bmj open · for peer review only breast cancer table 8: breast cancer and pre-diagnosis diet...

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

[email protected]

on Decem

ber 7, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-014530 on 19 February 2018. D

ownloaded from

http://bmjopen.bmj.com/

mailto:[email protected]


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



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

[email protected]

Institute of Cancer and Genomic Sciences, University of Birmingham,

B15 2TT Birmingham, United Kingdom

Page 1 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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

.


Page 10 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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.


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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

.


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



18


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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

.


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.


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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.


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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

.


Page 22 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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



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



Page 43 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




45



Page 45 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




47



Page 47 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Table S2: Summary of findings on a prudent / Mediterranean diet adherence and mortality in cancer survivors


exposure

assessment

Number

of cases /

sex (m/w)

Follow-up

period

(median

or

average)

Outcome

measures

Results


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




(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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


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



intake, BMI, smoking status,

Page 55 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




(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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Table S3: Summary of findings on a Western diet adherence and mortality in cancer survivors


exposure

assessment

Number

of cases /

sex (m/w)

Follow-up

(median

or

average)

Primary

outcome

Results


Adjustments

Bowel cancer

Meyerhardt (2007)

(Meyerhardt et al.,

2007)

Cancer and

Leukemia Group

B (CALGB)

study / United

States

Post-

diagnosis /

FFQ


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




physical activity level, and

time-varying total calories.

Zhu (2013) (Zhu et al.,

2013)

Familial CRC

registry in

Newfoundland /

Canada

Pre-diagnosis

/ FFQ


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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


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


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


intake, ethnicity, BMI at

enrolment, weight change

from before diagnosis to

baseline, smoking,

Page 59 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


mortality and

prostate cancer

mortality

RR1= 1.67; 95% CI 1.16–2.42


RR2= 2.53; 95% CI 1.00-6.42



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960





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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


Adjustments

Bladder cancer

Tang (2010)

(Tang et al.,

2010)

Roswell Park

Cancer

Institute

(RPCI) /

United States

Pre-diagnosis

/ FFQ


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Saxe (1999)

(Saxe et al.,

1999)

Michigan

University

Follow-up

Study / United

States

Pre- diagnosis /

FFQ


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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,


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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Project / China

and United

States

1.13 surgery, chemotherapy,

radiotherapy, hormonal

therapy, smoking, BMI,

exercise, menopausal

status, ethnicity, education


Han (2010)

(Han et al.,

2010a)

Yale

Connecticut

Tumor

Registry New

York (CTR) /

United States

Pre-diagnosis /

FFQ


mortality,

cancer-specific

mortality


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-


age, education, stage, B-

symptom, initial treatment,

total energy intake

Page 67 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


mortality


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


mortality

Citrus fruit:

HR1= 0.76; 95% CI 0.49-

age at diagnosis, clinical

stage, occurrence of new

Page 68 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




(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



Page 69 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Table S5: Summary of findings on meat and fish consumption and mortality in cancer survivors


country

Timeframe /

exposure

assessment

Number of

cases / sex

(m/w)

Follow-up period

(median or

average)

Outcome

measures

Results


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


post-diagnosis:

RR1= 1.02; 95% CI 0.76-

1.38

RR2= 1.28; 95% CI 0.74-

2.21


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


stage at diagnosis, and

post-diagnostic energy

intake (sex-specific

quartiles), weight change

Page 70 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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-


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-


between 1992 pre-

diagnostic, post-

diagnostic questionnaires,

and 1992 pre-diagnostic

meat intake.

Combined pre- and post-

diagnosis: age at


stage at diagnosis, 1992

pre-diagnostic energy

intake, and post-

diagnostic energy intake

Page 71 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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,



birth and parity, number

Page 73 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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





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


stage and occurrence of

new primaries

Page 74 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




ethnicity, multivitamin

and vitamin E

supplements, random

assignment to aspirin or

beta-carotene, tumour

stage, grade at diagnosis,

clinical presentation of

case



Page 76 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Table S6: Summary of findings on dairy consumption and mortality in cancer survivors


country

Timeframe /

exposure

assessment

Number of

cases / sex

(m/w)

Follow-up period

(median or

average)

Outcome

measures

Results


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



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)





calendar year of diagnosis,

body mass index, oral

contraceptive use,

menopausal status,



Page 77 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960







Kroenke (2013)

(Kroenke et al.,

2013)

Life After

Cancer

Epidemiology

(LACE) study

/ United States

Post-diagnosis /

FFQ


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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



Page 79 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Table S7: Summary of findings on coffee and tea consumption and mortality in cancer survivors


country

Timeframe /

exposure

assessment

Number of

cases / sex

(m/w)

Follow-up period

(median or

average)

Outcome

measures

Results


Adjustments

Bladder cancer

Wakai (1993)

(Wakai et al.,

1993)

Study in

metropolitan

Nagoya / Japan

Pre-diagnosis /

FFQ


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




(Dray et al.,

2003)

Côte d’Or area

/ France

FFQ mortality RR1= 1.46; 95% CI 0.64-

3.22


intake

Breast cancer

Harris (2012)

(Harris et al.,

2012b)

Swedish

Mammography

Cohort

(SMC) /

Sweden

Post-diagnosis /

FFQ


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



Page 81 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Table S8: Summary of findings on alcohol consumption and mortality in cancer survivors


country

Timeframe /

exposure

assessment

Number of

cases / sex

(m/w)

Follow-up period

(median or

average)

Outcome

measures

Results


Adjustments

Bladder cancer

Wakai (1993)

(Wakai et al.,

1993)

Study in

metropolitan

Nagoya / Japan

Pre- and post-

diagnosis / FFQ


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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Park (2006)

(Park et al.,

2006)

National

Health

Insurance

Corporation

Study

(NHICS) /

Korea

Pre-diagnosis /

FFQ


mortality

HR1= 0.92; 95% CI 0.72-

1.19


BMI, fasting serum

glucose level,




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-


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


95% CI 0.65–1.57 for

age, stage, treatment,

clinically relevant factors

Page 83 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960





Liquor:

HR1= 1.13; 95% CI 0.56–

2.29 for familial CRC


95% CI 0.51–1.24 for


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




(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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


mortality,

cancer-specific

mortality

HR1= 1.19; 95% CI 0.87-

1.62

HR2= 1.51; 95% CI 1.00-

2.29


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




McDonald

(2002)

(McDonald et

al., 2002)

Grampian

University

Hospitals

Follow-up

Study / United

States

Pre- diagnosis /

FFQ


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


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


mortality,

HR1= 0.99; 95% CI 0.80-

1.22

age at diagnosis, year of

diagnosis, TNM stage,

Page 88 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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,


Page 90 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960







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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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)





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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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–


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Serur (1995)

(Serur et al.,

1995)

Brooklyn and

Kings

County study /

United States

Pre- diagnosis /

FFQ


specific

mortality

HR2= 1.4; 95% CI 0.7-1.9 age <50 yr, advanced

stage, smoking, heroin and

cocaine use, public

hospital admission


Battaglioli

(2006)

(Battaglioli et

al., 2006)

Follow-up of

patients who

were originally

recruited in a

case-control

study / Italy

Pre-diagnosis /

FFQ


mortality

HR1= 1.41; 95% CI 1.10–


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


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


mortality

HR1= 0.90; 95% CI 0.70–

1.17

age, education, smoking,

disease stage, B-symptom

Page 94 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


mortality

HR1= 0.86; 95% CI 0.42-

1.76


BMI, fasting serum

glucose level,




Page 95 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Korea comorbidities

Laryngeal cancer

Crosignani

(1996)

(Crosignani et

al., 1996)

Lombardy

Cancer

Registry

(LCR) / Italy

Pre-diagnosis /

FFQ


mortality

HR1= 1.12; 95% CI 0.70-

1.80


stage, occurrence of new

primaries

Prostate cancer

Park (2006)

(Park et al.,

2006)

National

Health

Insurance

Corporation

Study

(NHICS) /

Korea

Pre-diagnosis /

FFQ


mortality

HR1= 1.85; 95% CI 0.79-

4.34


BMI, fasting serum

glucose level,




comorbidities



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


mortality,

HR1= 0.84; 95% CI 0.65-

1.08

age at diagnosis,

race/ethnicity, education,

Page 97 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


mortality

Pre-diagnosis:

HR1= statistically non-

significant (not shown)

Post-diagnosis:

HR1= 1.07; 95% CI 0.80 –

1.43





calendar year of diagnosis,

body mass index, oral

contraceptive use,

menopausal status,






Kwan (2011)

(Kwan et al.,

Life After

Cancer

Post-diagnosis

and combined


mortality,

Post-diagnosis:

HR1= 0.92; 95% CI 0.70-

age at diagnosis, ethnicity,

education, stage positive

Page 99 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




2011) Epidemiology

(LACE) study

/ United States

pre- and post-

diagnosis / FFQ

cancer-specific

mortality

1.20


diagnosis:

HR1= 0.79; 95% CI 0.56-

1.12

Post-diagnosis:

HR2= 0.88; 95% CI 0.61-

1.28


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960






Page 101 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from





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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from




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


For more information, visit: www.prisma-statement.org.

Page 2 of 2

Page 110 of 110


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


Epidemiology

Secondary Subject Heading: Oncology, Public health, Epidemiology

Keywords: cancer survivors, mortality, cancer recurrence, food, dietary pattern, diet


BMJ Open on D



jopen.bmj.com

/B

MJ O


ebruary 2018. Dow

nloaded from



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

[email protected] 20

Institute of Cancer and Genomic Sciences, University of Birmingham, 21

B15 2TT Birmingham, United Kingdom 22

23

Page 1 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


for dietary patterns/indices and foods of the main food groups amongst prostate cancer 457

survivors. 458

459

Page 20 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


Page 26 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


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 - - - -


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




34


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




35


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from

http://creativecommons.org/licenses/by-nc-nd/4.0/




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


this variable was

unnecessary?*

Is the confounding

domain measured

validly and

reliably by this

variable (or these

variables)?

OPTIONAL: Is


this variable

(alone) expected to

favour the

experimental

intervention or the

Page 57 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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







(ii) Additional co-interventions relevant to the setting of this particular study, or which the

study authors identified as important




not administered)?













Page 59 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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 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?


1.6. Did the authors control for any post-intervention

variables that could have been affected by the intervention?


Questions relating to baseline and time-varying confounding


controlled for all the important confounding domains and

for time-varying confounding?






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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


influenced by the outcome or a cause of the

outcome?



2.4. Do start of follow-up and start of intervention

coincide for most participants?


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?


Risk of bias judgement Low / Moderate / Serious / Critical /

NI


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?


3.3 Could classification of intervention status have

been affected by knowledge of the outcome or risk

of the outcome?


Risk of bias judgement Low / Moderate / Serious / Critical / NI

Optional: What is the predicted direction of bias

due to classification of interventions?


comparator / Towards null /Away from

null / Unpredictable

Page 61 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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?


5.2 Were participants excluded due to missing data

on intervention status?



on other variables needed for the analysis?


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?


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?



Optional: What is the predicted direction of bias due

to missing data?




Page 62 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



27

Bias in measurement of outcomes

6.1 Could the outcome measure have been

influenced by knowledge of the intervention

received?


6.2 Were outcome assessors aware of the

intervention received by study participants?


6.3 Were the methods of outcome assessment

comparable across intervention groups?


6.4 Were any systematic errors in measurement

of the outcome related to intervention

received?



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?


7.2 ... multiple analyses of the intervention-

outcome relationship?


7.3 ... different subgroups? Y / PY / PN / N / NI



bias due to selection of the reported result?


Towards null /Away from null / Unpredictable

Overall bias


Optional: What is the overall predicted

direction of bias for this outcome?





Page 63 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from





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.


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



29

Page 65 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



30

Domain Signalling questions Response options

Description/Support

for judgement

Bias arising from

the

randomization

process

1.1 Was the allocation

sequence random?



sequence concealed until

participants were

recruited and assigned to

interventions?


1.3 Were there baseline

imbalances that suggest a

problem with the

randomization process?


Risk of bias judgement Low / High / Some

concerns

Optional: What is the

predicted direction of

bias arising from the


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?


2.2. Were carers and trial

personnel aware of

participants' assigned


trial?


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?


2.5. Did study

participants adhere to the

assigned intervention

regimen?


Page 66 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



31


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


concerns



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?


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


concerns



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?


Page 67 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



32


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


concerns



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?


5.2 ... multiple analyses of the data?



concerns



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


bias for this outcome?

Favours

experimental /

Favours comparator

/ Towards null

/Away from null /

Unpredictable

Page 68 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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’


Page 70 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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 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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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



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


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



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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)


2,284 m/w (40-92) 7.5 Total dairy, milk Pre- and post-diagnosis



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


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)


3,292 m/w (25-70) 4.2 Dietary indices: WCRF/AICR score

Pre-diagnosis



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



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Breast cancer

Table S3: Summary of studies breast cancer

Author (year)

Study / country Number of participants /

sex (age)

Follow-up

period (yrs)


Exposure assessment



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


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)


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)


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)


2,619 w (30-55) 9 PCA: prudent diet, Western diet


FFQ usual intake 4 years before diagnosis and FFQ at least one year after diagnosis


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



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


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)


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



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


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)


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)


4,103 w (30-55) 9.3 Dietary indices: AHEI, DASH

Post-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



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


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


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


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


FFQ usual diet in 1992 (before diagnosis) and usual diet at least 1 year after diagnosis (after diagnosis)


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Laryngeal cancer

Table S4: Summary of studies laryngeal cancer

Author (year) Study / country Number of participants / sex

(age)


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from




Table S5: Summary of studies non-Hodgkin lymphoma

Author (year)


sex (age)



Exposure assessment


Adjustment

Han et al. (2010)

Yale Connecticut Tumor Registry New York (CTR) / USA

568 w (21-84) 7.7 Fruit, vegetables Pre-diagnosis



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Prostate cancer

Table S6: Summary of studies prostate cancer

Author (year)


sex (age)



Exposure assessment


Adjustment

Chavarro et al. (2008)

Physician’s Health Study (PHS) / USA

2,161 m 19 Total fish Pre-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



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)


926 m (40-84) 9.6 Total dairy, high-fat dairy, low-fat dairy

Post-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)


926 m (40-84) 8.7 PCA: prudent diet, Western diet

Post-diagnosis

FFQ usual after diagnosis


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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’


Page 84 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


Cruciferous vegetables: HR= 0.87; 95% CI 0.60-1.26


Raw cruciferous vegetables: HR= 0.73; 95% CI 0.50-1.06



Total fruit: HR= 1.09; 95% CI 0.66-1.81








Bowel cancer

Table 2: Bowel cancer and pre-diagnosis diet quality indices

Outcomes Hazard Ratio / Relative Risk (95% CI)



Comments

Overall mortality

HEI-2005: HR= 0.95; 95% CI 0.78-1.16


WCRF/AICR score: HR= 0.79; 95% CI 0.65-0.98



HEI-2005: HR= 0.99; 95% CI 0.77-1.27


WCRF/AICR score: HR= 0.70; 95% CI 0.56–0.89


Post-diagnosis diet quality indices

Overall mortality

AHEI-2010: HR= 0.71; 95% CI 0.52-0.98


DASH: HR= 0.98; 95% CI 0.71-1.35


AMED: HR= 0.87; 95% CI 0.63-1.21


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



AHEI-2010L HR= 0.72; 95% CI 0.43-1.21


DASH: HR= 0.87; 95% CI 0.52-1.45


AMED: HR= 0.84; 95% CI 0.50-1.42


Page 87 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Table 3: Bowel cancer and pre-diagnosis prudent/healthy diet




Comments

Overall mortality

Prudent vegetable pattern: HR= 1.03; 95% CI 0.61-1.75





Post-diagnosis prudent/healthy diet

Cancer recurrence

Prudent diet: HR= 1.13; 95% CI 0.77-1.67


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)




Table 4: Bowel cancer and pre-diagnosis Western/unhealthy diet




Comments

Overall mortality Processed meat pattern: HR= 1.53; 95% CI 0.85-2.74


High sugar pattern: HR= 1.27; 95% CI 0.72-2.25



Processed meat pattern: HR= 1.82; 95% CI 1.07-3.09




Post-diagnosis Western/unhealthy diet

Cancer recurrence

Western diet: HR= 2.85; 95% CI 1.75-4.63


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



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)


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


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


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


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


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


Page 88 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Table 6: Bowel cancer and pre-diagnosis protein foods




Comments

Cancer recurrence Red and processed meat: HR= 1.03; 95% CI 0.80-1.33


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



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




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


Death from other causes

Processed meat: RR= 1.27; 95% CI 0.92-1.75


Unprocessed red meat: RR= 1.19; 95% CI 0.87-1.64


Red and processed meat: RR= 1.39; 95% CI 1.00-1.92


Post-diagnosis protein foods

Overall mortality Processed meat: RR= 1.11; 95% CI 0.82-1.49




















Page 89 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Table 7: Bowel cancer and pre-diagnosis dairy and alternatives




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


Yoghurt: HR= 1.08; 95% CI 0.92-1.28


Cheese: HR= 0.87; 95% CI 0.74-1.04



Total dairy: HR= 1.17; 95% CI 0.96-1.43 RR= 0.89; 95% CI 0.65-1.22


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


Cheese: HR= 0.93; 95% CI 0.76-1.14


Post-diagnosis dairy and alternatives

Overall mortality Total dairy: RR= 0.75; 95% CI 0.56-1.01


Milk: RR= 0.72; 95% CI 0.55-0.94



Total dairy: RR= 0.73; 95% CI 0.44-1.23


Milk: RR= 0.93; 95% CI 0.59-1.49


Page 90 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Breast cancer

Table 8: Breast cancer and pre-diagnosis diet quality indices




Comments

Overall mortality ACS: RR= 1.00; 95% CI 0.84-1.18



ACS: RR= 1.06; 95% CI 0.79-1.42



ACS: RR= 1.02; 95% CI 0.79-1.31


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


DQIR: RR= 0.78; 95% CI 0.58-1.07


RFS: RR= 1.03; 95% CI 0.74-1.42


AMED: RR= 0.87; 95% CI 0.64-1.17


ACS: RR= 0.93; 95% CI 0.73-1.18



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


RFS: RR= 1.54; 95% CI 0.95-2.47


AMED: RR= 1.15; 95% CI 0.74-1.77


DASH: RR= 0.85; 95% CI 0.61-1.19


ACS: RR= 1.44; 95% CI 0.90-2.30



HEI-2005: HR= 0.58; 95% CI 0.38-0.87


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


RFS: RR= 0.86; 95% CI 0.54-1.37


AMED: RR= 0.80; 95% CI 0.50-1.26


DASH: RR= 0.72; 95% CI 0.53-0.99


ACS: RR= 0.78; 95% CI 0.56-1.07


Page 91 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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




Comments

Cancer recurrence Prudent diet: HR= 0.71; 95% CI 0.48-1.06


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)


Prudent diet: Not shown HR= 0.89; 95% CI 0.59-1.35





Breast cancer and post-diagnosis prudent/healthy diet



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


Prudent diet: RR= 1.07; 95% CI 0.66-1.73 HR= 0.79; 95% CI 0.43-1.43




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Table 11: Breast cancer and pre-diagnosis Western/unhealthy diet




Comments

Cancer recurrence Western diet: HR= 0.91; 95% CI 0.61-1.36


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


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



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




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



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)



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Table 12: Breast cancer and pre-diagnosis fruit and vegetables




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!)



Total vegetables: Not shown HR= 1.01; 95% CI 0.70-1.46


Total fruit: HR= 0.86; 95% CI 0.59-1.25


Total fruit + vegetables: HR= 1.26; 95% CI 0.96–1.64 (low versus high intake!)


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


Cruciferous vegetables: HR= 1.02; 95% CI 0.80-1.30 HR= 0.99; 95% CI 0.86-1.13


Total fruit: HR= 1.38; 95% CI 0.88-2.17







Total fruit: HR= 1.39; 95% CI 0.64-2.99


Table 13: Breast cancer and pre-diagnosis grain foods




Comments

Overall mortality Bread: HR= 1.31; 95% CI 0.93-1.83



Bread: HR= 1.10; 95% CI 0.74-1.63


Page 94 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Table 14: Breast cancer and pre-diagnosis protein foods




Comments

Overall mortality Poultry RR= 0.60; 95% CI 0.39-0.92


Fish RR= 0.94; 95% CI 0.62-1.43


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


Sesame/flaxseeds: HR= 0.90; 95% CI 0.68-1.19



Poultry: Not shown


Fish: Not shown


Red meat: Not shown






Breast cancer and post-diagnosis protein foods

Cancer recurrence Red meat: RR= 1.12; 95% CI 0.66-1.89


Overall mortality Red meat: RR= 1.06; 95% CI 0.76–1.49


Total meat (poultry, fish, beef, and processed meat): HR= 1.12; 95% CI 0.83-1.51



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


Fish: Not shown




Page 95 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Table 15: Breast cancer and pre-diagnosis dairy and alternatives




Comments

Overall mortality Total dairy RR= 0.71; 95% CI 0.44-1.14 HR= 1.18; 95% CI 0.90-1.54



Total dairy: HR= 0.94; 95% CI 0.56-1.59


Breast cancer and post-diagnosis dairy and alternatives

Cancer recurrence Total dairy: HR= 1.13; 95% CI 0.83-1.54


Low fat dairy: HR= 1.01; 95% CI 0.78-1.32


High fat dairy: HR= 1.22; 95% CI 0.92-1.55


Overall mortality Total dairy RR= 0.72; 95% CI 0.52–1.00 HR= 1.39; 95% CI 1.02-1.90







Total dairy: Not shown


Table 16: Breast cancer and post-diagnosis oils and spreads




Comments

Cancer recurrence Butter/margarine/lard: RR= 1.30; 95% CI 1.03-1.64



Butter/margarine/lard: RR= 1.03; 95% CI 0.61-1.76


Page 96 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Laryngeal cancer

Table 17: Laryngeal cancer and pre-diagnosis fruit and vegetables




Comments

Overall mortality Citrus fruits: HR= 0.76; 95% 0.49-1.19


Other fruits: HR= 0.65; 95% CI 0.39-1.07


Vegetables: HR= 0.57; 95% CI 0.35-0.94


Table 18: Laryngeal cancer and pre-diagnosis protein foods




Comments

Overall mortality Meat: HR= 0.50; 95% CI 0.30-0.83


Poultry: HR= 0.90; 95% CI 0.55-1.46


Fish: HR= 0.91; 95% CI 0.59-1.39


Eggs: HR= 1.22; 95% CI 0.74-2.00


Table 19: Laryngeal cancer and pre-diagnosis grain foods




Comments



Pasta: HR= 1.25; 95% CI 0.76-2.04


Potatoes: HR= 1.02; 95% CI 0.64-1.64


Table 20: Laryngeal cancer and pre-diagnosis dairy and alternatives




Comments

Overall mortality Milk: HR= 1.58; 95% CI 0.99-2.55


Cheese: HR= 0.70; 95% CI 0.44-1.12


Table 21: Laryngeal cancer and pre-diagnosis oils and spreads




Comments

Overall mortality Butter: HR= 1.11; 95% CI 0.69-1.80


Olive oil: HR= 0.71; 95% CI 0.44-1.16


Page 97 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Non-Hodgkin Lymphoma

Table 22: Non-Hodgkin Lymphoma and pre-diagnosis fruit and vegetables




Comments

Overall mortality Total fruit and vegetables: HR= 0.68; 95% CI 0.49-0.95



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


Bean vegetables: HR= 1.14; 95% CI 0.85-1.54


Green leafy vegetables: HR= 0.71; 95% CI 0.51-0.98


Red vegetables: HR= 1.03; 95% CI 0.76-1.38


Yellow vegetables: HR= 0.93; 95% CI 0.69-1.25


Citrus fruits: HR= 0.73; 95% CI 0.54-0.99


Cancer-specific mortality Total fruit and vegetables: HR= 0.70; 95% CI 0.45-1.10


















Table 23: Non-Hodgkin Lymphoma and pre-diagnosis protein foods




Comments

Overall mortality Red meat: HR= 1.00; 95% CI 0.87-1.15


Processed meat: HR= 0.94; 95% CI 0.82-1.08


Fish: HR= 0.90; 95% CI 0.78-1.03


Legumes: HR= 0.88; 95% CI 0.76-1.01


Cancer-specific mortality Red meat: HR= 0.95; 95% CI 0.81-1.13


Processed meat: HR= 0.94; 95% CI 0.79-1.12


Fish: HR= 0.91; 95% CI 0.76-1.08


Legumes: HR= 0.86; 95% CI 0.72-1.02


Page 98 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Table 24: Non-Hodgkin Lymphoma and pre-diagnosis dairy and alternatives




Comments

Overall mortality Soy foods: HR= 0.93; 95% CI 0.76-1.14


Total dairy: HR= 1.14; 95% CI 1.00-1.31


Cancer-specific mortality Soy foods: HR= 0.92; 95% CI 0.72-1.18


Total dairy: HR= 1.16; 95% CI 0.98-1.37


Page 99 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Prostate cancer

Table 25: Prostate cancer and post-diagnosis diet quality indices




Comments

Overall mortality Med diet score: HR= 0.78; 95% CI 0.67-0.90



Med diet score: HR= 1.01; 95% CI 0.75-1.38


Table 26: Prostate cancer and post-diagnosis prudent/healthy diet




Comments

Overall mortality Prudent diet: RR= 0.64; 95% CI 0.44-0.93



Prudent diet: RR= 0.46; 95% CI 0.17-1.24


Table 27: Prostate cancer and post-diagnosis Western/unhealthy diet




Comments

Overall mortality Western diet: RR= 1.67; 95% CI 1.16-2.42



Western diet: RR= 2.53; 95% CI 1.00-6.42


Table 28: Prostate cancer and pre-diagnosis protein foods




Comments


Fish: HR= 0.52; 95% CI 0.30-0.91


Table 29: Prostate cancer and pre-diagnosis dairy and alternatives




Comments

Overall mortality Total dairy: HR = 1.76; 95% CI 1.21-2.55


High-fat dairy: HR= 1.22; 95% CI 1.08-1.38


Low-fat dairy: HR= 1.17; 95% CI 1.05-1.29



Total dairy: HR = 2.41; 95% CI 0.96-6.02






Page 100 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



PRISMA Checklist


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


6-8


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


6


6-8



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



PRISMA Checklist



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


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


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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



Page 2 of 2

Page 103 of 103


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



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


Epidemiology




BMJ Open on D



jopen.bmj.com

/B

MJ O


ebruary 2018. Dow

nloaded from






4



7

(1) Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, 8

United Kingdom 9


University, Maastricht, The Netherlands 11


University of Birmingham, Birmingham, United Kingdom 13


Maastricht, The Netherlands 15

16

17






23

Page 1 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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




- Most studies investigating dietary patterns/indices and foods before diagnosis do not 56




60

Page 3 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



4

INTRODUCTION 61




cancer and increase overall life expectancy[1,2], the suggestion that epigenetic aberrations 65










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







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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



6

110


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










for cancer treatment. Excluded papers did not state hazard ratios (HRs) or relative risks 123


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




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



7

Cancer Research Fund and the American Institute for Cancer Research (WCRF/AIRC) 135




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





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


vegetable oils, spreads. Although processed (red) meats are not included in the main food 159

Page 7 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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







176

Assessment risk of bias and level of quality 177




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



10


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





Page 10 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



11




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





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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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






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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



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








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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



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


Page 21 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



22



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



23










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



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



24










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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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








598



601

Funding 602

Page 25 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



26


for-profit sectors. This systematic review has not as yet been registered. 604

605



608

Page 26 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



2013;178:590–602. doi:10.1093/aje/kwt006 620



2015;27:588–600. doi:10.1177/1010539515595860 623





doi:10.1017/S0007114515001798 628



2007;15:621–30. doi:10.1007/s00520-006-0207-6 631



doi:10.1002/cncr.29488 634





doi:10.3322/caac.21142 639






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


Page 27 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



28






doi:10.1093/aje/kwp393 655











doi:10.1200/JCO.2012.45.4462 666








doi:10.1016/S0002-8223(99)00168-6 674



2008;168:713. doi:10.1001/archinte.168.7.713 677







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



Nutr 2014;17:2769–82. doi:10.1017/S1368980013003169 689

Page 28 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



29







2017). 696










doi:10.1136/bmj.39489.470347.AD 706







2015. 713



doi:10.1158/1055-9965.EPI-10-0008 716



2007;298:754. doi:10.1001/jama.298.7.754 719



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






Page 29 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



30


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












doi:10.1080/01635581.2014.847472 748






2013;31:2773–82. doi:10.1200/JCO.2013.49.1126 754






Jama 2007;298:289. doi:10.1001/jama.298.3.289 760







2014;23:575–83. doi:10.1158/1055-9965.EPI-13-1162 767



2013;65:820–6. doi:10.1080/01635581.2013.804939 770


Page 30 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



31







26. doi:10.1200/JCO.2008.19.4035 778



doi:10.1038/bjc.2012.521 781




doi:10.1207/s15327914nc5502_3 785



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



0142(19990901)86:5<826::AID-CNCR19>3.0.CO;2-0 795



doi:10.1093/jnci/djt027 798



doi:10.1007/s10549-010-1323-z 801



2011;105:1151–7. doi:10.1038/bjc.2011.374 804



doi:10.1023/A:1006056915001 807






Page 31 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



32

doi:10.3109/10428191003690364 813






2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 819



2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 822



doi:10.1016/j.eururo.2013.08.009 825






2015;137:2462–9. doi:10.1002/ijc.29608 831







2015;218:59–70. doi:10.1242/jeb.107110 838


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



34




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 - - - -


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




35




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 - - - -



population

Page 35 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




36


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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




37





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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960








Page 38 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



2




- 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)


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



3



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





Exclusion criteria

Animal studies

In vitro studies


Inclusion criteria



Exclusion criteria



Exclusion criteria


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]


Page 40 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



4






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.







Present results for any of the following outcomes: • Overall mortality • Cancer-specific mortality • Death from other causes • Cancer recurrence

Page 41 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



5




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.


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



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



6



Page 43 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



7


Page 44 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



8


Page 45 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



9



Page 46 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



10


Page 47 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



11



Page 48 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



12


Page 49 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



13



Page 50 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



14



Page 51 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



15


Page 52 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



16


Page 53 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



17



Page 54 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



18


Page 55 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



19






Participants

Experimental

intervention

Comparator

Outcomes



impact on outcomes

Page 56 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from





20




over)

Participants

Experimental

intervention

Comparator


Specify the outcome







Page 57 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



21






Confounding

domain

Measured

variable(s)

Is there evidence


this variable was

unnecessary?*

Is the confounding

domain measured

validly and

reliably by this

variable (or these

variables)?

OPTIONAL: Is


this variable

(alone) expected to

favour the

experimental

intervention or the

comparator?

Yes / No / No

information

Favour

experimental /

Favour comparator

/ No information



Confounding

domain

Measured

variable(s)

Is there evidence


this variable was

unnecessary?*

Is the confounding

domain measured

validly and

reliably by this

variable (or these

variables)?

OPTIONAL: Is


this variable

(alone) expected to

favour the

experimental

intervention or the

Page 58 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



22

comparator?

Yes / No / No

information

Favour

experimental /

Favour comparator

/ No information



Page 59 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



23








not administered)?


















not administered)?













Page 60 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



24




formatting is used.


options






be considered

Y / PY / PN / N








NA / Y / PY /

PN / N / NI







NA / Y / PY /

PN / N / NI






















Critical / NI


confounding?



Unpredictable

Page 61 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



25













outcome?











NI










intervention?




of the outcome?








Page 62 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



26






Y / PY / PN /

N / NI



NA / Y / PY /

PN / N / NI




N / NI


N / NI


N / NI



NA / Y / PY /

PN / N / NI






participants?














missing data?




to missing data?




Page 63 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



27




received?










received?







Unpredictable
















Overall bias








Page 64 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from





28




Study design



of bias






assessed.





Trial protocol









Research)



Page 65 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



29

Page 66 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



30


Description/Support

for judgement

Bias arising from

the

randomization

process


sequence random?




participants were


interventions?




problem with the




concerns





Favours

experimental /

Favours comparator

/ Towards null

/Away from null /

Unpredictable

Bias due to

deviations from

intended

interventions




trial?



personnel aware of



trial?





across intervention

groups?

NA / Y / PY / PN /

N / NI


implemented

successfully?


2.5. Did study



regimen?


Page 67 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



31


Description/Support

for judgement

2.6. If N/PN/NI to 2.3,

2.4 or 2.5: Was an




the intervention?

NA / Y / PY / PN /

N / NI


concerns




from intended

interventions?

Favours

experimental /

Favours comparator

/ Towards null

/Away from null /

Unpredictable

Bias due to

missing outcome

data




randomized?







across intervention

groups?

NA / Y / PY / PN /

N / NI


there evidence that


presence of missing

outcome data?

NA / Y / PY / PN /

N / NI


concerns



bias due to missing

outcome data?

Favours

experimental /

Favours comparator

/ Towards null

/Away from null /

Unpredictable

Bias in

measurement of

the outcome



Page 68 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



32


Description/Support

for judgement


NA / Y / PY / PN /

N / NI


concerns




of the outcome?

Favours

experimental /

Favours comparator

/ Towards null

/Away from null /

Unpredictable

Bias in selection

of the reported

result







concerns





Favours

experimental /

Favours comparator

/ Towards null

/Away from null /

Unpredictable


concerns

Optional:

What is the overall



Favours

experimental /

Favours comparator

/ Towards null

/Away from null /

Unpredictable

Page 69 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



33






Scre

enin

g In

clu

ded

El

igib

ility

Id

enti

fica

tio

n


(n= 8)











(n= 38)

Page 70 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from






Page 71 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Bladder cancer


Author (year)




Exposure assessment



Adjustment

Tang et al. (2010)







Page 72 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Bowel cancer



Follow-up

period (yrs)


Exposure assessment


Adjustment




Post-diagnosis







2,315 m/w (40-93) 7.5 Red and processed meat, unprocessed red meat




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


Zhu et al. (2013)


529 m/w (20-75) 6.4 PCA: prudent vegetable pattern

Pre-diagnosis



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)



Pre-diagnosis



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from




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)



Post-diagnosis








Pre-diagnosis



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)



Pre-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


Page 74 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Yang et al. (2014)







Carr et al. (2016)



Pre-diagnosis




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




Pre-diagnosis





Ward et al. (2016)


3,789 m/w (25-70) 4.1 Red meat, unprocessed meat, poultry

Pre-diagnosis



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




Post-diagnosis




Page 75 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Breast cancer


Author (year)


sex (age)

Follow-up

period (yrs)


Exposure assessment



Adjustment




Post-diagnosis






















Page 76 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from













Post-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





Post-diagnosis











Kwan et al. (2009)



Post-diagnosis








Post-diagnosis



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


Page 77 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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)



Pre-diagnosis









Post-diagnosis



HEI: HR1= 0.40; 95% CI 0.17-0.94 HR2= 0.12; 95% CI 0.02-0.99


Kim et al. (2011)


2,729 w (30-55) not stated


Post-diagnosis





Izano et al. (2013)



Post-diagnosis








FFQ diet at diagnosis when cancer recurs before 6 yrs -


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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











Pre-diagnosis








Post-diagnosis








9.8-9.9 Dietary indices: ACS Fruit and vegetables, red and processed meat




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


Page 79 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



RR2= 0.88; 95% CI 0.54-1.43 RR3= 0.57; 95% CI 0.39-0.82

Page 80 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Laryngeal cancer



(age)



Adjustment





Overall mortality



Page 81 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Non-Hodgkin lymphoma (NHL)


Author (year)


sex (age)



Exposure assessment


Adjustment

Han et al. (2010)







Leo et al. (2015)


2,339 m/w (45-75)

4.5 Fruit, vegetables, dairy, legumes, fish, red meat

Pre-diagnosis



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:


Page 82 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



HR1= 1.14; 95% CI 1.00-1.31 HR2= 1.16; 95% CI 0.98-1.37

Page 83 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Prostate cancer


Author (year)


sex (age)



Exposure assessment


Adjustment






Total fish: HR2= 0.52; 95% CI 0.30-0.91





Post-diagnosis



MDS: HR1= 0.78; 95% CI 0.67-0.90 HR2= 1.01; 95% CI 0.75-1.38


Yang et al. (2015a)



Post-diagnosis




Yang et al. (2015b)



Post-diagnosis











Page 84 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from







Page 85 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from









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.



Grade Definition







Study Consequence





Factor Consequence









↑ 1 level

Page 86 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from





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





Flawed measurement of both exposure and outcome Differences in measurement of exposure













Page 87 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



7. Summary of findings tables

Bladder cancer





Comments










Total fruit: HR= 1.09; 95% CI 0.66-1.81








Bowel cancer





Comments

Overall mortality

HEI-2005: HR= 0.95; 95% CI 0.78-1.16





HEI-2005: HR= 0.99; 95% CI 0.77-1.27





Overall mortality

AHEI-2010: HR= 0.71; 95% CI 0.52-0.98


DASH: HR= 0.98; 95% CI 0.71-1.35


AMED: HR= 0.87; 95% CI 0.63-1.21


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



AHEI-2010L HR= 0.72; 95% CI 0.43-1.21


DASH: HR= 0.87; 95% CI 0.52-1.45


AMED: HR= 0.84; 95% CI 0.50-1.42


Page 88 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from







Comments

Overall mortality







Cancer recurrence



Overall mortality










Comments














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)







Comments











Page 89 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from





No of participants (studies) Quality of the evidence (GRADE) Comments



Overall mortality Unprocessed red meat: RR= 1.12; 95% CI 0.92-1.38 HR= 0.95; 95% CI 0.78-1.14




Poultry: HR= 0.87; 95% CI 0.73-1.03


Cancer- specific mortality





Poultry: HR= 0.91; 95% CI 0.75-1.10







Bowel cancer and post-diagnosis protein foods

Overall mortality Unprocessed red meat: RR= 0.75; 95% CI 0.55-1.03














Page 90 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from







Comments



Milk: HR= 1.21; 95% CI 1.03-1.43 RR= 0.95; 95% CI 0.79-1.15


Yoghurt: HR= 1.08; 95% CI 0.92-1.28


Cheese: HR= 0.87; 95% CI 0.74-1.04





Milk: HR= 1.21; 95% CI 0.99-1.48 RR= 0.98; 95% CI 0.73-1.32


Yoghurt: HR= 1.09; 95% CI 0.88-1.34


Cheese: HR= 0.93; 95% CI 0.76-1.14





Milk: RR= 0.72; 95% CI 0.55-0.94



Total dairy: RR= 0.73; 95% CI 0.44-1.23


Milk: RR= 0.93; 95% CI 0.59-1.49


Page 91 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Breast cancer





Comments




ACS: RR= 1.06; 95% CI 0.79-1.42



ACS: RR= 1.02; 95% CI 0.79-1.31




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


DQIR: RR= 0.78; 95% CI 0.58-1.07


RFS: RR= 1.03; 95% CI 0.74-1.42


AMED: RR= 0.87; 95% CI 0.64-1.17


ACS: RR= 0.93; 95% CI 0.73-1.18



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


DQIR: RR= 0.81; 95% CI 0.53-1.24


RFS: RR= 1.54; 95% CI 0.95-2.47


AMED: RR= 1.15; 95% CI 0.74-1.77


DASH: RR= 0.85; 95% CI 0.61-1.19


ACS: RR= 1.44; 95% CI 0.90-2.30



HEI-2005: HR= 0.58; 95% CI 0.38-0.87


AHEI: RR= 0.52; 95% CI 0.32-0.83 RR= 0.57; 95% CI 0.42-0.77


DQIR: RR= 0.85; 95% CI 0.54-1.34


RFS: RR= 0.86; 95% CI 0.54-1.37


AMED: RR= 0.80; 95% CI 0.50-1.26


Page 92 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



DASH: RR= 0.72; 95% CI 0.53-0.99


ACS: RR= 0.78; 95% CI 0.56-1.07






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)


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)





Comments















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)







Page 93 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from







Comments




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)



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














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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from







Comments


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)


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!)





Total fruit: HR= 0.86; 95% CI 0.59-1.25


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)


Total fruit + vegetables: RR= 1.11; 95% CI 0.81-1.52









Total fruit: HR= 1.38; 95% CI 0.88-2.17






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)



Total fruit: HR= 1.39; 95% CI 0.64-2.99











Comments




Bread: HR= 1.10; 95% CI 0.74-1.63


Page 95 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from







Comments



Fish RR= 0.94; 95% CI 0.62-1.43


Red meat: Not shown









Poultry: Not shown


Fish: Not shown


Red meat: Not shown






















2454 (2)


Poultry: Not show


Fish: Not shown









Page 96 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from







Comments

Overall mortality Total dairy RR= 0.71; 95% CI 0.44-1.14



Cancer recurrence

Total dairy: HR= 1.13; 95% CI 0.83-1.54






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)






Total dairy: Not shown HR= 0.94; 95% CI 0.56-1.59






Comments






Page 97 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Laryngeal cancer





Comments





Vegetables: HR= 0.57; 95% CI 0.35-0.94






Comments



Poultry: HR= 0.90; 95% CI 0.55-1.46


Fish: HR= 0.91; 95% CI 0.59-1.39


Eggs: HR= 1.22; 95% CI 0.74-2.00






Comments



Pasta: HR= 1.25; 95% CI 0.76-2.04


Potatoes: HR= 1.02; 95% CI 0.64-1.64






Comments



Cheese: HR= 0.70; 95% CI 0.44-1.12






Comments



Olive oil: HR= 0.71; 95% CI 0.44-1.16


Page 98 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from








Comments






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)



































Comments



Fish: HR= 0.90; 95% CI 0.78-1.03


Legumes: HR= 0.88; 95% CI 0.76-1.01




Fish: HR= 0.91; 95% CI 0.76-1.08


Legumes: HR= 0.86; 95% CI 0.72-1.02


Page 99 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from







Comments

Overall mortality Total dairy: HR= 1.14; 95% CI 1.00-1.31


Cancer-specific mortality Total dairy: HR= 1.16; 95% CI 0.98-1.37


Page 100 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



Prostate cancer





Comments










Comments










Comments










Comments


Fish: HR= 0.52; 95% CI 0.30-0.91






Comments








Total dairy: HR = 2.41; 95% CI 0.96-6.02






Page 101 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



PRISMAPRISMAPRISMAPRISMA ChecklistChecklistChecklistChecklist


TITLE


ABSTRACT


2

INTRODUCTION



5

METHODS


9




5


37




5


5-8



8-9


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960





Page 1 of 2



8-9


NA

RESULTS


9 and flowchart in supplemental file


Supplemental file



Supplemental file


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


21-23


24

Page 103 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960





Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research.

24-25

FUNDING


25-26



Page 2 of 2

Page 104 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




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



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


Epidemiology




BMJ Open on D



jopen.bmj.com

/B

MJ O


ebruary 2018. Dow

nloaded from






4



7

(1) Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, 8

United Kingdom 9


University, Maastricht, The Netherlands 11


University of Birmingham, Birmingham, United Kingdom 13


Maastricht, The Netherlands 15

16

17






23

Page 1 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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




- Most studies investigating dietary patterns/indices and foods before diagnosis do not 56




60

Page 3 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



4

INTRODUCTION 61




cancer and increase overall life expectancy[1,2]. The suggestion that epigenetic aberrations 65










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







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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



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


Page 5 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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










for cancer treatment. Excluded papers did not state hazard ratios (HRs) or relative risks 122


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




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


Page 6 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



7



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





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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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







175

Assessment risk of bias and level of quality 176




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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


and cancer recurrence is provided. Additionally, tables with the number of studies for pre-210

Page 9 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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





Page 10 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



11




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





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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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






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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



439



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








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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



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


Page 21 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



22



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



23










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



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



24










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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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








598



601

Funding 602


for-profit sectors. This systematic review has not as yet been registered. 604

Page 25 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



26

605



608

Page 26 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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



2013;178:590–602. doi:10.1093/aje/kwt006 618



2015;27:588–600. doi:10.1177/1010539515595860 621





doi:10.1017/S0007114515001798 626



2007;15:621–30. doi:10.1007/s00520-006-0207-6 629



doi:10.1002/cncr.29488 632





doi:10.3322/caac.21142 637






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




Page 27 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



28




doi:10.1093/aje/kwp393 653











doi:10.1200/JCO.2012.45.4462 664








doi:10.1016/S0002-8223(99)00168-6 672



2008;168:713. doi:10.1001/archinte.168.7.713 675







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



Nutr 2014;17:2769–82. doi:10.1017/S1368980013003169 687



Page 28 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



29





2017). 694










doi:10.1136/bmj.39489.470347.AD 704







2015. 711



doi:10.1158/1055-9965.EPI-10-0008 714



2007;298:754. doi:10.1001/jama.298.7.754 717



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







doi:10.3945/jn.116.244129 730

Page 29 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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












doi:10.1080/01635581.2014.847472 746






2013;31:2773–82. doi:10.1200/JCO.2013.49.1126 752






Jama 2007;298:289. doi:10.1001/jama.298.3.289 758







2014;23:575–83. doi:10.1158/1055-9965.EPI-13-1162 765



2013;65:820–6. doi:10.1080/01635581.2013.804939 768




Page 30 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



31





26. doi:10.1200/JCO.2008.19.4035 776



doi:10.1038/bjc.2012.521 779




doi:10.1207/s15327914nc5502_3 783



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



0142(19990901)86:5<826::AID-CNCR19>3.0.CO;2-0 793



doi:10.1093/jnci/djt027 796



doi:10.1007/s10549-010-1323-z 799



2011;105:1151–7. doi:10.1038/bjc.2011.374 802



doi:10.1023/A:1006056915001 805






doi:10.3109/10428191003690364 811


Page 31 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



32





2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 817



2015;8:545–51. doi:10.1158/1940-6207.CAPR-14-0442 820



doi:10.1016/j.eururo.2013.08.009 823






2015;137:2462–9. doi:10.1002/ijc.29608 829







2015;218:59–70. doi:10.1242/jeb.107110 836


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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on Decem


http://bmjopen.bm

j.com/

BM



ownloaded from



34




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 - - - -



population

Page 34 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




35




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 - - - -



population

Page 35 of 104


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




36





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


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

