is gi/gl of the diet important in diabetes prevention and ... · fiber hypothesis dr. dennis...

Is GI/GL of the Diet Important in Diabetes Prevention and

Management?

Cyril W.C. Kendall

Department of Nutritional Sciences, Faculty of Medicine, University of Toronto;

Clinical Nutrition & Risk Factor Modification Center, St. Michael’s Hospital;

Toronto, Ontario Canada.

Fiber Hypothesis

Dr. Dennis Burkitt (observational studies in Africa)

•  African diet - ↑ fiber & ↓ fat (↓ energy density) •  ↑ stool output •  ↓ chronic disease (cancer, heart disease diabetes)

Relationship of saturated fat intake to Coronary Heart Disease Before 1970

Mancini and Stamler ; Nutr Metab Cardiovasc Dis (2004) 14:52-57 (Keys A Circulation (1970) 41 (Suppl. 1): 1-211)

Diet (Macronutrient Intake) Trends

Increasing prevalence of Type 2 DM in the US with increasing

Consumption of Carbohydrates

Diabetes and Carbohydrate Consumption

300

325

350

375

400

425

450

475

500

525

0

5

10

15

20

25

30

35

60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96

Carbo

hydra

te (g/

d)

US Di

abeti

cs (p

er tho

usan

d)

Year

Total Carbohydrate Diabetics

Increasing prevalence of type 2 DM in the US with increasing

consumption of refined carbohydrates

National Nutrient Databank Centers for Disease Control and Prevention

Gross et al AJCN 2004

Relationship of saturated fat intake to Coronary Heart Disease

Siri-Tarino et al Am J Clin Nutr 2010;91:535–46.

FIGURE 2. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between saturated fat intake in relation to coronary heart disease. SAT, saturated fat intake; IV, inverse variance.

Traditional, Developing Countries

Modern, Developed Countries

Traditional Diets (Low Glycemic Index)

•  Latin America •  Mediterranean •  Middle East •  Africa •  South and East Asia

Hyperglycemia (high 2-h blood glucose or HbA1c) is associated with all-cause and CVD mortality

in non-diabetic populations

•  Helsinki Policeman Study (Pyorala et al. 1979) •  The Framingham Study (Singer et al. 1980) •  The Whitehall Study (Fuller et al. 1983) •  The Chicago Heart Study (Pan et al. 1986) •  The Rancho Bernardo Study (Park et al. 1996; Barrett-Connor et al., 1998) •  Hoorn study (De Vegt et al. Diabetologia, 1999) •  ARIC study (Vitelli et al. 1997) •  Meta-analysis DECODE study (Lancet 1999) •  EPIC-Norfolk (Khaw et al. 2001) •  The Cardiovascular Health Study (Smith et al. 2002) • ….many more now

Chiasson J-L, Josse RG, Hanefeld M, et al. JAMA 2003;346:393.

STOP-NIDDM Trial:

N= 1,386 subjects with Impaired Glucose Tolerance (high risk of developing diabetes) ~50% of subjects received acarbose (α-glucosidase

inhibitor, ↓ rate of glucose absorption). RESULT: ↓ 36% progression to diabetes

Chiasson J-L, Josse RG, Hanefeld M, et al. JAMA 2003;346:393.

CV Events

Acarbose

↓ 49% RR

STOP-NIDDM Trial: CV events

GLYCEMIC INDEX RANKING Foods grouped into high, intermediate and low classes according to GI :

High GI (GI >90)   Most Breads   Plain Crackers   Most Breakfast Cereals

  Instant Mashed Potatoes   Millet   Corn chips

Medium GI (GI 70-89)   All Bran   Oatmeal   Most biscuits or cookies   Buckwheat

  Sweet Corn   Boiled New Potatoes   Yam   Sweet Potatoes

Low GI (GI <69)   Pasta   Pumpernickel Bread   Parboiled Rice   Most legumes (dried)

  Nuts   Barley   Bulgur (cracked wheat)

High GI

Intermediate GI

Low GI

0

1

2

3

4

5

6

7

8

Time (h)

Blo

od G

luco

se (m

mol

/L)

Spaghetti

White Bread

Wholemeal Bread

0 1 2 3

<0.05

<0.005<0.001 <0.005

<0.02

Jenkins et al. Diab Care 1983;6:155-9

Blood Glucose Increments after Spaghetti, White and Wholemeal Bread (n=9)

4-hr Glucose Response Curves

4.00

4.50

5.00

5.50

6.00

6.50

7.00

7.50

0 30 60 90 120 150 180 210 240

Time (min)

Glu

cose

Con

cent

ratio

n (m

mol

/L)

50 Bread

100 Bread

50 Pasta

100 Pasta

GI & GL and Blood Glucose Response

(unpublished)

Glycemic Response of Tested Foods

0

25

50

75

100

125

150

GI (

%)

50 Bread100 Bread50 Pasta100 Pasta

a

bcControl

Blood Glucose: Area Under Curve High vs Low GI/GL Meals

(unpublished)

GI & GL and Blood Insulin Response

4-hr Insulin Response Curves

0

50

100

150

200

250

300

350

400

0 30 60 90 120 150 180 210 240

Time (min)

Insu

lin C

once

ntra

tion

(pm

ol/L

)

50 Bread

100 Bread

50 Pasta

100 Pasta

(unpublished)

Insulin IAUC (4-hr)

0

10,000

20,000

30,000

40,000

50,000

Insu

lin c

once

ntra

tions

(p

mol

/L)

50 Bread

100 Bread

50 Pasta

100 Pasta

Blood Insulin: Area Under Curve High vs Low GI/GL Meals

a

b c control

(unpublished)

Low vs High GI Diet: a Meta-Analysis % Difference in Glycated Proteins

% -35 -25 -15 -5 5 15!

Gilbertson et al. 2001 Komindr et al. 2001 Giacco et al 2000 Luscombe et al 1999 Jarvi et al 1999 Lafrance et al 1998 Frost et al 1993 Wolever et al 1992 Wolever et al 1992 Fontvieille et al 1992 Brand et al 1991 Jenkins et al 1991 Fontvieille et al 1988 Collier et al 1988 !

Mean %difference in 14 studies = -7.4% (CI -8.8 to -6.0%)!OVERALL RESULT!

Brand-Miller et al, Diabetes Care 2003;26:2261-67

Glycemic Load and Type 2 Diabetes – Cohort Studies

Livesey et al, Am J Clin Nutr (2013)

Overall Effect: RR=1.27/100g GL

P<0.001

Parameter ADA BDA CDA EASD Weight Reduction Modest weight loss Desirable weight Healthy weight Desirable weight

(5-7%) Physical Activity Increase Increase Increase Increase Carbohydrate 50-60% 50-55% 50-60% 45-60% Polysaccharides Whole grains, Fiber rich foods Whole grain cereals Low GI foods

fruits, vegetables & legumes Sugars No limit <25 g/d ≤10% added <10% calories Glycemic index Does not recommend Discusses Include low GI foods low GI foods

for general use recommended Fiber As for general >30 g/d 25-35 g/d Increase

population (low GI foods) Protein 15-20% 10-15% 11% 10-20% Total fat 25-35% 30-35% < 30% < 35% SFA/Trans <10% <10% < 10% < 10% MUFA 10-20% 10-15% 10-15% 10-20% PUFA ~10% <10% < 10% < 10% Cholesterol <300 mg/d <300 mg/d - < 300 mg

International Diabetes Associations’ Lifestyle and Dietary Guidelines

Anderson, Randles, Kendall, Jenkins. J Am Coll Nutr (2003)

Low GI Diet in Diabetes Control: Effectiveness Study (n=210)

Study Design

Low Glycemic Index diet

Month

Week

1 2 3 4

1 2 3 4 5 6 7 8 9 10 12 13 14 15 16

♠ ♠ ♠ ♠ ♠ ♠

11

π π π π π π π

High Cereal Fiber diet♠ Blood, BP & 7-day diet record

π Anthropometry

Figure 3. Schematic representation of the study protocol

Pretest +Randomization

(2 months)

17 18 19 20 21 22 23 24

5 6

π♠

Low Glycemic Index diet

Month

Week

1 2 3 4

1 2 3 4 5 6 7 8 9 10 12 13 14 15 16

♠ ♠ ♠ ♠ ♠ ♠

11

π π π π π π π

High Cereal Fiber diet♠ Blood, BP & 7-day diet record

π Anthropometry

Figure 3. Schematic representation of the study protocol

Pretest +Randomization

(2 months)

17 18 19 20 21 22 23 24

5 6

π♠

Bran flakesOatmeal

Whole wheat crackersBulgur

Vegetable soupBeans, Chickpeas, Lentils

Whole wheat breadPumpernickel bread (Dimpflmeier)

Rye crackersBarley

Shredded WheatRed River cereal

White / Brown riceParboiled rice

Potato (baked, mashed)Pasta (al dente)

High Cereal Fiber Diet (Control)

Low GI Diet (Test)

Bran flakesOatmeal

Whole wheat crackersBulgur

Vegetable soupBeans, Chickpeas, Lentils

Whole wheat breadPumpernickel bread (Dimpflmeier)

Rye crackersBarley

Shredded WheatRed River cereal

White / Brown riceParboiled rice

Potato (baked, mashed)Pasta (al dente)

High Cereal Fiber Diet (Control)

Low GI Diet (Test)

GI < 70 GI = 80

Recommended Study Foods

23

Low Glycemic Index Diets and Diabetes Control

Jenkins DJ, Kendall CW et al. JAMA 2008.

Control Low GI Energy (kcal) 1,648 1,664 Fat (%) 29.4 32.6

SFA 8.9 9.1 MUFA 11.6 12.9 PUFA 5.9 6.5 Cholesterol (mg/100kcal) 145 143

Protein (%) 21.0 21.5 Carbohydrate (%) 48.1 44.4 Fiber (g/1000 kcal) 16.5 19.7

GI 84 70

HbA1c (intent to treat, n=210)

0 4 8 12 16 20 24

Week

6.306.406.506.606.706.806.907.007.107.207.30

HbA

1c (%

)

Low Glycemic Index

High Cereal Fiber

P<0.001


Fasting Glucose

114

118

122

126

130

134

138

142

146

Fast

ing

gluc

ose

(mg/

dL)

Low Glycemic Index

High Cereal Fiber

P=0.022

0 4 8 12 16 20 24

Week Jenkins DJ, Kendall CW et al. JAMA 2008.

Body Weight (kg)

8081828384858687888990

Bod

y w

eigh

t (kg

)

Low Glycemic Index

High Cereal Fiber

P=0.053

0 4 8 12 16 20 24


HDL-C

38

39

40

41

42

43

44

45

46

HD

L-C

(mg/

dL)

Low Glycemic Index

High Cereal Fiber

P=0.005

0 4 8 12 16 20 24


LDL/HDL-C

2.1

2.2

2.3

2.4

2.5

2.6

2.7

LDL/HDL-C

Low Glycemic Index

High Cereal Fiber

P=0.047

0 4 8 12 16 20 24


Diastolic BP

68

70

72

74

76

Dia

stol

ic B

P (m

mH

g)

Low Glycemic Index

High Cereal Fiber


0 4 8 12 16 20 24

Week

P=0.37

Effect of pulses alone on Fasting Blood Glucose

Study or SubgroupAnderson et al. [45]Wursh et al. [46]Cobiac et al. [47]Fleming et al. [48]Fruhbeck et al. [49]Jang et al. [50] CADJang et al. [50] DM2Nestel et al. [51]Winham et al. [53]Winham et al. [52]Pittaway et al. [54]

Total (95% CI)Heterogeneity: Tau² = 0.71; Chi² = 79.65, df = 10 (P < 0.00001); I² = 87%Test for overall effect: Z = 2.95 (P = 0.003)

Weight9.3%8.3%

10.1%8.6%7.9%7.9%8.3%9.8%9.3%

10.2%10.3%

100.0%

IV, Random, 95% CI-0.68 [-1.37, 0.01]0.67 [-0.25, 1.59]

-0.12 [-0.57, 0.33]0.04 [-0.80, 0.88]

-2.96 [-3.98, -1.94]-3.55 [-4.57, -2.53]-1.63 [-2.55, -0.71]-0.88 [-1.43, -0.33]

0.00 [-0.69, 0.69]-0.14 [-0.57, 0.29]

-0.52 [-0.93, -0.11]

-0.82 [-1.36, -0.27]

Year19841988199019901997200120012004200720072007

Std. Mean Difference Std. Mean DifferenceIV, Random, 95% CI

-4 -2 0 2 4Favours Pulses Alone Favours Control

Sievenpiper, Kendall, et al. Diabetologia (2009)

Effect of pulses alone on Fasting Blood Insulin

Study or SubgroupWursh et al. [46]Fleming et al. [48]Fruhbeck et al. [49]Jang et al. [50] DM2Jang et al. [50] CADNestel et al. [51]Winham et al. [52]Pittaway et al. [54]Winham et al. [53]

Total (95% CI)Heterogeneity: Tau² = 0.32; Chi² = 30.02, df = 8 (P = 0.0002); I² = 73%Test for overall effect: Z = 2.14 (P = 0.03)

Weight7.7%7.9%9.4%

11.9%14.2%12.1%12.2%13.2%11.5%

100.0%

IV, Random, 95% CI-0.23 [-1.39, 0.93]0.00 [-1.14, 1.14]

-2.69 [-3.63, -1.75]-0.63 [-1.30, 0.04]

-0.61 [-1.02, -0.20]-0.14 [-0.79, 0.51]-0.26 [-0.89, 0.37]0.16 [-0.37, 0.69]

-0.24 [-0.95, 0.47]

-0.49 [-0.93, -0.04]

Year198819901997200120012004200720072007


-4 -2 0 2 4Favours Pulses Alone Favours Control


Effect of pulses in low-GI diets on Glycated Proteins

Study or Subgroup

Jenkins et al. [13]Jenkins et al. [14]Brand et al. [16]Wolever et al. [19]Fontvielle et al. [17]Wolever et al. [18]Frost et al. [20]Gilbertson et al. [21]Komindr et al. [22]Jimenez-Cruz et al. [24]Rizkalla et al. [27]Jimenez-Cruz et al. [25]Frost et al. [26]Wolever et al. [31]Ma et al. [30]

Total (95% CI)

Heterogeneity: Tau² = 0.04; Chi² = 37.66, df = 14 (P = 0.0006); I² = 63%Test for overall effect: Z = 3.93 (P < 0.0001)

Weight

3.9%5.2%4.8%

11.3%10.6%9.8%4.1%6.1%6.7%8.6%6.1%1.8%6.7%8.6%5.8%

100.0%

IV, Random, 95% CI

-0.93 [-1.52, -0.34]-0.62 [-1.09, -0.15]-0.55 [-1.06, -0.04]-0.24 [-0.38, -0.10]

0.00 [-0.18, 0.18]-0.31 [-0.53, -0.09]-0.40 [-0.97, 0.17]

-0.49 [-0.90, -0.08]-0.07 [-0.44, 0.30]

-0.56 [-0.83, -0.29]-0.69 [-1.10, -0.28]

0.00 [-0.98, 0.98]0.10 [-0.27, 0.47]0.00 [-0.27, 0.27]0.09 [-0.34, 0.52]

-0.28 [-0.42, -0.14]

Year

198719881991199219921992199420012001200320042004200420082008


-4 -2 0 2 4Favours Pulses (LGI) Favours Control


Effect of pulses in high-fibre diets on

Fasting Blood Glucose Study or Subgroup

Simpson et al. [55] DM1Simpson et al. [55] DM2Kinmonth et al. [56]Riccardi et al. [58]Karlstrom et al. [57]O'Dea et al. [59]Anderson et al. [60]Singh et al. [61]Giacco et al. [63]Chandalia et al. [62]Hall et al. [64]

Total (95% CI)

Heterogeneity: Tau² = 0.07; Chi² = 193.75, df = 10 (P < 0.00001); I² = 95%Test for overall effect: Z = 3.66 (P = 0.0003)

Weight

2.2%10.6%10.0%10.3%10.3%10.0%10.2%10.0%6.9%

10.0%9.4%

100.0%

IV, Random, 95% CI

-0.73 [-1.77, 0.31]-0.69 [-0.75, -0.63]-0.44 [-0.58, -0.30]-0.19 [-0.29, -0.09]-0.07 [-0.17, 0.03]

-0.34 [-0.48, -0.20]-0.28 [-0.40, -0.16]-0.30 [-0.44, -0.16]-0.68 [-1.07, -0.29]-0.35 [-0.49, -0.21]

0.14 [-0.06, 0.34]

-0.32 [-0.49, -0.15]

Year

19811981198219841987198919911996200020002005


-4 -2 0 2 4Favours Pulses (HF) Favours Control


Effect of pulses in high-fibre diets on Glycated Proteins

Study or Subgroup

Simpson et al. [55] DM1Simpson et al. [55] DM2Karlstrom et al. [57]O'Dea et al. [59]Anderson et al. [60]Giacco et al. [63]Chandalia et al. [62]

Total (95% CI)

Heterogeneity: Tau² = 0.05; Chi² = 159.20, df = 6 (P < 0.00001); I² = 96%Test for overall effect: Z = 2.92 (P = 0.003)

Weight

15.0%15.3%15.3%15.0%15.0%9.0%

15.5%

100.0%

IV, Random, 95% CI

-0.10 [-0.20, -0.00]-0.50 [-0.58, -0.42]-0.59 [-0.67, -0.51]

0.05 [-0.05, 0.15]-0.12 [-0.22, -0.02]-0.42 [-0.81, -0.03]-0.24 [-0.30, -0.18]

-0.27 [-0.45, -0.09]

Year

1981198119871989199120002000


-4 -2 0 2 4Favours Pulses (HF) Favours Control


Low GI Pulse Based Diet in Diabetes Mean Daily Nutrient Intakes

High Fiber Control

Pulse Diet Test

Energy, Kcal 1440 1491*

Fat (% energy) 29.3 31.0*

Protein (% energy) 21.7 23.0*

Available carbohydrate (% of energy) 49.0 46.2

Fiber (g/1000 kcal) 18.6 25.2*

Glycemic index 82 66*

Pulses (g/d) 20 196*

Jenkins, Kendall, et al. Arch Intern Med (2012)

　

　

Between Treatment Difference (P-value)

High Fiber Control

Pulse Diet Test

(n=64) (n=67) Weight (kg) -1.7* -2.6* 0.219

Glucose (mgl/dL) -7 -11* 0.934 HbA1c (%) -0.3* -0.5* 0.001

Lipids (mg/dL)

Total Cholesterol -2 -9* 0.001 LDL-C -1 -3* 0.063 HDL-C 1* 0 0.001

Triglycerides -9* -19* 0.064 Total Chol:HDL-C ratio -0.12* -0.14* 0.360

Blood Pressure (mmHg)

Systolic 0 -4* 0.010 Diastolic 0 -3* 0.001

Low GI Pulse Based Diet in Diabetes Change Across Treatments (mean) (n=131)

10y CHD Risk (%) -0.5 -1.1* 0.010

Jenkins, Kendall, et al. Arch Intern Med (2012)

Dietary Approaches to Managing Diabetes - Low Carbohydrate Diets -

Ajala O, et al. AJCN 2013.

Overall Effect HbA1c: -0.12%, P=0.04

Dietary Approaches to Managing Diabetes - Low Glycemic Index -



Dietary Approaches to Managing Diabetes - Mediterranean Diet -



Dietary Approaches to Managing Diabetes - High Protein Diets -



　 Muffins Both Nuts

Energy (kcal/d) 471 474 477

Total fat (g/d) 21 32 44

MUFA (g/d) 3 14 25

PUFA (g/d) 12 12 12

SFA (g/d) 5 5 5

Dietary Cholesterol (mg/d) 2 1 0

Total Protein (g/d) 12 12 12

Plant Protein (g/d) 7 10 12

Available Carbohydrate (g/d) 58 34 9

Fiber (g/d) 6 6 6

Glycemic Load 36 19 1

Nuts and Diabetes (n=117) Nutrients Provided by Study Supplements

.

Jenkins, Kendall, et al. Diab Care (2011)

-0.30

-0.25

-0.20

-0.15

-0.10

-0.05

0.00

Cha

nge

in %

HbA

1c

Full Dose Nuts

Half Dose Nuts and Muffins

Full Dose Muffins

P=0.0012

P=0.3855 P=0.4834

Nuts and Diabetes Change in % HbA1c Across Treatments (n=117)

Jenkins, Kendall, et al. Diab Care (2011)

Full vs Half P=0.004 Full vs Muffin P=0.001

　　

P-value Treatment Difference

Nuts Both Muffins Characteristics (n=40) (n=38) (n=39)

Weight (kg) 79.7±2.28 86.50±2.54 82.90±2.35 0.327

Glucose (mmol/L) 7.15±0.22 7.41±0.33 7.18±0.25 0.425

HbA1c (%) 6.88±0.10 7.02±0.12 7.06±0.09 0.039

Lipids (mmol/L) 　　　　　　　

Total Cholesterol 4.16±0.18 3.98±0.13 4.32±0.17 0.002

LDL-C 2.34±0.15 2.21±0.10 2.47±0.13 0.007

HDL-C 1.10±0.04 1.09±0.04 1.14±0.06 0.701

Triglycerides 1.59±0.16 1.50±0.14 1.64±0.16 0.204

Total Chol:HDL-C ratio 3.88±0.16 3.84±0.19 3.97±0.21 0.015

LDL-C:HDL-C ratio 2.17±0.13 2.14±0.13 2.29±0.14 0.028

Blood pressure (mmHg) 　　　　　　　

Systolic 120±2 124±2 124±2 0.678

Diastolic 70±1 71±1 72±2 0.182

End of Treatment Differences for Intent-to-Treat Analyses (n=117) (mean + sem)

Nuts consumption >5x/wk 18 - 51% reduction in CHD

risk.

Kris-Etherton PM, Zhao G, Binkoski AE, Coval SM, Etherton TD. Nutr. Rev. 2001

Nuts & CHD Risk: Epidemiologic Studies

45

Subjects: High risk of CVD (aged 55-80y) Study Design: Randomized Control Trial  Three Study groups:

  Control: Advice to reduce dietary fat   Test 1: Mediterranean Diet plus Olive Oil

(1litre per week)   Test 2: Mediterranean Diet plus Nuts (30g per day)

 Duration: average 4.8 y follow-up

PREDIMED Trial (7,447 men and women)

Estruch, Ros, Salas-Salvado et al. N Engl J Med 2013

46

Estruch, Ros, Salas-Salvado et al. N Engl J Med 2013

PREDIMED Trial (7,447 men and women)

Low GL Canola Based Diet in Diabetes Mean Daily Nutrient Intakes (n=141)

　

Whole Wheat Control Diet

Low GI Canola P-value

Energy (kcal/d) 1617 1525 0.089

Total fat (%) 27 37 <0.001

SFA (%) 8 8 0.684

MUFA (%) 10 18 <0.001

PUFA (%) 7 9 <0.001

Diet Cholest (mg/1000 kcal) 137 135 0.601

Total Protein (%) 20 20 0.370

Plant Protein (%) 8 9 <0.001

Available Carbohydrate (%) 49 38 <0.001

Fiber (g/1000 kcal) 20 19 0.261

Glycemic Index 87 73 <0.001

Glycemic Load 122 75 <0.001

　

　

Between Treatment Difference (P-value)

Whole Wheat Control

Low GI Canola

(n=69) (n=70) Weight (kg) -1.9* -2.4* 0.133

Glucose (mmol/L) -0.45* -0.43* 0.759 HbA1c (%) -0.35* -0.48* 0.001

Lipids (mmol/L)

Total Cholesterol -0.02 -0.25* <0.001 LDL-C 0.00 -0.15* <0.001 HDL-C -0.01 -0.02 0.096

Triglycerides -0.05 -0.16* 0.034 Total Chol:HDL-C ratio 0.01 -0.15* 0.002

Blood Pressure (mmHg)

Systolic -5* -5* 0.982 Diastolic -3* -4* 0.579

Low GL Canola Based Diet in Diabetes Change Across Treatments (mean) (n=141)

Summary •  There is increasing scientific evidence from

clinical studies that low GI/GL diets improve glycemic control in diabetes.

•  Low GI/GL diets also appear to have other metabolic advantages including improvements in blood lipids, blood pressure and possibly body weight for individuals with diabetes.

•  Part of the effect appears to be slowing carbohydrate absorption.

Summary

•  Low GI/GL foods are part of traditional diets.

•  We should be promoting traditional diets and low GI foods as part of a healthy dietary pattern.

is gi/gl of the diet important in diabetes prevention and ... · fiber hypothesis dr. dennis...

Documents