of the study are not shown in the article (downgrade one level for publication bias)
Table 16: Breast cancer and post-diagnosis oils and spreads
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer recurrence Butter/margarine/lard: RR= 1.30; 95% CI 1.03-1.64
472 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Butter/margarine/lard: RR= 1.03; 95% CI 0.61-1.76
472 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 97 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Laryngeal cancer
Table 17: Laryngeal cancer and pre-diagnosis fruit and vegetables
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Citrus fruits: HR= 0.76; 95% 0.49-1.19
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Other fruits: HR= 0.65; 95% CI 0.39-1.07
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Vegetables: HR= 0.57; 95% CI 0.35-0.94
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 18: Laryngeal cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Meat: HR= 0.50; 95% CI 0.30-0.83
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Poultry: HR= 0.90; 95% CI 0.55-1.46
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish: HR= 0.91; 95% CI 0.59-1.39
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Eggs: HR= 1.22; 95% CI 0.74-2.00
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 19: Laryngeal cancer and pre-diagnosis grain foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Bread: HR= 0.54; 95% CI 0.32-0.90
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Pasta: HR= 1.25; 95% CI 0.76-2.04
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Potatoes: HR= 1.02; 95% CI 0.64-1.64
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 20: Laryngeal cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Milk: HR= 1.58; 95% CI 0.99-2.55
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cheese: HR= 0.70; 95% CI 0.44-1.12
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 21: Laryngeal cancer and pre-diagnosis oils and spreads
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Butter: HR= 1.11; 95% CI 0.69-1.80
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Olive oil: HR= 0.71; 95% CI 0.44-1.16
213 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 98 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Non-Hodgkin Lymphoma
Table 22: Non-Hodgkin Lymphoma and pre-diagnosis fruit and vegetables
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total fruit and vegetables: HR= 0.68; 95% CI 0.49-0.95
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit: HR= 0.91; 95% CI 0.70-1.18 HR= 1.03; 95% CI 0.90-1.19
2907 (2) + Cohort ++ and downgraded one level; the estimates are inconsistent HR>1 and HR<1 (downgrade for consistency lack of agreement between studies). Additionally, one study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Total vegetables: HR= 0.58; 95% CI 0.38-0.89 HR= 0.98; 95% CI 0.85-1.12
2907 (2) + Cohort ++ and downgraded one level; one study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Cruciferous vegetables: HR= 0.91; 95% CI 0.67–1.24
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bean vegetables: HR= 1.14; 95% CI 0.85-1.54
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Green leafy vegetables: HR= 0.71; 95% CI 0.51-0.98
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red vegetables: HR= 1.03; 95% CI 0.76-1.38
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Yellow vegetables: HR= 0.93; 95% CI 0.69-1.25
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Citrus fruits: HR= 0.73; 95% CI 0.54-0.99
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality Total fruit and vegetables: HR= 0.70; 95% CI 0.45-1.10
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Total fruit: HR= 1.04; 95% CI 0.74-1.45 HR= 1.04; 95% CI 0.88-1.24
2907 (2) + Cohort ++ and downgraded one level; one study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Total vegetables: HR= 0.58; 95% CI 0.33-1.03 HR= 0.98; 95% CI 0.83-1.16
2907 (2) + Cohort ++ and downgraded one level; one study analysed women only whilst the other study analysed men and women together (downgrade one level for directness)
Cruciferous vegetables: HR= 0.75; 95% CI 0.49–1.14
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Bean vegetables: HR= 1.05; 95% CI 0.71-1.55
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Green leafy vegetables: HR= 0.82; 95% CI 0.54-1.23
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Red vegetables: HR= 1.11; 95% CI 0.76-1.62
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Yellow vegetables: HR= 1.11; 95% CI 0.77-1.61
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Citrus fruits: HR= 0.81; 95% CI 0.54-1.20
568 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 23: Non-Hodgkin Lymphoma and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Red meat: HR= 1.00; 95% CI 0.87-1.15
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish: HR= 0.90; 95% CI 0.78-1.03
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Legumes: HR= 0.88; 95% CI 0.76-1.01
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality Red meat: HR= 0.95; 95% CI 0.81-1.13
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Fish: HR= 0.91; 95% CI 0.76-1.08
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Legumes: HR= 0.86; 95% CI 0.72-1.02
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 99 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Table 24: Non-Hodgkin Lymphoma and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy: HR= 1.14; 95% CI 1.00-1.31
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality Total dairy: HR= 1.16; 95% CI 0.98-1.37
2339 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 100 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
Prostate cancer
Table 25: Prostate cancer and post-diagnosis diet quality indices
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Med diet score: HR= 0.78; 95% CI 0.67-0.90
4538 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Med diet score: HR= 1.01; 95% CI 0.75-1.38
4538 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 26: Prostate cancer and post-diagnosis prudent/healthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Prudent diet: RR= 0.64; 95% CI 0.44-0.93
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Prudent diet: RR= 0.46; 95% CI 0.17-1.24
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 27: Prostate cancer and post-diagnosis Western/unhealthy diet
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Western diet: RR= 1.67; 95% CI 1.16-2.42
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Western diet: RR= 2.53; 95% CI 1.00-6.42
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 28: Prostate cancer and pre-diagnosis protein foods
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Cancer-specific mortality
Fish: HR= 0.52; 95% CI 0.30-0.91
2161 (1) + Cohort ++ and downgraded one level; data of only 1 study
Table 29: Prostate cancer and pre-diagnosis dairy and alternatives
Outcomes Hazard Ratio / Relative Risk (95% CI)
No of participants (studies)
Quality of the evidence (GRADE)
Comments
Overall mortality Total dairy: HR = 1.76; 95% CI 1.21-2.55
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
High-fat dairy: HR= 1.22; 95% CI 1.08-1.38
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Low-fat dairy: HR= 1.17; 95% CI 1.05-1.29
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Cancer-specific mortality
Total dairy: HR = 2.41; 95% CI 0.96-6.02
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
High-fat dairy: HR= 1.30; 95% CI 0.97-1.73
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Low-fat dairy: HR= 1.16; 95% CI 0.88-1.53
926 (1) + Cohort ++ and downgraded one level; data of only 1 study
Page 101 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on Decem
ber 7, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2016-014530 on 19 February 2018. D
ownloaded from
For peer review only
PRISMAPRISMAPRISMAPRISMA ChecklistChecklistChecklistChecklist
Section/topic # Checklist item Reported on page #
TITLE
Title 1 Identify the report as a systematic review, meta-analysis, or both. 1
ABSTRACT
Structured summary 2 Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number.
2
INTRODUCTION
Rationale 3 Describe the rationale for the review in the context of what is already known. 4
Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).
5
METHODS
Protocol and registration 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.
9
Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered,
language, publication status) used as criteria for eligibility, giving rationale. 5-8
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.
5
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.
37
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable,
included in the meta-analysis). 6
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.
5
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.
5-8
Risk of bias in individual studies
12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.
8-9
Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means). 10
Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency
(e.g., I2) for each meta-analysis.
9-10
Page 102 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
PRISMAPRISMAPRISMAPRISMA ChecklistChecklistChecklistChecklist
Page 1 of 2
Section/topic # Checklist item Reported on page #
Risk of bias across studies 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within studies).
8-9
Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-specified.
NA
RESULTS
Study selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.
9 and flowchart in supplemental file
Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations.
Supplemental file
Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12). Data on risk of bias of each study can be obtained on request
Results of individual studies 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.
Supplemental file
Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency. NA
Risk of bias across studies 22 Present results of any assessment of risk of bias across studies (see Item 15). Supplemental file
Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]).
NA
DISCUSSION
Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).
21-23
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).
24
Page 103 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from
For peer review only
PRISMAPRISMAPRISMAPRISMA ChecklistChecklistChecklistChecklist
Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research.
24-25
FUNDING
Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for the systematic review.
25-26
From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097
For more information, visit: www.prisma-statement.org.
Page 2 of 2
Page 104 of 104
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on December 7, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-014530 on 19 February 2018. Downloaded from