Mitos da nutrição

Pedro Carrera Bastos

MITOS EM NUTRIÇÃO

RECOMENDAÇÕES NUTRICIONAIS

PIRÂMIDE AMERICANA

USDA, 1992

NOVA RODA PORTUGUESA

Cereais e derivados, tubérculos

4 a 11 doses

Hortaliças 3 a 5 doses Fruta 3 a 5 doses

Lácteos 2 a 3 doses Carnes, peixe e ovos 1,5 a 4,5 doses

Leguminosas 1 a 2 doses Gorduras e óleos 1 a 3 doses

Instituto do Consumidor, 2003

HARVARD SCHOOL OF PUBLIC HEALTH, 2005

http://www.hsph.harvard.edu/nutritionsource/what-should-you-eat/pyramid/

PIRÂMIDE VEGETARIANA

PIRÂMIDE MEDITERRÂNICA

PIRÂMIDE LATINO-AMERICANA

VÁRIAS DIETAS

LENDAS

Battle Creek J.H. Kellogg

W. Arbuthnot-Lane M. Bircher-Benner D. Burkitt

S. Graham

DINHEIRO

COMO CHEGAR A CONCLUSÕES

Epidemiologia

Cordain L. Dietary implications for the development of acne: a shifting paradigm. In: U.S. Dermatology Review II 2006, (Ed.,Bedlow, J).

Touch Briefings Publications, London, 2006.

CAUSA/EFEITO

Estudos com Animais

Estudos In vitro

Estudos com Humanos: Estudos Metabólicos de curto prazo

RCTs

Epidemiologia

Cordain L. Dietary implications for the development of acne: a shifting paradigm. In: U.S. Dermatology Review II 2006, (Ed.,Bedlow, J).

Touch Briefings Publications, London, 2006.

CAUSA/EFEITO

RECOMENDAÇÕES BASEADAS EM LENDAS E EPIDEMIOLOGIA

CONTROVÉRSIAS

COLESTEROL DIETÉTICO < 300 MG/DIA

Dietary Guidelines for Americans, USDA, 2005

ALIMENTO (100 G) COLESTEROL (MG)

Queijo de Azeitão 88

Nata 33% gordura 97

Costoleta de Porco (gorda) grelhada 111

Peito de vitela estufado 121

Camarão Cozido 198

Mexilhão cozido 360

Fígado de vitela grelhado 387

Ovo cozido 408

Tabela de Composição dos Alimentos. Centro de Segurança Alimentar e Nutrição. Instituto nacional de Saúde Dr. Ricardo Jorge, 2006

30

EFEITOS NO CURTO PRAZO

INCREMENTO DE 100 MG/D DE COLESTEROL DIETÉTICO AUMENTOU: ü  CT: 2.2 mg/dl ü  C- HDL: 0.3 mg/dl

Okuyama H, et al. World Rev Nutr Diet. 2007;96:1-17.

34

US NATIONAL HEALTH AND NUTRITIONAL SURVEY (1984–1994)

Okuyama H, et al. World Rev Nutr Diet. 2007;96:1-17.

CHOLESTEROL SYNTHESIS, TRANSPORT, & EXCRETION / 225

C

CE

C

LDL(apo B-100, E)

receptor

EXTRAHEPATICTISSUES

Synthesis

LDL

CE

TG

TG

CE

LRP receptor

IDL(VLDL remnant)

Chylomicronremnant

CC

C

CE

SynthesisBile acids

(total pool, 3–5 g)

BILE DUCT

VLDL Chylomicron

Bileacids

CE

CCE

Diet (0.4 g/d)HEPATIC PORTAL VEIN

GALLBLADDER

ENTEROHEPATIC CIRCULATION

C(0.6 g/d)

Bile acids(0.4 g/d)

Feces

ILEUM

Unesterifiedcholesterol

pool

CE

HDLA-ILC

AT

CETP

LPL

ACAT

HL

CEC

TGCEC

LIVERTGCEC

TGCEC

CE

TG, CE

TG

TGCEC

C

–8999 %

CEC

––

C

LDL(apo B-100, E)

receptor

Figure 26–6. Transport of cholesterol between the tissues in humans. (C, unesterified cholesterol; CE, cho-lesteryl ester; TG, triacylglycerol; VLDL, very low density lipoprotein; IDL, intermediate-density lipoprotein; LDL,low-density lipoprotein; HDL, high-density lipoprotein; ACAT, acyl-CoA:cholesterol acyltransferase; LCAT,lecithin:cholesterol acyltransferase; A-I, apolipoprotein A-I; CETP, cholesteryl ester transfer protein; LPL, lipopro-tein lipase; HL, hepatic lipase; LRP, LDL receptor-related protein.)

CHOLESTEROL IS EXCRETED FROM THEBODY IN THE BILE AS CHOLESTEROL ORBILE ACIDS (SALTS)About 1 g of cholesterol is eliminated from the bodyper day. Approximately half is excreted in the feces afterconversion to bile acids. The remainder is excreted ascholesterol. Coprostanol is the principal sterol in the

feces; it is formed from cholesterol by the bacteria inthe lower intestine.

Bile Acids Are Formed From CholesterolThe primary bile acids are synthesized in the liver fromcholesterol. These are cholic acid (found in the largestamount) and chenodeoxycholic acid (Figure 26–7).

ch26.qxd 3/16/04 10:58 AM Page 225

Murray R, et al. Harper’s Illustrated Biochemistry 26th Edition. McGraw-Hill, 2003

Revisiting Dietary Cholesterol Recommendations:Does the Evidence Support a Limit of 300 mg/d?

Maria Luz Fernandez & Mariana Calle

Published online: 4 August 2010# Springer Science+Business Media, LLC 2010

Abstract The perceived association between dietary cho-lesterol (DC) and risk for coronary heart disease (CHD) hasresulted in recommendations of no more than 300 mg/d forhealthy persons in the United States. These dietaryrecommendations proposed in the 1960s had little scientificevidence other than the known association betweensaturated fat and cholesterol and animal studies wherecholesterol was fed in amounts far exceeding normalintakes. In contrast, European countries, Asian countries,and Canada do not have an upper limit for DC. Further,current epidemiologic data have clearly demonstrated thatincreasing concentrations of DC are not correlated withincreased risk for CHD. Clinical studies have shown thateven if DC may increase plasma low-density lipoprotein(LDL) cholesterol in certain individuals (hyper-responders),this is always accompanied by increases in high-densitylipoprotein (HDL) cholesterol, so the LDL/HDL cholesterolratio is maintained. More importantly, DC reduces circu-lating levels of small, dense LDL particles, a well-definedrisk factor for CHD. This article presents recent evidencefrom human studies documenting the lack of effect of DCon CHD risk, suggesting that guidelines for DC should berevisited.

Keywords Dietary cholesterol . LDL cholesterol .

HDL cholesterol . LDL size . Clinical studies .

Epidemiologic data . Eggs

Introduction

The American Heart Association (AHA) recommends nomore than 300 mg/d of dietary cholesterol (DC) for healthypersons to prevent increased risk for coronary heart disease(CHD) [1]. These recommendations are mostly based onthe presence of both saturated fat and cholesterol in manyfoods and on data derived from animal studies wheresupraphysiologic doses of cholesterol, ranging from theequivalent of 1,000 mg to 20,000 mg/d for humans, werefed in order to produce atherosclerosis [2].

It is important to note that many other countries do nothave the same guidelines for DC. Canada [3••], Korea [4•],New Zealand [5], and India [6], for example, do not set anupper limit for DC, focusing instead on controlling theintake of saturated fat and trans fat, which are the majordeterminants of blood cholesterol concentrations. Similarly,the European guidelines on cardiovascular disease preven-tion have the following recommendations regarding healthyfood choices: “consume a wide variety of foods, adjustenergy intake to maintain a healthy weight, encourageconsumption of fruits and vegetables, replace saturated fatwith mono or polyunsaturated fatty acids and reduce saltintake” [7]. In contrast to US policies, Europeans have nodietary guidelines for DC [7]. A summary of the dietaryrecommendations for DC in different countries, includingtwo recent reports from the AHA, is presented in Table 1.

Epidemiologic studies do not support an associationbetween cholesterol intake and CHD [8–12]. This couldpartly be explained by the fluctuations in response to dietarycholesterol among all individuals, which varies from nochanges to large increases in plasma cholesterol. In addition,it is critical to note that for those individuals who havehypercholesterolemic response to dietary cholesterol (aboutone third of the population), the rise is typically due to

M. L. Fernandez (*) :M. CalleDepartment of Nutritional Sciences,the University of Connecticut,3624 Horsebarn Road Extension,Storrs, CT 06269, USAe-mail: [email protected]

Curr Atheroscler Rep (2010) 12:377–383DOI 10.1007/s11883-010-0130-7

an increase in CHD in patients diagnosed with diabetes atthe highest level of egg consumption [21]. Other studiesalso suggest that individuals with diabetes should limit eggintake [22, 23•]. In summary, the preponderance of theepidemiologic evidence from the past 14 years does notsupport a relationship between dietary cholesterol (or eggintake) and risk for CHD [4•, 8–14], [15•, 16••, 17••].Accordingly, neither Europe [7], Canada [3••], nor Asiancountries [4•, 5] restrict dietary cholesterol as part of therecommendations for a heart-healthy diet.

Eggs and Dietary Cholesterol

The AHA still recommends limiting other food items highin cholesterol if eggs are to be consumed [1] in spite ofrecent reports that show no association between egg intakeand risk for heart disease [8–11, 18••, 19–21]. In fact, thereare no studies with substantial evidence supporting theclaims of egg consumption involved in CHD risk. Incontrast, a recent analysis in which a risk-apportionmentapproach was used on the risk factors for CHD revealedthat egg intake contributes to less than 1% of the risk, andthe authors conclude that AHA dietary guidelines possiblyshould be revised [18••]. Eggs are the only food that is bothrich in cholesterol and low in saturated fat, perhapsexplaining why eggs are often used to evaluate the effectsof dietary cholesterol on plasma lipids and CHD risk [8–11,24, 25]. Other cholesterol-containing foods, such as dairyproducts, also contain high concentrations of saturated fat,which is a confounder for dietary cholesterol effects. Thismight be the reason why controversial results existregarding the effects of dairy products on CHD risk [26].

Clinical trials conducted in children [27], younger adults[24, 25], and the elderly [28, 29] have clearly demonstratedthat although dietary cholesterol provided by eggs signifi-cantly increases LDL-C in one third of the population, thoseindividuals considered hyper-responders to a cholesterolchallenge exhibit increases in both LDL-C and HDL-C, with

the result of no changes in the LDL-C/HDL-C ratio, a majorpredictor of CHD [30•]. These results indicate thatindividuals with initial plasma cholesterol concentrationsthat place them at a low risk for CHD do not develop anatherogenic lipoprotein profile following the consumptionof additional dietary cholesterol, regardless of theirresponse classification.

It is well established that nutritional interventions aimedat managing plasma lipids are known to be less effective inobese and overweight individuals [31]. During a weightloss intervention, intake of 3 eggs per day for 12 weeks wasshown to selectively increase plasma HDL-C withoutincreasing LDL-C in overweight men [32••]. Harman etal. [33•] also reported no changes in LDL-C afterconsuming 2 eggs per day for 12 weeks in a weight lossintervention study. Intake of only 1 egg per day increasedHDL-C without increasing LDL-C in men and women aged40–60 years, resulting in a lower LDL-C/HDL-C ratio [34].Similarly, in a study in which 56 participants with a meanage of 35 years were given an additional egg per day during12 weeks, significant increases were reported for HDL-Cwith no changes in LDL-C [35•]. A summary of plasmaLDL-C and HDL-C concentrations as a response to eggintake in recent clinical studies is presented in Table 2.

To evaluate whether insulin resistance, with or withoutobesity, influences the response to dietary cholesterol, Knoppet al. [36] recruited 197 healthy individuals into a random-ized crossover study in which 0, 2, and 4 eggs per day werefed for 4 weeks with a 1-month washout period in between.The participants were classified as insulin sensitive (n=65),insulin resistant (n=75), and obese insulin resistant (OIR,n=58). Insulin-resistant and insulin-sensitive individuals hadsignificant increases in LDL-C of 7.8% and 3.3%, respec-tively, after consuming 4 eggs per day, whereas OIRindividuals had no changes in LDL-C at any intake level.In contrast, HDL-C was significantly increased for all groupseven after the consumption of only 2 eggs per day. Thesestudies suggest that dietary management of OIR individualsneed not include restrictions on eggs.

Table 2 Changes in LDL-C, HDL-C, LDL size, and HDL size as a response to dietary cholesterol provided by eggs in various populations

Population Duration Additional dietary cholesterol LDL-C HDL-C LDL-C/HDL-C ratio LDL size HDL size

Children (n=54) [27] 4 wk 518 mg/d Increase Increase No change Increase ND

Women (n=51) [25] 4 wk 640 mg/d Increase Increase No change Increase ND

Men (n=28) [32••] 12 wk 640 mg/d No change Increase Decrease Increase Increase

Men/women (n=42) [34] 12 wk 215 mg/d No change Increase No change Increase Increase

Men/women (n=34) [28] 4 wk 640 mg/d Increase Increase No change Increase Increase

Men/women (n=56) [35•] 12 wk 250 mg/d No change Increase Decrease ND ND

Men/women (n=45) [33•] 12 wk 400 mg/d No change No change No change ND ND

HDL-C high-density lipoprotein cholesterol; LDL-C low-density lipoprotein cholesterol; ND not determined.

Curr Atheroscler Rep (2010) 12:377–383 379

LDL PEQUENAS E DENSAS

ü  sdLDL: ü  Tamanho da partícula <

25.5nm ü Densidade > 1.04g/ml

ü Aumenta significativamente o risco de DCV

ü  TG elevados e HDL-C baixo é um bom preditor de sdLDL

Griffin BA. Proc Nutr Soc 1999;58:163-69

41

FLUXO DE LDL PARA A INTIMA É > PARA LDL PEQUENAS E DENSAS UMA VEZ NA INTIMA, LDL PEQUENAS E DENSAS SÃO SUSCEPTÍVEIS DE SOFRER OXIDAÇÃO

Cordain, 2009

•  1.

Células espumosas

214 Esrey et al.

TABLE 3. Cox regression estimates of the effect of dietary intake measured at baseline on coronary heart disease death among 3925 men and women 30 to 59 years of age after 12.4 years foIIow-up in the North American Lipid Research Clinics Prevalence FoIIow-Up Study

Model 3 Model 1 Modei 2 (adjusted for age, sex,

(adjusted for age, sex, (adjusted for age, sex, energy intake, serum lipids, energy intake) energy intake, serum lipids”) and other risk factorsb)

Variable Estimate RR (95% CI) Estimate RR (95% CI) Estimate RR (95% CI)

% Fat Total 0.044" 1.04 (1.01-1.08) 0.043" 1.04 (1.01-1.08) 0.043' 1.04 (l.Ol-1.08) Saturated 0.104” 1.11 (1.04-1.18) 0.095" 1.10 (1.03-1.17) 0.103" 1.11 (1.04-1.18) Monounsaturated 0.095" 1.10 (1.03-1.17) 0.094' l 1.10 (1.03-1.18) 0.080' 1.08 (1.01-1.16) Polyunsaturated -0.040 0.96 (0.88-1.05) -0.027 0.97 (0.89-1.07) -0.014 0.99 (0.90-1.08)

% Carbohydrate -0.034" 0.97 (0.94-0.99) -0.040" 0.96 (0.93-0.99) -0.036' 0.96 (0.94-0.99) % Protein 0.020 1.02 (0.97-1.08) 0.013 1.01 (0.96-1.07) 0.011 1.01 (0.95-1.07) % Alcohol + alcohol’ -0.019 0.98 (0.92-1.05) 0.001 1.00 (0.94-1.06) 0.0001 1.00 (0.95-1.07) Cholesterol (10 mg) per 5000 kJ 0.004 1.00 (0.99-1.02) 0.006 1.01 (0.99-1.02) 0.004 1.00 (0.99-1.02)

?Serum lipids (mmoliliter) included total serum cholesterol and high-density lipoproteins. bOther risk factors were systolic blood pressure (mmHg), cigarette smoking status (current smoker/nonsmoker), body mass index (kg/m*), and glucose

intolerance (present/absent). ‘p < 0.05. “p < 0.01. RR = Relative risk for coronary heart disease mortality, defined as exp(@, associated with a I+unit increase in the dietary variable; 95% CI = 95% confidence

interval, defined as exp@ 2 1.96 SE)(z, - q), where SE is the standard error of /3, and 7, and Q represent the two levels of the dietary variable being compared; % =percentage of total energy provided by the dietary variable.

(total cholesterol and HDL) were added to the regression model (model 2), the estimated relative risks for the dietary variables remained stable. The magnitudes of these relationships were also unaf- fected by the addition of other known coronary risk factors (model 3).

Among the older age group, no relationships between dietary fat or its fatty acid components, carbohydrate, or protein, and coronary heart disease death were identified (Table 4). The estimares of the regression coefficients were al1 smaller than those among the younger age group

and none approached statistical significance. The magnitude and direction of some estimates changed between the three models evaluated, but within a small range. When a quadratic alcohol term was included in the Cox model, percentage of energy intake as alcohol was significantly associated with risk of coronary disease death within this older age group.

None of the dietary components examined were significantly associated with total mortality within either the young or the older age groups.

TABLE 4. Cox regression estimates of the effect of dietary intake measured at baseline on coronary heart disease death among 621 men and women 60 to 79 years of age after 12.4 years follow-up in the North American Lipid Research Clinics Prevalence Follow-Up Study

Variable

Model 1 (adjusted for age, sex,

energy intake) Estimate RR (95% CI)

Model 3 Model 2 (adjusted for age, sex,

(adjusted for age, sex, energy intake, serum lipids, energy intake, serum lipids”) and other risk factorsb) Estimate RR (95% CI) Estimate RR (95% CI)

% Fat Total -0.0004 1.00 (0.96-1.04) -0.002 1.00 (0.96-1.04) -0.011 0.99 (0.95-1.03) Saturated -0.031 0.97 (0.89-1.05) -0.026 0.97 (0.90-1.06) -0.038 0.96 (0.88-1.05) Monounsaturated 0.025 1.03 (0.95-1.11) 0.019 1.02 (0.94-l. 11) -0.005 1.00 (0.91-1.08) Polyunsaturated 0.006 1.01 (0.91-1.11) -0.004 1.00 (0.90-l. 10) -0.004 1.00 (0.90-1.10)

% Carbohydrate 0.004 1.00 (0.97-1.04) 0.003 1.00 (0.97-1.04) 0.016 1.02 (0.98-1.05) % Protein 0.017 1.02 (0.94-l. 10) 0.015 1.02 (0.94-1.10) 0.0002 1.00 (0.93-1.08) % Alcohol + alcohol’ -0.114" 0.89 (0.84-0.94) -0.116** 0.89 (0.83-0.95) -0.124” 0.88 (0.83-0.95) Cholesterol (10 mg) per 5000 kJ 0.011 1.01 (0.99-1.03) 0.011 1.01 (0.99-1.03) 0.006 1.01 (0.98-1.03)

“Serum lipids (mmoliliter) included total serum cholesterol and high-density lipoproteins. bOther risk factors were systolic blood pressure (mmHg), cigarette smoking status (current smoker/nonsmoker), body mass index (kg/m’), and glucose

intolerance foresent/absent).

RR = Relative risk for coronary heart disease mortality, defined as exp(@), associated with a l-unit increase in the dietary variable; 95% Cl = 95% confidence interval, defined as exp(p + 1.96 SE)(zt - Q), where SE is the standard error of @, and zt and q represent the two levels of the dietary variable being compared; % =percent of total energy provided by the dtetary variable.

214 Esrey et al.

TABLE 3. Cox regression estimates of the effect of dietary intake measured at baseline on coronary heart disease death among 3925 men and women 30 to 59 years of age after 12.4 years foIIow-up in the North American Lipid Research Clinics Prevalence FoIIow-Up Study

Model 3 Model 1 Modei 2 (adjusted for age, sex,

(adjusted for age, sex, (adjusted for age, sex, energy intake, serum lipids, energy intake) energy intake, serum lipids”) and other risk factorsb)

Variable Estimate RR (95% CI) Estimate RR (95% CI) Estimate RR (95% CI)

% Fat Total 0.044" 1.04 (1.01-1.08) 0.043" 1.04 (1.01-1.08) 0.043' 1.04 (l.Ol-1.08) Saturated 0.104” 1.11 (1.04-1.18) 0.095" 1.10 (1.03-1.17) 0.103" 1.11 (1.04-1.18) Monounsaturated 0.095" 1.10 (1.03-1.17) 0.094' l 1.10 (1.03-1.18) 0.080' 1.08 (1.01-1.16) Polyunsaturated -0.040 0.96 (0.88-1.05) -0.027 0.97 (0.89-1.07) -0.014 0.99 (0.90-1.08)

% Carbohydrate -0.034" 0.97 (0.94-0.99) -0.040" 0.96 (0.93-0.99) -0.036' 0.96 (0.94-0.99) % Protein 0.020 1.02 (0.97-1.08) 0.013 1.01 (0.96-1.07) 0.011 1.01 (0.95-1.07) % Alcohol + alcohol’ -0.019 0.98 (0.92-1.05) 0.001 1.00 (0.94-1.06) 0.0001 1.00 (0.95-1.07) Cholesterol (10 mg) per 5000 kJ 0.004 1.00 (0.99-1.02) 0.006 1.01 (0.99-1.02) 0.004 1.00 (0.99-1.02)

?Serum lipids (mmoliliter) included total serum cholesterol and high-density lipoproteins. bOther risk factors were systolic blood pressure (mmHg), cigarette smoking status (current smoker/nonsmoker), body mass index (kg/m*), and glucose

intolerance (present/absent). ‘p < 0.05. “p < 0.01. RR = Relative risk for coronary heart disease mortality, defined as exp(@, associated with a I+unit increase in the dietary variable; 95% CI = 95% confidence

interval, defined as exp@ 2 1.96 SE)(z, - q), where SE is the standard error of /3, and 7, and Q represent the two levels of the dietary variable being compared; % =percentage of total energy provided by the dietary variable.

(total cholesterol and HDL) were added to the regression model (model 2), the estimated relative risks for the dietary variables remained stable. The magnitudes of these relationships were also unaf- fected by the addition of other known coronary risk factors (model 3).

Among the older age group, no relationships between dietary fat or its fatty acid components, carbohydrate, or protein, and coronary heart disease death were identified (Table 4). The estimares of the regression coefficients were al1 smaller than those among the younger age group

and none approached statistical significance. The magnitude and direction of some estimates changed between the three models evaluated, but within a small range. When a quadratic alcohol term was included in the Cox model, percentage of energy intake as alcohol was significantly associated with risk of coronary disease death within this older age group.

None of the dietary components examined were significantly associated with total mortality within either the young or the older age groups.

TABLE 4. Cox regression estimates of the effect of dietary intake measured at baseline on coronary heart disease death among 621 men and women 60 to 79 years of age after 12.4 years follow-up in the North American Lipid Research Clinics Prevalence Follow-Up Study

Variable

Model 1 (adjusted for age, sex,

energy intake) Estimate RR (95% CI)

Model 3 Model 2 (adjusted for age, sex,

(adjusted for age, sex, energy intake, serum lipids, energy intake, serum lipids”) and other risk factorsb) Estimate RR (95% CI) Estimate RR (95% CI)

% Fat Total -0.0004 1.00 (0.96-1.04) -0.002 1.00 (0.96-1.04) -0.011 0.99 (0.95-1.03) Saturated -0.031 0.97 (0.89-1.05) -0.026 0.97 (0.90-1.06) -0.038 0.96 (0.88-1.05) Monounsaturated 0.025 1.03 (0.95-1.11) 0.019 1.02 (0.94-l. 11) -0.005 1.00 (0.91-1.08) Polyunsaturated 0.006 1.01 (0.91-1.11) -0.004 1.00 (0.90-l. 10) -0.004 1.00 (0.90-1.10)

% Carbohydrate 0.004 1.00 (0.97-1.04) 0.003 1.00 (0.97-1.04) 0.016 1.02 (0.98-1.05) % Protein 0.017 1.02 (0.94-l. 10) 0.015 1.02 (0.94-1.10) 0.0002 1.00 (0.93-1.08) % Alcohol + alcohol’ -0.114" 0.89 (0.84-0.94) -0.116** 0.89 (0.83-0.95) -0.124” 0.88 (0.83-0.95) Cholesterol (10 mg) per 5000 kJ 0.011 1.01 (0.99-1.03) 0.011 1.01 (0.99-1.03) 0.006 1.01 (0.98-1.03)

“Serum lipids (mmoliliter) included total serum cholesterol and high-density lipoproteins. bOther risk factors were systolic blood pressure (mmHg), cigarette smoking status (current smoker/nonsmoker), body mass index (kg/m’), and glucose

intolerance foresent/absent).

RR = Relative risk for coronary heart disease mortality, defined as exp(@), associated with a l-unit increase in the dietary variable; 95% Cl = 95% confidence interval, defined as exp(p + 1.96 SE)(zt - Q), where SE is the standard error of @, and zt and q represent the two levels of the dietary variable being compared; % =percent of total energy provided by the dtetary variable.

37,851 H seguidos por 8 anos

80.082 M seguidas por 14 anos

Baseado em Hu Fb, et al. JAMA. 1999;281:1387-1394 e adaptado por Okuyama H, et al. World Rev Nutr Diet. 2007;96:1-17.

P!0·45), analyses excluding those who were takingcholesterol-lowering drugs or those with total cholesterol con-centration $ 2200mg/l (trend P!0·72), analyses using quan-titative variables (cholesterol and systolic blood pressure) inthose who participated in a health check-up programme(P!0·09), all yielded the basically similar results: there wasno significant association between egg consumption andCHD incidence.

Total cholesterol concentration and coronary heart diseaseincidence – age and sex-adjusted and multivariate-adjustedCox analyses

The number of cases, their male percentage, person-years, ageand sex-adjusted as well as multivariate-adjusted hazard ratiosand their 95% CI for CHD incidence according to the totalcholesterol concentration categories are shown in Table 4.The percentage of men progressively decreased in thegroups with the higher total cholesterol concentration. Totalcholesterol concentration was significantly related to CHDincidence (hazard ratio of CHD in those with total cholesterolconcentration $2400mg/l was 2·17 (95% CI 1·22, 3·85) as

compared with those with total cholesterol concentration,1800mg/l; trend P!0·0018).

Discussion

In the present study, we found that eating eggs more fre-quently, up to almost daily, was not associated with any con-sistent adverse effect on CHD incidence. We confirmed apositive association between total cholesterol concentrationand CHD incidence in Japanese as in previous studies(Kodama et al. 1990; Kitamura et al. 1994; Okamura et al.2003). We also found an inverse correlation between egg con-sumption and mean total cholesterol concentration as well asthe frequency of the subjects with hypercholesterolaemia.The subjects with hypercholesterolaemia were more frequentamong the groups of subjects who ate fewer eggs than thosein the groups of subjects who ate more eggs. Controversiesexist as to the relationship between dietary egg consumptionand total cholesterol concentration. Some studies haveshown no relationship between egg consumption and totalcholesterol concentration (Mattson et al. 1972; Nichols et al.1976a,b; Kummerow et al. 1977; Porter et al. 1977; Franket al. 1978; Flynn et al. 1979; Dawber et al. 1982; Keys,

Table 3. Coronary heart disease incidence according to egg consumption categories in the Japan Public Health Center-based prospectivestudy (Hazard ratios (HR) and 95% CI)

,1 d/week 1–2 d/week 3–4 d/week Almost daily

Egg consumption. . . HR 95% CI HR 95% CI HR 95% CI HR 95% CI Trend P

Subjects at risk* (n) 10 491 20 802 31 182 28 260Person-years 96 748 213907 323856 292858CHD incidenceCases† (n) 64 110 147 141Men (%) 72 84 76 81Incidence (per 1000 person-years) 0·66 0·51 0·45 0·48Age and sex-adjusted HR 1·28 0·95, 1·72 1·11 0·86, 1·42 1·01 0·80, 1·27 1 – 0·11Multivariate-adjusted HR‡ 1·19 0·86, 1·64 1·00 0·77, 1·30 1·00 0·79, 1·26 1 – 0·45

* Total number of subjects was 90 735.†Subtotal of subjects was 462.‡Multivariate Cox analysis adjusted for age, sex, BMI, hypertension, diabetes, use of cholesterol-lowering drugs, smoking (never, ex-, and current smoker), alcohol

drinking (six categories), whether or not intended to avoid cholesterol-rich diets, consumption frequencies of meat, fish, vegetables, fruits, and cohort effects.

Table 4. Coronary heart disease incidence according to serum total cholesterol concentration categories in men and women (Japan Public HealthCenter-based prospective study) (Hazard ratios (HR) and 95% CI)

Total cholesterol,1800 1800–1999 2000–2199 2200–2399 $ 2400

concentration (mg/l). . . HR 95% CI HR 95% CI HR 95% CI HR 95% CI HR 95% CI Trend P

Subjects at risk* (n) 9162 7528 6896 4959 4484Men (%) 45·0 38·0 31·6 28·1 23·3 ,0·0001Person-years 96 027 77 847 70 822 50 438 45 423CHD incidence

Cases† (n) 28 20 33 22 23Men (%) 79 75 70 73 70Incidence (per 1000 person-years) 0·29 0·26 0·47 0·44 0·51Age and sex-adjusted HR 1 – 0·97 0·54, 1·71 1·93 1·16, 3·20 1·93 1·10, 3·38 2·48 1·42, 4·33 0·0001Multivariate-adjusted HR‡ 1 – 0·94 0·52, 1·68 1·85 1·11, 3·10 1·68 0·95, 3·00 2·17 1·22, 3·85 0·0018

*Total number of subjects was 33 029.†Subtotal of subjects was 126.‡Multivariate Cox analysis adjusted for age, sex, BMI, systolic blood pressure, diabetes, use of cholesterol-lowering drugs or anti-hypertensive drugs, smoking (never, ex-, and

current smoker), alcohol drinking (six categories), whether or not intended to avoid cholesterol-rich diets, consumption frequencies of egg, meat, fish, vegetables, fruits, andcohort effects.

Egg consumption and coronary heart disease 925

men and women to assess the association between eggconsumption and cardiovascular health. We found thatself-reported egg consumption frequency was not asso-ciated with increased mortality from CHD or stroke inmen or women after adjustment for established risk fac-tors. In addition, we did not observe increased CHD or

stroke mortality among diabetics with ‘high’ egg con-sumption (v. ‘low’) in subgroup analyses. A uniquefinding was the reduction in stroke mortality among menreporting ‘high’ egg consumption, but the 95 % CI waswide, indicating imprecision, and the results should beinterpreted in the present context.

Table 2 Egg consumption and mortality from CHD in NHANES III mortality follow-up

Egg consumption

,1 egg EO/week $1 to ,7 egg EO/week $7 egg EO/week

MenDeaths from CHD (person-years) 53 (17 597) 106 (34 060) 39 (7877)Egg EO per week (range) 0?27 (0–0?7) 1?93 (1–6?5) 7?54 (7–49?7)

HR HR 95 % CI HR 95 % CI

Model 1 1?00 1?44 0?89, 2?33 1?60 0?94, 2?72Model 2 1?00 1?34 0?82, 2?18 1?25 0?70, 2?22Model 3 1?00 1?26 0?79, 2?00 1?13 0?61, 2?11Model 4 1?00 1?38 0?85, 2?24 1?38 0?84, 2?26

WomenDeaths from CHD (person-years) 72 (28 626) 74 (35 871) 22 (5770)Egg EO per week (range) 0?24 (0–0?7) 1?79 (1–6?3) 7?41 (7–35?5)



NHANES III, Third National Health and Nutrition Examination Survey 1988–1994; EO, eating occasion; HR, hazard ratio; WHR, waist-to-hip ratio.Model 1: Age and energy.Model 2: Age, energy, marital status, educational status, race/ethnicity, smoking status, BMI, WHR, hypertension, diabetes.Model 3: Age, energy, marital status, educational status, race/ethnicity, smoking status, BMI, WHR, diabetes, hypertension and dietary variables.Model 4: Men – age, energy, marital status, race/ethnicity, BMI, diabetes, hypertension and alcohol intake.Model 4: Women – age, energy, marital status, educational status, race/ethnicity, WHR, diabetes, hypertension and vitamin E.

Table 3 Egg consumption and mortality from stroke in NHANES III mortality follow-up

Egg Consumption

,1 egg EO/week $1 to ,7 egg EO/week $7 egg EO/week

MenDeaths from stroke (person-years) 21 (17 597) 32 (34 060) 10 (7877)Egg EO per week (range) 0?27 (0–0?7) 1?93 (1–6?5) 7?54 (7–49?7)



WomenDeaths from stroke (person-years) 26 (28 626) 39 (35 871) 9 (5770)Egg EO per week (range) 0?24 (0–0?7) 1?79 (1–6?3) 7?41 (7–35?5)



NHANES III, Third National Health and Nutrition Examination Survey 1988–1994; EO, eating occasion; HR, hazard ratio; WHR, waist-to-hip ratio.Model 1: Age and energy.Model 2: Age, energy, marital status, educational status, race/ethnicity, smoking status, BMI, WHR, hypertension, diabetes.Model 3: Age, energy, marital status, educational status, race/ethnicity, smoking status, BMI, WHR, diabetes, hypertension and dietary variables.Model 4: Men – age, energy, race/ethnicity, alcohol and fibre intake.Model 4: Women – age, energy, BMI and grain intake.

266 CG Scrafford et al.

NUTRITION MYTHS

REDUZIR GORDURA DIETÉTICA DIMINUI RISCO DE DCV

“DIMINUIR GORDURA REDUZ O RISCO CARDIOVASCULAR”

970 G.J. NELSON ETAL.

(13-20). Thus, it is difficult to determine if simply reducing the fat content of the diet without changing the dietary FAC changes the tissue FAC or alters plasma lipid levels. Further- more, people eat natural and processed food as complex mix- tures that vary day to day and year to year. They rarely eat pure fats or oils, and certainly not continuously for weeks. In most previous studies (13-20) that have examined this issue, the FAC of the diet was changed markedly as well as the amount of calories from fat; thus, the experimental diets had different ratios of saturated and unsaturated FA in them. This confounds the interpretation of the results; one does not know if the changes observed are the result of changes in the caloric intake or changes in the type of FA in the diet. In an effort to clarify this point, we fed LF (22 en% fat) and HF (39 en% fat) diets to healthy volunteers for 50 d. The diets had FAC that were identical despite the intentional differences in the fat content in the diet.

MATERIALS A N D M E T H O D S

Materials. All natural foods were purchased at local food markets. FA standards were purchased from Nu-Chek-Prep (Elysian, MN). Organic solvents were obtained from Baxter Scientific, Burdick and Jackson Brand (McGaw Park, IL). The reagents for the enzymatic determination of plasma cholesterol and triglycerides were purchased from Sigma Diag- nostics (St. Louis, MO).

Subjects. The volunteers were recruited from the West Coast, and consisted of men between the ages of 20 and 35. The physical characteristics of the eleven male volunteers (HNS-27) who completed the study were age (years), 32.9 _+ 4.5; weight (kg), 72.9 _+ 8.2; body mass index (kg/M2), 23.1 _+ 1.6; blood pressure systolic, 115.9 _+ 9.5; blood pressure diastolic, 73.2 _+ 7.8; smokers, none. Initially twelve volunteers were included in the study, but one was unable to complete the protocol. The volunteers were given complete physical examinations. Body weights had to be within -10 to +20 percent of ideal body weights using the Metropolitan Life Insur- ance Company tables (medium frame values from the 1983 edition). Evidence of existing illness or chronic disease was an exclu~on criterion. Mild hypertension was not an exclusion criterion, but the group that was recruited tended to have blood pressures slightly below average for men in this age group. Smoking, excessive alcohol consumption, and evidence of narcotic abuse were also exclusionary.

Experimental design. All the volunteers were confined to the Nutrition Research Suite of the Western Human Nutrition Research Center (San Francisco, CA) for the duration of the study. As the subjects stay within the confines of the Nutri- tion Suite except for occasional supervised outings, they had no opportunity to consume any food except that provided by the Center. Thus, compliance with the protocol was, of neces- sity, 100%. In addition, all food intake was monitored, portions were weighed, and subjects were required to consume everything provided to them during their meals. (A rubber spatula was provided to ensure that all food was scraped from

the plates and eaten.) Food spills were carefully monitored and recorded, and fluid intake, while ad libitum, was also measured precisely.

The protocol for this study was approved by the Institu- tional Review Boards of the University of California at Davis (Davis, CA) and the USDA (Washington, D.C.). A crossover design was used so that the subjects acted as their own controls. The subjects were fed a stabilization diet, containing 39% of calories from fat for 20 d. The HF diet had a macronutrient composition of 39 en% fat, 16 en% protein, and 45 en% CHO. On day 20 of the study, they were divided randomly into two groups: Group A remained on the HF diet (39% of calories from fat) for 50 d; Group B was placed on a LF diet (22% of calories from fat). The LF diet had a macronutrient composition of 22 en% fat, 16 en% protein, and 62 en% CHO. After 50 d, day 70 of the study, the groups switched diets for the remaining 50 d of the study. After 120 d, all the volunteers were discharged from the Nutrition Suite and re- turned to an ad libitum diet.

During the confinement period, there were blood draws on study days 2, 20, 45, 70, 95, and 120. As measured parameters on the intermediate blood draws (study days 45 and 95) showed insignificant differences from the endpoints (study days 70 and 120), only data obtained from the endpoint blood draws were used for statistical analysis. Statistical comparisons of the measured parameters were made using day 20 as the baseline values for each subject.

Diets. The diets consisted of natural foods. No dietary supplements were given. The macronutrient composition of the diets is given in Table 1. (A complete description of the diets, listing all the major and minor nutrients, is available upon request.) A seven-day menu cycle was used throughout the study. Proximate analysis was made on seven individual diet composite samples taken from each menu once during the study for both the stabilization diet and intervention diets. The results for the seven composite samples were averaged to find the actual composition of the diets. No alcohol was in-

TABLE 1 Composition of HNS*27 Diets, Proximal Analysis a

Low-fat diets High-fat diets (% of total calories) Target Target

Measured value Measured value

Macronutrient energy distribution

Protein 15.9 16.0 15.7 16.0 Fat 22.2 20.0 38.7 40.0 Carbohydrate 61.9 64.0 45.7 44.0

Cholesterol content (mean, mg/day) - - 360 - - 360

Fatty acid energy distribution

Saturated 6.4 5 10.6 10 Monounsaturated 9.2 10 15.5 20 Polyunsaturated 6.6 5 12.6 10

P/S ratio 1.0 1.0 1.2 1.0 ap/s, polyunsaturated/saturated; HNS-2 7 diet.

Lipids, Vol. 30, no. 11 (1995)

Nelson GJ, Schmidt PC, Kelley DS. Lipids. 1995 Nov;30(11):969-76

Nelson GJ, Schmidt PC, Kelley DS. Lipids. 1995 Nov;30(11):969-76

EFFECT OF DIETARY FAT ON BLOOD LIPIDS 973

TABLE 5 Red Blood Cell Fatty Acid Composition (wt%) a

High-fat diet Low-fat diet FAME (means _+ SD) (means _+ SD) Pvalues

14:0 (myristate) 0.20 _+ 0.05 0.19 • 0.03 16:0 D M A 1.51 • 1 .62+0 .11 16:0 (palmitate) 16.57 • 1.57 16.16 + 0.38 16:1n-9 0 . 3 5 • 0.33_+0.02 17:0 2 . 8 7 • 3 . 0 4 • 18:0 D M A 0.35 • 0.05 0.38 _+ 0.04 18:1 n-7 D M A 1.64 • 0.28 1.75 _+ 0.22 18:0 (stearate) 10.25 + 0.58 9.74 + 0.36 0.01 18:1n-7t 1.11 • 1.06_+0.1 18:1n-9 (oleate) 9.51 _+0.82 9.11 _+0.8 18:1 n-7 0.85 _+ 0.06 0.88 • 0.08 18:1n-5 0.44 _+ 0.05 0.47 _+ 0.07 18:2t tand 19:0 0.11 _+0.02 0.11 _+0.01 18:2n-6 (l inoleate) 9.99 _+ 0.61 8.37 _+ 0.46 0.000004 20:0 (arachidate) 0.61 _+ 0.11 0.54 _+ 0.05 18:3n-3 (@-Iinolenate) 0.30 • 0.05 0.25 • 0.04 0.0005 20:3n-6 (homo-T-linolenate) 1.12 _+ 0.23 1.34 • 0.24 0.00003 22:0 (behenate) 2.22 • 0.47 2 _+ 0.19 20:4n-6 (arachidonate) 12.11 + 2.34 13.85 -+ 0.47 0.04 23:0 0.47 _+ 0.1 0.42 _+ 0.04 20:5n-3 (EPA) 0.52 _+ 0.14 0.61 • 0.11 24:0 (l ignocerate) 6.54 -+ 1.56 5.94 -+ 0.59 24:4n-6 3.12 • 0.73 3.49 _+ 0.34 24:1 n-9 (nervonate) 6.38 _+ 1.32 6.03 _+ 0.5 22:4n-3 0.56 • 0.15 0.74 _+ 0.13 0.007 24:2n-6 1.05 _+ 0.16 1.05 • 0.09 22:5n-3 2.6 _+ 0.69 2.99 • 0.3 22:6n-3 (DHA) 4.02 _+ 1.3 4.87 _+ 0.81 26:0 0.44 _+ 0.11 0.41 _+ 0.05

Total 99.04 _+ 0.16 98.95 _+ 0.13 Unknown 0.96 _+ 0.16 1.05 _+ 0.13

aSee Tables 2 and 3 for abbreviations.

ous lipoprotein fractions were not altered significantly, there was a redistribution of the lipoprotein spectrum during the two diet periods. The increase in the plasma triglyceride level indicated an increase in the plasma very low density lipoprotein (VLDL) level and a decrease in both the plasma HDL and LDL levels. Thus, the total cholesterol level remained con- stant because the increase in plasma VLDL cholesterol com- pensated for the reduction in HDL-cholesterol and LDL-cholesterol. Because of individual variations, the standard devia-

tion was large enough to prevent these values from reaching statistical significance.

D I S C U S S I O N

It has almost been dogma for the last forty years in the field of dietary fat effects on blood lipids that HF diets will raise blood cholesterol levels while LF diets will lower blood cholesterol (1-4). A problem one encounters when examining the scientific basis for this concept is that the FAC of LF and HF diets was rarely the same. The Keys et al. (5) and Hegsted et al. (6) formulas for the effect of dietary fat on blood cholesterol levels were published thirty years ago. Those equations say nothing about the total percentage of fat calories in the diet, but they imply that HF diets will raise blood cholesterol because they relate grams of fat ingested to blood cholesterol level, positively for saturated FA, except stearic (7,8), and negatively for polyunsaturated fat (n-3 FA are ignored). Di- etary cholesterol has only a small influence on blood cholesterol levels. If one uses the Keys equation (5) or the Hegsted equation (6) to calculate the change in blood cholesterol that one would expect with the protocol used in this study, there should have been an average difference between the HF and LF diets o f -2 3 mg/dl. The subjects' LF diet cholesterol level was 173 mg, and should have risen to 196 mg/dl if the Keys et al. (5) and Hegsted et al. (6) equations were an appropriate explanation for the physiological response to changes in dietary fat intakers. The observed total cholesterol value when the participants consumed the HF was 177 mg/dl, not significantly different from the value obtained when they consumed the LF diet.

There are several possible reasons that the Keys et al. (5) and Hegsted et aL (6) equations do not hold in this particular experiment: (i) The equations were developed using large population data bases (11,20). Here the sample number is too small to detect the effect. This answer is unlikely and unsatis- factory because the conditions of this experiment were carefully controlled and the statistical power of the protocol was excellent.

(ii) The average cholesterol level in the subjects was con- siderably below that in the European and American populations used to develop the Keys et aL (5) and Hegsted et al. (6)

TABLE 6 High- and Low-Fat Diets, Blood Cholesterol, Triglycerides, and Lipoprotein Values

Total HDL- LDL- cholesterol Triglycerides cholesterol cholesterol

Period Diet mean + SD mean _+ SD mean • SD mean • SD

Entry A d l ib i tum 176.3 + 33.1 85.8 + 28.4 46.3 _+ 14.0 112.8 • 26.8 Stabil ization High-fat 172.5 + 30.3 75.3 • 46.4 44.8 • 11.6 112.6 • 21.9 Intervention Low-fat 173.2 _+ 27.3 91.5 • 38.0 40.5 • 12.4 114.5 • 21.3 Intervention High-fat 176.9 _+ 32.9 66.4 • 41.7 43.2 + 13.4 119.5 _+ 24.3 Paired t-test, Pvalues a 0.425 0.002 0.258 0.238

~lhe t-test compares only the values at the end of the high- or low-fat diets with the values obtained at the end of the stabilization period, study day 20. Groups A and B values were taken at study day 70 and day 120, depending on the leg of the intervention diet for the each group; HDL, high density lipoprotein; LDL, low density lipoprotein.

Lipids, Vol. 30, no. 11 (1995)

ü Nenhum caso de morte por DAC

ü Nenhum caso de enfarte agudo do miocardio

(População = 2,600)

Bang HO, Dyerberg J. Adv Nutr Res 1980 3:1-22.

ENTRE 1968-1978, NA GRONELÂNDIA

ALIMENTOS VEGETAIS / ANIMAIS EM 13 SOCIEDADES PRIMITIVAS!

Cordain L et al. Eur J Clin Nutr 2002; 56 (suppl 1):S42-S52. !

American Institute for Cancer Research 11th Annual

Research Conference on Diet, Nutrition and Cancer

The Mediterranean Diets: What Is So Special about the Diet of Greece?

The Scientific Evidence

1

Artemis P. Simopoulos2

The Center for Genetics, Nutrition and Health, Washington, DC

ABSTRACT The term “Mediterranean diet,” implying that all Mediterranean people have the same diet, is amisnomer. The countries around the Mediterranean basin have different diets, religions and cultures. Their dietsdiffer in the amount of total fat, olive oil, type of meat and wine intake; milk vs. cheese; fruits and vegetables; andthe rates of coronary heart disease and cancer, with the lower death rates and longer life expectancy occurring inGreece. Extensive studies on the traditional diet of Greece (the diet before 1960) indicate that the dietary patternof Greeks consists of a high intake of fruits, vegetables (particularly wild plants), nuts and cereals mostly in the formof sourdough bread rather than pasta; more olive oil and olives; less milk but more cheese; more fish; less meat;and moderate amounts of wine, more so than other Mediterranean countries. Analyses of the dietary pattern of thediet of Crete shows a number of protective substances, such as selenium, glutathione, a balanced ratio of(n-6):(n-3) essential fatty acids (EFA), high amounts of fiber, antioxidants (especially resveratrol from wine andpolyphenols from olive oil), vitamins E and C, some of which have been shown to be associated with lower risk ofcancer, including cancer of the breast. These findings should serve as a strong incentive for the initiation ofintervention trials that will test the effect of specific dietary patterns in the prevention and management of patientswith cancer. J. Nutr. 131: 3065S–3073S, 2001.

KEY WORDS: c diet of Crete c (n-3) fatty acids c wild plants c antioxidants c cancer c (n-6) fatty acids

The health of the individual and the population in generalis the result of interactions between genetics and a number ofenvironmental factors. Nutrition is an environmental factor ofmajor importance (1–4). Our genetic profile has not changedover the past 10,000 y, whereas major changes have takenplace in our food supply and in energy expenditure and phys-ical activity (5–17). Today industrialized societies are charac-terized by the following: 1) an increase in energy intake anddecrease in energy expenditure; 2) an increase in saturated fat,(n-6) fatty acids and trans fatty acids and a decrease in (n-3)fatty acid intake; 3) a decrease in complex carbohydrates andfiber intake; 4) an increase in cereal grains and a decrease infruit and vegetable intake; and 5) a decrease in protein, anti-oxidant and calcium intake (5–17). Furthermore, the ratio of(n-6) to (n-3) fatty acids is 16.74:1, whereas during evolutionit was 2–1:1 (Table 1, Fig. 1).

Recent investigations of the dietary patterns and healthstatus of the countries surrounding the Mediterranean basinclearly indicate major differences among them in both dietary

intake and health status. Therefore, the term “Mediterraneandiet” is a misnomer. There is not just one Mediterranean dietbut in fact many Mediterranean diets (18), which is notsurprising because the countries along the Mediterranean ba-sin have different religions, economic and cultural traditionsand diets. Diets are influenced by religious habits, that is,Muslims do not eat pork or drink wine and other alcoholicdrinks, whereas Greek Orthodox populations usually do noteat meat on Wednesdays and Fridays but drink wine, and soon. Although Greece and the Mediterranean countries areusually considered to be areas of medium-high death rates(14.0–18.0 per 1000 inhabitants), death rates on the island ofCrete have been below this level continuously since before1930 (19). No other area in the Mediterranean basin has hadas low a death rate as Crete, according to data compiled by theUnited Nations in their demographic yearbook for 1948. Itwas 11.3–13.7 per 1000 inhabitants before World War II and;10.6 in 1946–1948 (19). Cancer and heart disease causedalmost three times as many deaths proportionally in theUnited States as in Crete (19). The diet of Crete representsthe traditional diet of Greece before 1960. The Seven Coun-tries Study was the first to establish credible data on cardio-vascular disease prevalence rates in contrasting populations(United States, Finland, The Netherlands, Italy, former Yu-goslavia, Japan and Greece), with differences found on theorder of 5- to 10-fold in coronary heart disease (20). In 1958,the field work started in Dalmatia in the former Yugoslavia.

1 Presented as part of the 11th Annual Research Conference on Diet, Nutritionand Cancer held in Washington, DC, July 16–17, 2001. This conference wassponsored by the American Institute for Cancer Research and was supported bythe California Dried Plum Board, The Campbell Soup Company, General Mills,Lipton, Mead Johnson Nutritionals, Roche Vitamins Inc. and Vitasoy USA. Guesteditors for this symposium publication were Ritva R. Butrum and Helen A.Norman, American Institute for Cancer Research, Washington, DC.

2 To whom correspondence should be addressed.E-mail: [email protected]

0022-3166/01 $3.00 © 2001 American Society for Nutritional Sciences.

3065S

by on September 27, 2006

jn.nutrition.orgD

ownloaded from





1











3065S


jn.nutrition.orgD

ownloaded from

GORDURA DIETÉTICA: 37% DA ENERGIA TOTAL CONSUMIDA





1











3065S


jn.nutrition.orgD

ownloaded from

0% 5% 10% 15% 20% 25% DAC Total DCV RISCO RELATIVO

Howard BV, et al. JAMA. 2006 Feb 8;295(6):655-66

ORIGINAL CONTRIBUTION

Low-Fat Dietary Patternand Risk of Cardiovascular DiseaseThe Women’s Health Initiative Randomized ControlledDietary Modification TrialBarbara V. Howard, PhD; LindaVan Horn, PhD; Judith Hsia, MD;JoAnn E. Manson, MD; Marcia L.Stefanick, PhD; SylviaWassertheil-Smoller, PhD; Lewis H.Kuller, MD; Andrea Z. LaCroix, PhD;Robert D. Langer, MD; Norman L.Lasser, MD; Cora E. Lewis, MD;Marian C. Limacher, MD; Karen L.Margolis, MD; W. Jerry Mysiw, MD;Judith K. Ockene, PhD; Linda M.Parker, DSc; Michael G. Perri, PhD;Lawrence Phillips, MD; Ross L.Prentice, PhD; John Robbins, MD;Jacques E. Rossouw, MD; Gloria E.Sarto, MD; Irwin J. Schatz, MD; Linda G.Snetselaar, PhD; Victor J. Stevens, PhD;Lesley F. Tinker, PhD; MaurizioTrevisan, MD; Mara Z. Vitolins, DrPH;Garnet L. Anderson, PhD; Annlouise R.Assaf, PhD; Tamsen Bassford, MD;Shirley A. A. Beresford, PhD; Henry R.Black, MD; Robert L. Brunner, PhD;Robert G. Brzyski, MD; BetteCaan, DrPH; Rowan T. Chlebowski, MD;Margery Gass, MD; Iris Granek, MD;Philip Greenland, MD; JenniferHays, PhD; David Heber, MD;Gerardo Heiss, MD; Susan L.Hendrix, DO; F. Allan Hubbell, MD;Karen C. Johnson, MD;Jane Morley Kotchen, MD

CLINICAL TRIALS AND OBSERVA-tional studies have identifiedstrong associations betweenlow-density lipoprotein cho-

lesterol (LDL-C) level and other cardio-vascular disease (CVD) risk factors anddietary intake of fats, particularly

See also pp 629, 643, and 693.Author Affiliations are listed at the end of thisarticle.Corresponding Author: Barbara V. Howard, PhD,

MedStar Research Institute, 6495 New Hampshire Ave,Suite 201, Hyattsville, MD 20783 ([email protected]).

Context Multiple epidemiologic studies and some trials have linked diet with car-diovascular disease (CVD) prevention, but long-term intervention data are needed.Objective To test the hypothesis that a dietary intervention, intended to be low infat and high in vegetables, fruits, and grains to reduce cancer, would reduce CVD risk.Design, Setting, and Participants Randomized controlled trial of 48 835 post-menopausal women aged 50 to 79 years, of diverse backgrounds and ethnicities, whoparticipated in the Women’s Health Initiative Dietary Modification Trial. Women wererandomly assigned to an intervention (19 541 [40%]) or comparison group (29 294[60%]) in a free-living setting. Study enrollment occurred between 1993 and 1998 in40 US clinical centers; mean follow-up in this analysis was 8.1 years.Intervention Intensive behavior modification in group and individual sessions de-signed to reduce total fat intake to 20% of calories and increase intakes of vegetables/fruits to 5 servings/d and grains to at least 6 servings/d. The comparison group receiveddiet-related education materials.Main Outcome Measures Fatal and nonfatal coronary heart disease (CHD), fataland nonfatal stroke, and CVD (composite of CHD and stroke).Results By year 6, mean fat intake decreased by 8.2% of energy intake in the inter-vention vs the comparison group, with small decreases in saturated (2.9%), monoun-saturated (3.3%), and polyunsaturated (1.5%) fat; increases occurred in intakes of veg-etables/fruits (1.1 servings/d) and grains (0.5 serving/d). Low-density lipoprotein cholesterollevels, diastolic blood pressure, and factor VIIc levels were significantly reduced by 3.55mg/dL, 0.31 mm Hg, and 4.29%, respectively; levels of high-density lipoprotein cho-lesterol, triglycerides, glucose, and insulin did not significantly differ in the interventionvs comparison groups. The numbers who developed CHD, stroke, and CVD (annual-ized incidence rates) were 1000 (0.63%), 434 (0.28%), and 1357 (0.86%) in the in-tervention and 1549 (0.65%), 642 (0.27%), and 2088 (0.88%) in the comparison group.The diet had no significant effects on incidence of CHD (hazard ratio [HR], 0.97; 95%confidence interval [CI], 0.90-1.06), stroke (HR, 1.02; 95% CI, 0.90-1.15), or CVD (HR,0.98; 95% CI, 0.92-1.05). Excluding participants with baseline CVD (3.4%), the HRs(95% CIs) for CHD and stroke were 0.94 (0.86-1.02) and 1.02 (0.90-1.17), respec-tively. Trends toward greater reductions in CHD risk were observed in those with lowerintakes of saturated fat or trans fat or higher intakes of vegetables/fruits.Conclusions Over a mean of 8.1 years, a dietary intervention that reduced total fatintake and increased intakes of vegetables, fruits, and grains did not significantly re-duce the risk of CHD, stroke, or CVD in postmenopausal women and achieved onlymodest effects on CVD risk factors, suggesting that more focused diet and lifestyleinterventions may be needed to improve risk factors and reduce CVD risk.Clinical Trials Registration ClinicalTrials.gov Identifier NCT00000611JAMA. 2006;295:655-666 www.jama.com

©2006 American Medical Association. All rights reserved. (Reprinted) JAMA, February 8, 2006—Vol 295, No. 6 655

by PEDROBASTOS, on January 18, 2007 www.jama.comDownloaded from



















0% 5% 10% 15% 20% 25% DAC Total DCV RISCO RELATIVO

Howard BV, et al. JAMA. 2006 Feb 8;295(6):655-66

0%

5%

10%

15%

20%

25%

DAC Total DCV

RISCO RELATIVO

EM MULHERES QUE TINHAM DCV

Burr ML, Fehily AM, Gilbert JF, et al. Lancet 1989; 2:757-761.

DIMINUIÇÃO DA GORDURA TOTAL DE 35% PARA 32.3%

AUMENTO DO RÁCIO PUFA/SAFA EM 100%

ÁCIDOS GORDOS SATURADOS

USDA, AHA: < 10% DO TOTAL CALÓRICO

Dietary Guidelines for Americans, USDA, 2010

Mensink RP, Zock PL, Kester AD, Katan MB. Am J Clin Nutr. 2003 May;77(5):1146-55

67

SAFA EM ALGUNS ALIMENTOS

Alimento (100 g)

AG Saturados (g)

AG Laúrico

12:0

AG Mirístico

14:0

AG Palmitico

16:0

AG Esteárico

18:0

Manteiga 51 3 7 2 10

Óleo de Coco

86,5 44,6 16,8 8,2 2,8

Cacau 8,07 0 0,02 3,69 4,25

Óleo de palma 49,3 0,1 1 43,496 4,3

Leite gordo 1,865 0,077 0,297 0,829 0,365

Fonte: USDA food database


Coronary heart disease (CHD) incidence and mor-tality re! ect complex interactions between genetic susceptibilities and environmental factors. Although several CHD susceptibility genes have been identi" ed [1], several lines of evidence indicate that environ-ment rather than genetics is the main driver of CHD risk [2]. Globally, age-adjusted CHD incidence and mortality vary as much as 10-fold across populations [3,4]. CHD incidence and risk factors are sensitive to lifestyle changes. When immigrants from traditionally low-risk regions adopt the habits of high-risk popula-tions, their CHD incidence rises to approach that of

resident inhabitants, especially with increasing dura-tion of residence [5–7].

For instance, CHD is historically far more common in the United States than in Japan [4]. Among men of Japa-nese ancestry, CHD risk is lowest in Japan, intermediate in Hawaii, and highest in California [8,9]. These differences appear to re! ect the replacement of traditional Japanese cultural traditions with Western habits [8]. Indeed, Japa-nese Americans who maintained traditional customs and habits had a CHD risk similar to that of their counterparts residing in Japan, whereas those who adopted Western cul-ture had a three- to " vefold excess in CHD prevalence [8].

Dietary Fat Quality and Coronary Heart Disease Prevention: A Unifi ed Theory Based on Evolutionary, Historical, Global, and Modern PerspectivesChristopher E. Ramsden, MDKeturah R. Faurot, PA, MPHPedro Carrera-Bastos, BALoren Cordain, PhDMichel De Lorgeril, MD, PhDLaurence S. Sperling, MDCorresponding authorChristopher E. Ramsden, MDDepartment of Physical Medicine and Rehabilitation, Program on Integrative Medicine, University of North Carolina–Chapel Hill School of Medicine, CB# 7200, Chapel Hill, NC 27599, USA.E-mail: [email protected]

Current Treatment Options in Cardiovascular Medicine 2009, 11:289–301Current Medicine Group LLC ISSN 1092-8464Copyright © 2009 by Current Medicine Group LLC

Opinion statementA large and growing body of evidence indicates that dietary fatty acids regulate crucial metabolic processes involved in the pathogenesis of coronary heart disease (CHD). Despite this evidence, optimal dietary fatty acid intakes for CHD preven-tion remain unclear. Signifi cant gaps in the modern nutrition literature and contra-dictions in its interpretation have precluded broad consensus. These shortcomings can be addressed through the incorporation of evolutionary, historical, and global perspectives. The objective of this review is to propose a unifi ed theory of optimal dietary fatty acid intake for CHD prevention that integrates critical insights from evolutionary, historical, global, and modern perspectives. This broad approach may be more likely than previous methods to characterize optimal fatty acid intakes.

Introduction

Coronary heart disease (CHD) incidence and mor-tality re! ect complex interactions between genetic susceptibilities and environmental factors. Although several CHD susceptibility genes have been identi" ed [1], several lines of evidence indicate that environ-ment rather than genetics is the main driver of CHD risk [2]. Globally, age-adjusted CHD incidence and mortality vary as much as 10-fold across populations [3,4]. CHD incidence and risk factors are sensitive to lifestyle changes. When immigrants from traditionally low-risk regions adopt the habits of high-risk popula-tions, their CHD incidence rises to approach that of

resident inhabitants, especially with increasing dura-tion of residence [5–7].

For instance, CHD is historically far more common in the United States than in Japan [4]. Among men of Japa-nese ancestry, CHD risk is lowest in Japan, intermediate in Hawaii, and highest in California [8,9]. These differences appear to re! ect the replacement of traditional Japanese cultural traditions with Western habits [8]. Indeed, Japa-nese Americans who maintained traditional customs and habits had a CHD risk similar to that of their counterparts residing in Japan, whereas those who adopted Western cul-ture had a three- to " vefold excess in CHD prevalence [8].

Dietary Fat Quality and Coronary Heart Disease Prevention: A Unifi ed Theory Based on Evolutionary, Historical, Global, and Modern PerspectivesChristopher E. Ramsden, MDKeturah R. Faurot, PA, MPHPedro Carrera-Bastos, BALoren Cordain, PhDMichel De Lorgeril, MD, PhDLaurence S. Sperling, MDCorresponding authorChristopher E. Ramsden, MDDepartment of Physical Medicine and Rehabilitation, Program on Integrative Medicine, University of North Carolina–Chapel Hill School of Medicine, CB# 7200, Chapel Hill, NC 27599, USA.E-mail: [email protected]

Current Treatment Options in Cardiovascular Medicine 2009, 11:289–301Current Medicine Group LLC ISSN 1092-8464Copyright © 2009 by Current Medicine Group LLC

Opinion statementA large and growing body of evidence indicates that dietary fatty acids regulate crucial metabolic processes involved in the pathogenesis of coronary heart disease (CHD). Despite this evidence, optimal dietary fatty acid intakes for CHD preven-tion remain unclear. Signifi cant gaps in the modern nutrition literature and contra-dictions in its interpretation have precluded broad consensus. These shortcomings can be addressed through the incorporation of evolutionary, historical, and global perspectives. The objective of this review is to propose a unifi ed theory of optimal dietary fatty acid intake for CHD prevention that integrates critical insights from evolutionary, historical, global, and modern perspectives. This broad approach may be more likely than previous methods to characterize optimal fatty acid intakes.

Introduction

CHINA RURAL: < 5%

TOKELAU: 40%

EUA: 11-12% CRS: 4-18%

MAASAI: 30-35% KITAVA: 17%

A meta-analysis of prospective epidemiologic studies showed that there is no significant evidence for concluding that dietary saturated fat is associated with an increased risk of CHD or CVD.

NUTRITION MYTHS

ÓLEOS VEGETAIS RICOS EM ÓMEGA-6 DIMINUEM O RISCO DE DCV

LA levels predicted lower risk of stroke, particularlyischemic stroke.55 LA intakes are not associated with riskfor cancer.26 Therefore, observational studies generallysuggest an overall modest benefit of omega-6 PUFA intakeon CHD risk and no significant effect on stroke or cancer.These studies, some of which included LA intakes of up to10% to 12% of energy, contradict the supposition thathigher omega-6 PUFA intakes increase risk for CHD.

Omega-6 PUFA Consumption and CHDEvents: Randomized Controlled Trials

Several randomized trials have evaluated the effects ofreplacing saturated fatty acids with PUFAs on CHDevents.56 – 65 Intakes of PUFA (almost entirely omega-6PUFA) ranged from 11% to 21%. In addition to theinability to double-blind these studies, many had designlimitations such as small sample size (n!54),65 the provi-sion of only "50% of meals,56 outcomes composed largelyof “soft” ECG end points,59,60 randomization of sites ratherthan individuals with open enrollment and high turnover ofsubjects,59,60 use of vegetable oils that also contained theplant omega-3 fatty acid !-linolenic acid,57,59,60 and simul-taneous recommendations to increase fish and cod liver oiluse.58 Nevertheless, a meta-analysis including 6 of thesetrials56 – 60,62– 64 indicated that replacing saturated fatty ac-ids with PUFAs lowered the risk for CHD events by24%.66 Of the remaining 4 studies, 1 reported a significant45% reduction in risk,59 whereas no significant effect wasseen in the others.60,61,65

These trials tested the effect of replacing saturated fattyacids; no randomized trial has reported the effects ofreplacing carbohydrate or protein with omega-6 PUFAs onCHD risk. Although limitations are present for each trial,the combined results of these studies and the observationaltrials provide evidence that replacing saturated fatty acidor refined carbohydrate (eg, sugars, white bread, whiterice, potatoes) with omega-6 PUFAs reduces CHD risk. Onthe basis of the intakes of omega-6 PUFAs used in therandomized trials, metabolic studies, and nonhuman pri-mate studies discussed below, reductions in CHD riskmight be expected with omega-6 PUFA intakes of 10% to21% of energy compared with lower intakes, with noclinical evidence for adverse events.

Recommended Intakes of Omega-6Fatty Acids

Dietary recommendations for omega-6 PUFAs tradition-ally focused on the prevention of essential fatty aciddeficiency but are now increasingly seeking to define“optimal” intakes to reduce risk for chronic disease,particularly CHD. The Institute of Medicine’s Food andNutrition Board, in their Dietary Reference Intake Reportfor Energy and Macronutrients,67 defines an adequateintake of LA as 17 g/d for men and 12 g/d for women (5%to 6% of energy) 19 to 50 years of age, approximately thecurrent median US intake. Both the Dietary ReferenceIntake Report and the 2005 Dietary Guidelines for Amer-icans68 support an acceptable macronutrient distributionrange (the range of intakes for a particular energy source

that is associated with reduced risk of chronic diseasewhile providing adequate intakes of essential nutrients) of5% to 10% dietary energy from omega-6 PUFAs. TheThird Adult Treatment Panel of the National CholesterolEducation Program recommends PUFA consumption up to10%, noting that “there are no large populations that haveconsumed large quantities of polyunsaturated fatty acidsfor long periods. Thus, high intakes have not been provensafe in large populations; this introduces a note of cautionfor recommending high intakes.”69 On the other hand,evidence from trials in nonhuman primates has demon-strated cardiovascular benefits and no evidence of harmwith LA intakes of 25% of energy for up to 5 years,70,71

and randomized trials in humans have shown reduced CHDrisk with omega-6 PUFA intakes of 11% to 21% of energyfor up to 11 years with no evidence of harm.

Other governmental health recommendations foromega-6 fatty acid intakes (on a percent energy basis) areas follows: European Commission, 4% to 8%72; Food andAgriculture Organization/World Health Organization, 5%to 8%73; British Nutrition Foundation, 6% to 6.5% (max-imum, 10%)74; the Department of Health and Ageing,Australia and New Zealand, 4% to 5% (maximum, 10%)75;and the American Dietetic Association/Dietitians of Can-ada, 3% to 10%.76 The American Heart Association placesprimary emphasis on healthy eating patterns rather than onspecific nutrient targets.

Advice to reduce omega-6 PUFA intakes is typicallyframed as a call to lower the ratio of dietary omega-6 toomega-3 PUFAs.1– 4 Although increasing omega-3 PUFAtissue levels does reduce the risk for CHD,77,78 it does notfollow that decreasing omega-6 levels will do the same.Indeed, the evidence considered here suggests that it wouldhave the opposite effect. Higher omega-6 PUFA intakescan inhibit the conversion of !-linolenic acid to eicosa-pentaenoic acid,79 but such conversion is already quitelow,80 and whether additional small changes would havenet effects on CHD risk after the other benefits of LAconsumption are taken into account is not clear. The focuson ratios, rather than on levels of intake of each type ofPUFA, has many conceptual and biological limitations.81

ConclusionsThis advisory was undertaken to summarize the currentevidence on the consumption of omega-6 PUFAs, partic-ularly LA, and CHD risk. Aggregate data from randomizedtrials, case-control and cohort studies, and long-termanimal feeding experiments indicate that the consumptionof at least 5% to 10% of energy from omega-6 PUFAsreduces the risk of CHD relative to lower intakes. The dataalso suggest that higher intakes appear to be safe and maybe even more beneficial (as part of a low–saturated-fat,low-cholesterol diet). In summary, the AHA supports anomega-6 PUFA intake of at least 5% to 10% of energy inthe context of other AHA lifestyle and dietary recommen-dations. To reduce omega-6 PUFA intakes from theircurrent levels would be more likely to increase than todecrease risk for CHD.

904 Circulation February 17, 2009

by on July 26, 2010 circ.ahajournals.orgDownloaded from

Omega-6 Fatty Acids and Risk for Cardiovascular DiseaseA Science Advisory From the American Heart Association Nutrition

Subcommittee of the Council on Nutrition, Physical Activity, andMetabolism; Council on Cardiovascular Nursing; and Council on

Epidemiology and Prevention

William S. Harris, PhD, FAHA, Chair; Dariush Mozaffarian, MD, DrPH, FAHA;Eric Rimm, ScD, FAHA; Penny Kris-Etherton, PhD, FAHA; Lawrence L. Rudel, PhD, FAHA;

Lawrence J. Appel, MD, MPH, FAHA; Marguerite M. Engler, PhD, FAHA;Mary B. Engler, PhD, FAHA; Frank Sacks, MD, FAHA

Alarge body of literature suggests that higher intakes ofomega-6 (or n-6) polyunsaturated fatty acids (PUFAs)

reduce risk for coronary heart disease (CHD). However, forthe reasons outlined below, some individuals and groups haverecommended substantial reductions in omega-6 PUFA in-take.1–4 The purpose of this advisory is to review evidence onthe relationship between omega-6 PUFAs and the risk ofCHD and cardiovascular disease.

Omega-6 PUFAsOmega-6 PUFAs are characterized by the presence of at least2 carbon-carbon double bonds, with the first bond at the sixthcarbon from the methyl terminus. Linoleic acid (LA), an18-carbon fatty acid with 2 double bonds (18:2 omega-6), isthe primary dietary omega-6 PUFA. LA cannot be synthe-sized by humans, and although firm minimum requirementshave not been established for healthy adults, estimates de-rived from studies in infants and hospitalized patients receiv-ing total parenteral nutrition suggest that an LA intake of!0.5% to 2% of energy is likely to suffice. After consump-tion, LA can be desaturated and elongated to form otheromega-6 PUFAs such as !-linolenic and dihomo-!-linolenicacids. The latter is converted to the metabolically importantomega-6 PUFA arachidonic acid (AA; 20:4 omega-6), thesubstrate for a wide array of reactive oxygenated metabolites.Because LA accounts for 85% to 90% of the dietary omega-6PUFA, this advisory focuses primarily on this fatty acid,recognizing that dietary AA, which can affect tissue AA

levels,5 may have physiological sequelae.6–8 LA comesprimarily from vegetable oils (eg, corn, sunflower, safflower,soy). The average US intake of LA, according to NationalHealth and Nutrition Examination Survey 2001 to 2002 datafor adults "19 years of age, is 14.8 g/d.9 On the basis of anaverage intake of 2000 kcal/d, LA intake is 6.7% of energy.AA (!0.15 g/d) is consumed preformed in meat, eggs, andsome fish.

Omega-6 PUFAs and InflammationArguments for reduced LA intakes are based on theassumption that because CHD has an inflammatory com-ponent10 and because the omega-6 fatty acid, AA, is thesubstrate for the synthesis of a variety of proinflammatorymolecules, reducing LA intakes should reduce tissue AAcontent, which should reduce the inflammatory potentialand therefore lower the risk for CHD. The evidence,derived primarily from human studies, regarding this lineof reasoning is examined below.

AA is the substrate for the production of a wide variety ofeicosanoids (20-carbon AA metabolites). Some are proin-flammatory, vasoconstrictive, and/or proaggregatory, such asprostaglandin E2, thromboxane A2, and leukotriene B4. How-ever, others are antiinflammatory/antiaggregatory, such asprostacyclin, lipoxin A4,11 and epoxyeicosatrienoic acids.12

Epoxyeicosatrienoic acids are fatty acid epoxides producedfrom AA by a cytochrome P450 epoxygenase. Epoxyeicosa-trienoic acids also have important vasodilator properties via

The American Heart Association makes every effort to avoid any actual or potential conflicts of interest that may arise as a result of an outsiderelationship or a personal, professional, or business interest of a member of the writing panel. Specifically, all members of the writing group arerequired to complete and submit a Disclosure Questionnaire showing all such relationships that might be perceived as real or potential conflictsof interest.

This statement was approved by the American Heart Association Science Advisory and Coordinating Committee on November 6, 2008. A copy of thestatement is available at http://www.americanheart.org/presenter.jhtml?identifier"3003999 by selecting either the “topic list” link or the “chronologicallist” link (No. LS-1966). To purchase additional reprints, call 843-216-2533 or e-mail [email protected].

Expert peer review of AHA Scientific Statements is conducted at the AHA National Center. For more on AHA statements and guidelines development,visit http://www.americanheart.org/presenter.jhtml?identifier"3023366.

Permissions: Multiple copies, modification, alteration, enhancement, and/or distribution of this document are not permitted without the expresspermission of the American Heart Association. Instructions for obtaining permission are located at http://www.americanheart.org/presenter.jhtml?identifier"4431. A link to the “Permission Request Form” appears on the right side of the page.

(Circulation. 2009;119:902-907.)© 2009 American Heart Association, Inc.

Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.108.191627

902

AHA Science Advisory


Simopoulos AP, Cleland LG (eds): Omega–6/Omega–3 Essential Fatty Acid Ratio: The Scientific Evidence. World Rev Nutr Diet. Basel, Karger, 2003, vol 92, pp 81–91

Omega–6/Omega–3 Fatty Acid Ratio:The Israeli Paradox

Gal Dubnov, Elliot M. Berry

Department of Human Nutrition and Metabolism, Hebrew University, Hadassah Medical School, Jerusalem, Israel

While the amount of fat is very important in terms of public health indealing with the current epidemic of obesity, an equally significant issue is thetype of fat consumed. As polyunsaturated fatty acids (PUFA) have long beenshown to possess cholesterol-lowering effects [1], increasing their consumptionhas been promoted in the management of coronary artery disease (CAD) [2].These recommendations followed both experimental and population basedstudies that showed decreasing rates of CAD in countries with increasingpolyunsaturated/saturated fat (P/S) ratios over the past years.

The dietary habits in Israel appear to be as recommended: low in totalcalories, in total fat and in saturated fat, while high in hypolipidemic omega–6fatty acids (!6) as compared with other western countries [3, 4]. Unexpectedly, therates of modern-world illnesses are about the same as they are in the USA andEurope [3, 5, 6]. The reason for this is not clear. Recent evidence suggests thata high intake of omega–6 fatty acids may prove harmful [2, 7–9]: these fattyacids may elevate the risk of hyperinsulinemia and its associated metabolic dis-orders, atherogenesis, and cancer. Another group of PUFA, the omega–3 fattyacids (!3), have demonstrated cardioprotection in observational [10–15] andintervention studies for both secondary [16–18] and primary [18] prevention.An example for this is shown in figure 1: an Indo-Mediterranean diet, rich inthe plant-derived omega–3 fatty acid alpha-linolenic acid, markedly decreasedthe risk for a cardiac event among both those with established coronary arterydisease, or those only with risk factors [18]. A recent meta-analysis showed thatboth dietary and non-dietary sources are equally beneficial [19], and the healthbenefits of plant- derived or fish- derived omega–3 fatty acids now seem tohave a sound basis [20]. As the omega–6 and omega–3 fatty acids compete for

ü  ÓLEO MAIS CONSUMIDO: ÓLEO DE SOJA

ü  RÁCIO P/S = 0,9 – 1,2

ü  ÁCIDO LINOLEICO: 10% DA ENERGIA TOTAL DIÁRIA

ü  % LA NO ADIPÓCITO: 25%

RÁCIO N-6/N-3 > 20/1

PREVALÊNCIA DE DAC É COMPARÁVEL À DE OUTROS PAÍSES OCIDENTAIS

non-fatal myocardial infarction (MI) + CHD death.

n-6 specific PUFA trials non significantly increased the risk of non-fatal MI + CHD death by 13%

(risk ratio (RR) 1·13; 95% CI 0·84, 1·53; P=0·427)

Omega-6 Fatty Acids and Risk for Cardiovascular DiseaseA Science Advisory From the American Heart Association Nutrition

Subcommittee of the Council on Nutrition, Physical Activity, andMetabolism; Council on Cardiovascular Nursing; and Council on

Epidemiology and Prevention

William S. Harris, PhD, FAHA, Chair; Dariush Mozaffarian, MD, DrPH, FAHA;Eric Rimm, ScD, FAHA; Penny Kris-Etherton, PhD, FAHA; Lawrence L. Rudel, PhD, FAHA;

Lawrence J. Appel, MD, MPH, FAHA; Marguerite M. Engler, PhD, FAHA;Mary B. Engler, PhD, FAHA; Frank Sacks, MD, FAHA

Alarge body of literature suggests that higher intakes ofomega-6 (or n-6) polyunsaturated fatty acids (PUFAs)

reduce risk for coronary heart disease (CHD). However, forthe reasons outlined below, some individuals and groups haverecommended substantial reductions in omega-6 PUFA in-take.1–4 The purpose of this advisory is to review evidence onthe relationship between omega-6 PUFAs and the risk ofCHD and cardiovascular disease.

Omega-6 PUFAsOmega-6 PUFAs are characterized by the presence of at least2 carbon-carbon double bonds, with the first bond at the sixthcarbon from the methyl terminus. Linoleic acid (LA), an18-carbon fatty acid with 2 double bonds (18:2 omega-6), isthe primary dietary omega-6 PUFA. LA cannot be synthe-sized by humans, and although firm minimum requirementshave not been established for healthy adults, estimates de-rived from studies in infants and hospitalized patients receiv-ing total parenteral nutrition suggest that an LA intake of!0.5% to 2% of energy is likely to suffice. After consump-tion, LA can be desaturated and elongated to form otheromega-6 PUFAs such as !-linolenic and dihomo-!-linolenicacids. The latter is converted to the metabolically importantomega-6 PUFA arachidonic acid (AA; 20:4 omega-6), thesubstrate for a wide array of reactive oxygenated metabolites.Because LA accounts for 85% to 90% of the dietary omega-6PUFA, this advisory focuses primarily on this fatty acid,recognizing that dietary AA, which can affect tissue AA

levels,5 may have physiological sequelae.6–8 LA comesprimarily from vegetable oils (eg, corn, sunflower, safflower,soy). The average US intake of LA, according to NationalHealth and Nutrition Examination Survey 2001 to 2002 datafor adults "19 years of age, is 14.8 g/d.9 On the basis of anaverage intake of 2000 kcal/d, LA intake is 6.7% of energy.AA (!0.15 g/d) is consumed preformed in meat, eggs, andsome fish.

Omega-6 PUFAs and InflammationArguments for reduced LA intakes are based on theassumption that because CHD has an inflammatory com-ponent10 and because the omega-6 fatty acid, AA, is thesubstrate for the synthesis of a variety of proinflammatorymolecules, reducing LA intakes should reduce tissue AAcontent, which should reduce the inflammatory potentialand therefore lower the risk for CHD. The evidence,derived primarily from human studies, regarding this lineof reasoning is examined below.

AA is the substrate for the production of a wide variety ofeicosanoids (20-carbon AA metabolites). Some are proin-flammatory, vasoconstrictive, and/or proaggregatory, such asprostaglandin E2, thromboxane A2, and leukotriene B4. How-ever, others are antiinflammatory/antiaggregatory, such asprostacyclin, lipoxin A4,11 and epoxyeicosatrienoic acids.12

Epoxyeicosatrienoic acids are fatty acid epoxides producedfrom AA by a cytochrome P450 epoxygenase. Epoxyeicosa-trienoic acids also have important vasodilator properties via

The American Heart Association makes every effort to avoid any actual or potential conflicts of interest that may arise as a result of an outsiderelationship or a personal, professional, or business interest of a member of the writing panel. Specifically, all members of the writing group arerequired to complete and submit a Disclosure Questionnaire showing all such relationships that might be perceived as real or potential conflictsof interest.

This statement was approved by the American Heart Association Science Advisory and Coordinating Committee on November 6, 2008. A copy of thestatement is available at http://www.americanheart.org/presenter.jhtml?identifier"3003999 by selecting either the “topic list” link or the “chronologicallist” link (No. LS-1966). To purchase additional reprints, call 843-216-2533 or e-mail [email protected].

Expert peer review of AHA Scientific Statements is conducted at the AHA National Center. For more on AHA statements and guidelines development,visit http://www.americanheart.org/presenter.jhtml?identifier"3023366.

Permissions: Multiple copies, modification, alteration, enhancement, and/or distribution of this document are not permitted without the expresspermission of the American Heart Association. Instructions for obtaining permission are located at http://www.americanheart.org/presenter.jhtml?identifier"4431. A link to the “Permission Request Form” appears on the right side of the page.

(Circulation. 2009;119:902-907.)© 2009 American Heart Association, Inc.

Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.108.191627

902

AHA Science Advisory


LA levels predicted lower risk of stroke, particularlyischemic stroke.55 LA intakes are not associated with riskfor cancer.26 Therefore, observational studies generallysuggest an overall modest benefit of omega-6 PUFA intakeon CHD risk and no significant effect on stroke or cancer.These studies, some of which included LA intakes of up to10% to 12% of energy, contradict the supposition thathigher omega-6 PUFA intakes increase risk for CHD.

Omega-6 PUFA Consumption and CHDEvents: Randomized Controlled Trials

Several randomized trials have evaluated the effects ofreplacing saturated fatty acids with PUFAs on CHDevents.56 – 65 Intakes of PUFA (almost entirely omega-6PUFA) ranged from 11% to 21%. In addition to theinability to double-blind these studies, many had designlimitations such as small sample size (n!54),65 the provi-sion of only "50% of meals,56 outcomes composed largelyof “soft” ECG end points,59,60 randomization of sites ratherthan individuals with open enrollment and high turnover ofsubjects,59,60 use of vegetable oils that also contained theplant omega-3 fatty acid !-linolenic acid,57,59,60 and simul-taneous recommendations to increase fish and cod liver oiluse.58 Nevertheless, a meta-analysis including 6 of thesetrials56 – 60,62– 64 indicated that replacing saturated fatty ac-ids with PUFAs lowered the risk for CHD events by24%.66 Of the remaining 4 studies, 1 reported a significant45% reduction in risk,59 whereas no significant effect wasseen in the others.60,61,65

These trials tested the effect of replacing saturated fattyacids; no randomized trial has reported the effects ofreplacing carbohydrate or protein with omega-6 PUFAs onCHD risk. Although limitations are present for each trial,the combined results of these studies and the observationaltrials provide evidence that replacing saturated fatty acidor refined carbohydrate (eg, sugars, white bread, whiterice, potatoes) with omega-6 PUFAs reduces CHD risk. Onthe basis of the intakes of omega-6 PUFAs used in therandomized trials, metabolic studies, and nonhuman pri-mate studies discussed below, reductions in CHD riskmight be expected with omega-6 PUFA intakes of 10% to21% of energy compared with lower intakes, with noclinical evidence for adverse events.

Recommended Intakes of Omega-6Fatty Acids

Dietary recommendations for omega-6 PUFAs tradition-ally focused on the prevention of essential fatty aciddeficiency but are now increasingly seeking to define“optimal” intakes to reduce risk for chronic disease,particularly CHD. The Institute of Medicine’s Food andNutrition Board, in their Dietary Reference Intake Reportfor Energy and Macronutrients,67 defines an adequateintake of LA as 17 g/d for men and 12 g/d for women (5%to 6% of energy) 19 to 50 years of age, approximately thecurrent median US intake. Both the Dietary ReferenceIntake Report and the 2005 Dietary Guidelines for Amer-icans68 support an acceptable macronutrient distributionrange (the range of intakes for a particular energy source

that is associated with reduced risk of chronic diseasewhile providing adequate intakes of essential nutrients) of5% to 10% dietary energy from omega-6 PUFAs. TheThird Adult Treatment Panel of the National CholesterolEducation Program recommends PUFA consumption up to10%, noting that “there are no large populations that haveconsumed large quantities of polyunsaturated fatty acidsfor long periods. Thus, high intakes have not been provensafe in large populations; this introduces a note of cautionfor recommending high intakes.”69 On the other hand,evidence from trials in nonhuman primates has demon-strated cardiovascular benefits and no evidence of harmwith LA intakes of 25% of energy for up to 5 years,70,71

and randomized trials in humans have shown reduced CHDrisk with omega-6 PUFA intakes of 11% to 21% of energyfor up to 11 years with no evidence of harm.

Other governmental health recommendations foromega-6 fatty acid intakes (on a percent energy basis) areas follows: European Commission, 4% to 8%72; Food andAgriculture Organization/World Health Organization, 5%to 8%73; British Nutrition Foundation, 6% to 6.5% (max-imum, 10%)74; the Department of Health and Ageing,Australia and New Zealand, 4% to 5% (maximum, 10%)75;and the American Dietetic Association/Dietitians of Can-ada, 3% to 10%.76 The American Heart Association placesprimary emphasis on healthy eating patterns rather than onspecific nutrient targets.

Advice to reduce omega-6 PUFA intakes is typicallyframed as a call to lower the ratio of dietary omega-6 toomega-3 PUFAs.1– 4 Although increasing omega-3 PUFAtissue levels does reduce the risk for CHD,77,78 it does notfollow that decreasing omega-6 levels will do the same.Indeed, the evidence considered here suggests that it wouldhave the opposite effect. Higher omega-6 PUFA intakescan inhibit the conversion of !-linolenic acid to eicosa-pentaenoic acid,79 but such conversion is already quitelow,80 and whether additional small changes would havenet effects on CHD risk after the other benefits of LAconsumption are taken into account is not clear. The focuson ratios, rather than on levels of intake of each type ofPUFA, has many conceptual and biological limitations.81

ConclusionsThis advisory was undertaken to summarize the currentevidence on the consumption of omega-6 PUFAs, partic-ularly LA, and CHD risk. Aggregate data from randomizedtrials, case-control and cohort studies, and long-termanimal feeding experiments indicate that the consumptionof at least 5% to 10% of energy from omega-6 PUFAsreduces the risk of CHD relative to lower intakes. The dataalso suggest that higher intakes appear to be safe and maybe even more beneficial (as part of a low–saturated-fat,low-cholesterol diet). In summary, the AHA supports anomega-6 PUFA intake of at least 5% to 10% of energy inthe context of other AHA lifestyle and dietary recommen-dations. To reduce omega-6 PUFA intakes from theircurrent levels would be more likely to increase than todecrease risk for CHD.

904 Circulation February 17, 2009


85

LYON HEART STUDY

Okuyama H, Ichikawa Y, Sun Y, Hamazaki T, Lands WE. World Rev Nutr Diet. 2007;96:83-103.

NUTRITION MYTHS

e-Book

MITO 1

• É errado pensar-se que as pessoas com Diabetes terão que ter uma alimentação diferente das outras pessoas. A alimentação de um Diabético deve reger-se pelos princípios de uma alimentação saudável, devendo, por isso, ser equilibrada, diversificada e completa.

• Contudo, tal como para a população em geral, pode ser benéfica a adopção de alguns princípios básicos, tais como: distribuir os alimentos de forma equilibrada por várias refeições, com intervalos regulares e não omitindo refeições; beber água ao longo do dia (em média 1,5l a 2l por dia); limitar o consumo de fritos e molhos gordos, dando preferência a outras confecções culinárias, como os grelhados, os cozidos, os estufados e os assados com pouca gordura.

As pessoas com Diabetes terão que ter uma alimentação completamente diferente das pessoas sem doença.

MITO 2

• Estes alimentos são grandes fornecedores de hidratos de carbono (HC), os nutrientes que mais influenciam os níveis de glicemia após as refeições. No entanto, ao contrário dos alimentos ricos em açúcar, estes alimentos contêm HC de absorção lenta, permitindo um melhor controlo da glicemia ao longo do dia.

• A sua ingestão é indispensável, pois devem fornecer a maior parte da energia que o nosso organismo necessita, cerca de 45 a 60% das calorias totais por dia.

• Desta forma, estes alimentos devem fazer parte de todas as refeições realizadas ao longo do dia.

As pessoas com Diabetes devem evitar comer arroz, massa, batata ou pão.

MITO 2





IG E INSULINA!

Last AR, Wilson SA. Low-Carbohydrate Diet. Am Fam Physician 2006;73:1942-8

Alimento IG Dose Hidratos Carbono

CG

Abóbora 75 80 grs 6 grs 4 Beterraba 64 80 grs 7 grs 4 Cenoura 47 80 grs 6 grs 3 Batata com pel 60 150 grs 30 grs 18 Batata no forno 85 150 grs 30 grs 27 Batata frita congelada

85 150 grs 29 grs 22

Puré de Batata 74 150 grs 20 grs 15 Batata Doce 61 150 grs 28 grs 17 Mandioca 70 150 grs 57 grs 40 Inhame 37 150 grs 36 grs 13

ÍNDICE GLICÉMICO E CARGA GLICÉMICA

Foster-Powell K, Holt SH, Brand-Miller JC. Am J Clin Nutr. 2002 Jul;76(1):5-56.

Alimento IG Dose Hidratos Carbono

CG

Pão branco 70 60 grs 30 grs 21 Barra francesa 62 70 grs 42 grs 26 Pão centeio 50 60 grs 24 grs 12 Cheerios 74 30 grs 20 grs 15 Chocapic 84 30 grs 25 grs 21 CornFlakes 92 30 grs 26 grs 24 Golden Grahams 71 30 grs 25 grs 18 Special K 84 30 grs 24 grs 20 Bran Flakes 74 30 grs 16 grs 13 Cream Crackers 65 25 grs 17 grs 11 Alpen Muesli 55 30 grs 19 grs 10



Alimento IG Dose HC CG Arroz branco Uncle Ben’s 10 min 68 150 grs 37 grs 25 Arroz branco 56 150 grs 41 grs 23

Arroz branco BasmaP 58 150 grs 38 grs 22 Arroz integral 55 150 grs 33 grs 18 Millet 71 150 grs 36 grs 25 Bulgur 48 150 grs 26 grs 12 Esparguete de milho 78 180 grs 42 grs 32 FeRucine com ovo 40 180 grs 46 grs 18 Gnocchi 68 180 grs 48 grs 33 Linguini 46 180 grs 48 grs 22 Macarroni 47 180 grs 48 grs 23 Ravioli 40 180 grs 42 grs 32 SpagheV, cozido 5 min 38 180 grs 48 grs 18 SpagheV, cozido 20 min 61 180 grs 44 grs 27 Esparguete integral 37 180 grs 42 grs 16



RECOMENDAÇÃO

• Ingerir diariamente 4 a 11 porções de cereais, derivados e tubérculos segundo as recomendações da Roda dos Alimentos. Estas porções devem ser distribuídas por 5 a 7 refeições.

O número de porções recomendado depende das necessidades energéticas individuais. As crianças de 1 a 3 anos devem guiar-se pelos limites inferiores

e os homens activos e os rapazes adolescentes pelos limites superiores; a restante população deve orientar-se pelos valores intermédios.

1 Porção de Cereais e derivados, tubérculos representa: 1 pão (50g)

1 fatia fina de broa (70g)

1 e ½ batata – tamanho médio (125g)

5 colheres de sopa de cereais de

pequeno-almoço (35g)

6 bolachas Maria/água e sal (35g)

2 colheres de sopa de arroz/massa crus (35g)

4 colheres de sopa de arroz/massa

cozinhados (110g)

Nota: pesar os alimentos poderá

ser uma boa forma de compreender as

porções.

RECOMENDAÇÃO

• Ingerir diariamente 4 a 11 porções de cereais, derivados e tubérculos segundo as recomendações da Roda dos Alimentos. Estas porções devem ser distribuídas por 5 a 7 refeições.

O número de porções recomendado depende das necessidades energéticas individuais. As crianças de 1 a 3 anos devem guiar-se pelos limites inferiores

e os homens activos e os rapazes adolescentes pelos limites superiores; a restante população deve orientar-se pelos valores intermédios.

1 Porção de Cereais e derivados, tubérculos representa: 1 pão (50g)

1 fatia fina de broa (70g)

1 e ½ batata – tamanho médio (125g)

5 colheres de sopa de cereais de

pequeno-almoço (35g)

6 bolachas Maria/água e sal (35g)

2 colheres de sopa de arroz/massa crus (35g)

4 colheres de sopa de arroz/massa

cozinhados (110g)

Nota: pesar os alimentos poderá

ser uma boa forma de compreender as

porções.

IG E INSULINA!

Last AR, Wilson SA. Low-Carbohydrate Diet. Am Fam Physician 2006;73:1942-8

HIPERINSULINEMIA à RESISTÊNCIA À INSULINA

Rizza RA, et al. Diabetologia. 1985 Feb;28(2):70-5 Del Prato S, et al. Diabetologia. 1994 Oct;37(10):1025-35. Flores-Riveros JR, McLenithan JC, Ezaki O, Lane MD. Proc Natl Acad Sci 1993;90:512–6.

Zammit et al. J Nutr 2001;131:2074-77.

"

!Resistência à Insulina é a primeira alteração metabólica da Síndrome Metabólica!

!

"""""""

""

Ludwig D.JAMA 2002;287:2414–2423.

Para diagnosticar Síndrome Metabólico:

Apresentar, pelo menos, 3 sintomas

Alguns homens apresentam alterações metabólicas com perímetro da cintura entre 94-102 cm

* Propensão genética para insulinorresistência

ACSM, 2005

99

FLUXO DE LDL PARA A INTIMA É > PARA LDL PEQUENAS E DENSAS UMA VEZ NA INTIMA, LDL PEQUENAS E DENSAS SÃO SUSCEPTÍVEIS DE SOFRER OXIDAÇÃO

Cordain, 2009

0

20

40

60

80

0 10 20 30 40 50

% P

esso

as c

om

> %

sdL

DL

% Gord % CHO 75 65 55 45 35

r = -0.95 P< 0.001

Krauss RM. J Nutr 2001;131:340s-43s Griffin BA. Proc Nutr Soc 1999;58:163-69

LDL PEQUENAS E DENSAS

Effect of a low-glycaemic index–low-fat–high protein diet on theatherogenic metabolic risk profile of abdominally obese men

Jean G. Dumesnil1,2*, Jacques Turgeon1,2,3, Angelo Tremblay2,5, Paul Poirier1,2, Marcel Gilbert1,2,Louise Gagnon1,2, Sylvie St-Pierre2,5, Caroline Garneau1,3, Isabelle Lemieux1,4, Agnes Pascot1,4,

Jean Bergeron4 and Jean-Pierre Despres1,4

1Quebec Heart Institute, Laval Hospital Research Center, Quebec, Canada2Faculty of Medicine, Laval University, Quebec, Canada3Faculty of Pharmacy, Laval University, Quebec, Canada

4Lipid Research Center, CHUQ Research Center, Quebec, Canada5Division of Kinesiology, Department of Preventive and Social Medicine, Laval University, Quebec, Canada

(Received 29 October 1999 – Revised 14 May 2001 – Accepted 28 May 2001)

It has been suggested that the current dietary recommendations (low-fat–high-carbohydrate diet)may promote the intake of sugar and highly refined starches which could have adverse effects on themetabolic risk profile.We have investigated the short-term (6-d) nutritional and metabolic effects ofanad libitum low-glycaemic index–low-fat–high-protein diet (prepared according to theMontignacmethod) comparedwith theAmericanHeart Association (AHA) phase I diet consumed ad libitum aswell as with a pair-fed session consisting of the same daily energy intake as the former but with thesame macronutrient composition as the AHA phase I diet. Twelve overweight men (BMI 33:0(SD 3:5) kg/m2)without other diseaseswere involved in three experimental conditionswithaminimalwashout period of 2 weeks separating each intervention. By protocol design, the first two conditionswere administered randomly whereas the pair-fed session had to be administered last. During the adlibitum version of the AHA diet, subjects consumed 11695:0 (SD 1163:0) kJ/d and this diet induced a28% increase in plasma triacylglycerol levels (1:77 (SD 0:79) v. 2:27 (SD 0:92)mmol/l,P,0:05) anda 10% reduction in plasmaHDL-cholesterol concentrations (0:92 (SD 0:16) v. 0:83 (SD 0:09)mmol/l,P,0:01) which contributed to a significant increase in cholesterol:HDL-cholesterol ratio (P,0:05),this lipid index being commonly used to assess the risk of coronary heart disease. In contrast, the low-glycaemic index–low-fat–high-protein diet consumed ad libitum resulted in a spontaneous 25%decrease (P,0:001) in total energy intake which averaged 8815:0 (SD 738:0) kJ/d. As opposed to theAHA diet, the low-glycaemic index–low-fat–high-protein diet produced a substantial decrease(235%) in plasma triacylglycerol levels (2:00 (SD 0:83) v. 1:31 (SD 0:38) mmol/l, P,0:0005), asignificant increase (+1:6%) inLDLpeak particle diameter (251 (SD 5) v. 255 (SD 5) A,P,0:02) andmarked decreases in plasma insulin levels measured either in the fasting state, over daytime andfollowing a 75g oral glucose load. During the pair-fed session, in which subjects were exposed to adiet with the same macronutrient composition as the AHA diet but restricted to the same energyintake as during the low-glycaemic index–low-fat–high-proteindiet, therewas a trend for a decreasein plasma HDL-cholesterol levels which contributed to the significant increase in cholesterol:HDL-cholesterol ratio noted with this condition. Furthermore, a marked increase in hunger (P,0:0002)and a significant decrease in satiety (P,0:007) were also noted with this energy-restricted diet.Finally, favourable changes in themetabolic risk profile notedwith the ad libitum consumptionof thelow-glycaemic index–low-fat–high-protein diet (decreases in triacyglycerols, lack of increase incholesterol:HDL-cholesterol ratio, increase in LDL particle size) were significantly different fromthe response of these variables to theAHAphase I diet. Thus, a low-glycaemic index–low-fat–high-protein content dietmay have unique beneficial effects comparedwith the conventionalAHAdiet forthe treatment of the atherogenic metabolic risk profile of abdominally obese patients. However, thepresent study was a short-term intervention and additional trials are clearly needed to document thelong-term efficacy of this dietary approach with regard to compliance and effects on the metabolicrisk profile.

Low-glycaemic index diet: Atherogenesis: Abdominal obesity: Dietary interventions

*Corresponding author: Dr Jean G. Dumesnil, fax +1 418 656 4562, email [email protected]: AHA, American Heart Association.

British Journal of Nutrition (2001), 86, 557–568 DOI: 10.1079/BJN2001427q Nutrition Society 2001

INSULINA!

Although the low-glycaemic index–low-fat–high-protein content diet has enjoyed considerable popularsuccess (Montignac, 1994), especially in Europe, therehave been, to our knowledge, few scientific investi-gations of this method, particularly when offered tosubjects on an ad libitum basis. The fact that this dietappeared to be successful while being consumed adlibitum appeared particularly intriguing and was thebasic anecdotal observation which led us to undertakethe present short-term study. Hence, an important andrelatively unexpected result of our study was that this dietproduced a substantial (about 25%) reduction in energyintake without any change in hunger or satiety, as measuredobjectively using a visual analogue scale (Doucet et al.2000). To our knowledge, a reduction in spontaneousenergy intake of that magnitude without inducing hungercannot be achieved without pharmacotherapy, and chronichunger is a major barrier to compliance when patients areasked to follow a reduced-energy diet.

A significant reduction inweight and decreases inwaist andhip circumferences were observed after only 1 week on thelow-glycaemic index–low-fat–high-protein diet. This findingshould not be overemphasized as initial changes in weight donot necessarily reflect adequately the loss of body fat. Therestricted AHA phase I diet, although clamped for energyintake with the low-glycaemic index–low-fat–high-protein

diet, also produced weight loss but had a less favourableimpact on the metabolic risk profile than the low-glycaemicindex diet. The notion that the glycaemic index of foods isimportant is far from new (Jenkins et al. 1994). Our resultsare concordant with the numerous studies which havesuggested that a diet rich in carbohydrates with a lowglycaemic index may be helpful for the management ofinsulin-resistant or dyslipidaemic patients.However, the contribution of such a diet in inducing a

substantial reduction in spontaneous energy intake is a muchless studied issue. Ludwig et al. (1999) have recentlyreported a 81% greater voluntary energy intake after a high-glycaemic index meal (5:8MJ) than after a low-glycaemicindex meal (3:2MJ), a finding which is concordant with theresults of the present study. In the present study, the low-glycaemic–low-fat–high-protein diet was less energy-dense than the AHA diet. However, this factor did notentirely explain the reduced energy intake on the lowglycaemic index diet since this difference in energy intakeremained statistically significant after adjustment for energydensity. We are also aware that absence of standardizationof the meal frequency of the diet conditions 1 and 2 ascompared with condition 3 may have influenced hunger andsatiety ratings. Long-term studies will also be necessary todetermine whether the spontaneous reduction in energyintake is maintained over time and which are the key factors

Fig. 2. Mean fasting plasma insulin (a), apolipoprotein B (b) and LDL size (c) before (A) and after (B) each of the three dietary regimens.Mean changes in fasting insulin (d), apolipoprotein B (e) and LDL size (f) in response to the three dietary regimens. Standard errors arerepresented by vertical bars. *Mean values were significantly different from baseline, P , 0:05: a,bMean responses to diets with unlike letterswere significantly different, P,0:05. For details of diets and procedures, see p. 558–559.

Low-glycaemic index diet and metabolic risk profile 563











Dieta clássica ad libitum

Dieta low GI ad libitum

Dieta clássica hipocalórica

DISLIPIDEMIA!

Although the low-glycaemic index–low-fat–high-protein content diet has enjoyed considerable popularsuccess (Montignac, 1994), especially in Europe, therehave been, to our knowledge, few scientific investi-gations of this method, particularly when offered tosubjects on an ad libitum basis. The fact that this dietappeared to be successful while being consumed adlibitum appeared particularly intriguing and was thebasic anecdotal observation which led us to undertakethe present short-term study. Hence, an important andrelatively unexpected result of our study was that this dietproduced a substantial (about 25%) reduction in energyintake without any change in hunger or satiety, as measuredobjectively using a visual analogue scale (Doucet et al.2000). To our knowledge, a reduction in spontaneousenergy intake of that magnitude without inducing hungercannot be achieved without pharmacotherapy, and chronichunger is a major barrier to compliance when patients areasked to follow a reduced-energy diet.

A significant reduction inweight and decreases inwaist andhip circumferences were observed after only 1 week on thelow-glycaemic index–low-fat–high-protein diet. This findingshould not be overemphasized as initial changes in weight donot necessarily reflect adequately the loss of body fat. Therestricted AHA phase I diet, although clamped for energyintake with the low-glycaemic index–low-fat–high-protein

diet, also produced weight loss but had a less favourableimpact on the metabolic risk profile than the low-glycaemicindex diet. The notion that the glycaemic index of foods isimportant is far from new (Jenkins et al. 1994). Our resultsare concordant with the numerous studies which havesuggested that a diet rich in carbohydrates with a lowglycaemic index may be helpful for the management ofinsulin-resistant or dyslipidaemic patients.However, the contribution of such a diet in inducing a

substantial reduction in spontaneous energy intake is a muchless studied issue. Ludwig et al. (1999) have recentlyreported a 81% greater voluntary energy intake after a high-glycaemic index meal (5:8MJ) than after a low-glycaemicindex meal (3:2MJ), a finding which is concordant with theresults of the present study. In the present study, the low-glycaemic–low-fat–high-protein diet was less energy-dense than the AHA diet. However, this factor did notentirely explain the reduced energy intake on the lowglycaemic index diet since this difference in energy intakeremained statistically significant after adjustment for energydensity. We are also aware that absence of standardizationof the meal frequency of the diet conditions 1 and 2 ascompared with condition 3 may have influenced hunger andsatiety ratings. Long-term studies will also be necessary todetermine whether the spontaneous reduction in energyintake is maintained over time and which are the key factors

Fig. 2. Mean fasting plasma insulin (a), apolipoprotein B (b) and LDL size (c) before (A) and after (B) each of the three dietary regimens.Mean changes in fasting insulin (d), apolipoprotein B (e) and LDL size (f) in response to the three dietary regimens. Standard errors arerepresented by vertical bars. *Mean values were significantly different from baseline, P , 0:05: a,bMean responses to diets with unlike letterswere significantly different, P,0:05. For details of diets and procedures, see p. 558–559.

Low-glycaemic index diet and metabolic risk profile 563











A low–glycemic index diet combined with exercise reduces insulinresistance, postprandial hyperinsulinemia, and glucose-dependentinsulinotropic polypeptide responses in obese, prediabetic humans1–4

Thomas PJ Solomon, Jacob M Haus, Karen R Kelly, Marc D Cook, Julianne Filion, Michael Rocco, Sangeeta R Kashyap,Richard M Watanabe, Hope Barkoukis, and John P Kirwan

ABSTRACTBackground: The optimal lifestyle intervention that reverses dia-betes risk factors is not known.Objective: We examined the effect of a low–glycemic index (GI)diet and exercise intervention on glucose metabolism and insulinsecretion in obese, prediabetic individuals.Design: Twenty-two participants [mean 6 SEM age: 66 6 1 y;body mass index (in kg/m2): 34.4 6 0.8] underwent a 12-wkexercise-training intervention (1 h/d for 5 d/wk at ’85% of max-imum heart rate) while randomly assigned to receive either a low-GIdiet (LoGIX; 40 6 0.3 units) or a high-GI diet (HiGIX; 80 6 0.6units). Body composition (measured by using dual-energy X-rayabsorptiometry and computed tomography), insulin sensitivity(measured with a hyperinsulinemic euglycemic clamp with[6,6-2H2]-glucose), and oral glucose–induced insulin and incretinhormone secretion were examined.Results: Both groups lost equal amounts of body weight (28.8 60.9%) and adiposity and showed similar improvements in peripheraltissue (+76.2 6 14.9%) and hepatic insulin sensitivity (+27.1 67.1%) (all P , 0.05). However, oral glucose–induced insulin secre-tion was reduced only in the LoGIX group (6.59 6 0.86 nmol in theprestudy compared with 4.70 6 0.67 nmol in the poststudy, P ,0.05), which was a change related to the suppressed postprandialresponse of glucose-dependent insulinotropic polypeptide. Whencorrected for changes in b cell glucose exposure, changes in insulinsecretion were attenuated in the LoGIX group but became signifi-cantly elevated in the HiGIX group.Conclusions: Although lifestyle-induced weight loss improves in-sulin resistance in prediabetic individuals, postprandial hyperinsu-linemia is reduced only when a low-GI diet is consumed. In contrast,a high-GI diet impairs pancreatic b cell and intestinal K cell functiondespite significant weight loss. These findings highlight the impor-tant role of the gut in mediating the effects of a low-GI diet on type2 diabetes risk reduction. Am J Clin Nutr 2010;92:1359–68.

INTRODUCTION

The reversal of progressive pancreatic b cell dysfunction andoral glucose intolerance is a primary goal for prediabetic in-dividuals (1). However, solely treating glucose tolerance andperipheral and hepatic insulin resistance (IR) in overweight andobese individuals may delay, but is unlikely to prevent, the fu-ture onset of type 2 diabetes (T2D). Indeed, evidence suggests

that preserving b cell function in these at-risk populations isa critical factor in the prevention of T2D onset (2). Therefore,therapeutic management of such individuals should consider allprocesses involved in maintaining glucose homeostasis.

Although the recent Diabetes Prevention Program OutcomesStudy underlined the large effect that intensive lifestyle in-tervention can have on the prevention of diabetes onset inoverweight individuals (3), the mechanisms by which lifestyleintervention prevents the development of glucose intolerance arenot fully understood. Exercise training interventions can suc-cessfully improve glucose tolerance (4, 5), and several groupshave shown that exercise can reverse peripheral tissue and hepaticIR (6–9). In addition, it has been shown that exercise training canalter insulin secretion and improve b cell function in obesehumans (10–13). Caloric restriction has also been shown to im-prove these components of glucose tolerance (8, 14). However,certain nutrients (high-fat feeding) and elevated glycemic con-centrations can induce IR and impair b cell function (15–17),whereas a high-carbohydrate and high-fiber diet can improveperipheral insulin sensitivity (18). Thus, dietary composition mayplay a critical role in determining the ultimate success of suchinterventions.

The concept of a dietary glycemic index (GI) has receivedmuch attention: the consumption of high-GI diets may increase

1 From the Department of Pathobiology, Lerner Research Institute (TPJS,JMH, KRK, MDC, JF, and JPK), and the Departments of CardiovascularMedicine (MC), Endocrinology, Diabetes, and Metabolism (SRK), and Gas-troenterology/Hepatology (JPK), Cleveland Clinic, Cleveland, OH; the De-partments of Physiology (JMH and JPK) and Nutrition (KRK and JPK), CaseWestern Reserve University, Cleveland, OH; and the Department of Preven-tive Medicine, Physiology, and Biophysics, University of Southern Califor-nia, Los Angeles, CA (RMW).

2 Supported by the National Institutes of Health (NIH) (grants RO1AG12834; to JPK) and the NIH National Center for Research Resources(Cleveland, OH) (Clinical and Translational Science Award1UL1RR024989).

3 Current address for TPJS: Centre of Inflammation and Metabolism, Rig-shospitalet, Section 7641, Blegdamsvej 9, 2100 København Ø, Denmark.E-mail: [email protected].

4 Address correspondence to JP Kirwan, Department of Pathobiology,Lerner Research Institute, Cleveland Clinic, 9500 Euclid Avenue/NE-40,Cleveland, OH 44195. E-mail: [email protected].

Received May 14, 2010. Accepted for publication September 19, 2010.First published online October 27, 2010; doi: 10.3945/ajcn.2010.29771.

Am J Clin Nutr 2010;92:1359–68. Printed in USA. ! 2010 American Society for Nutrition 1359

at Lund University Libraries on M

ay 5, 2011w

ww

.ajcn.orgD

ownloaded from




INTRODUCTION












ay 5, 2011w

ww

.ajcn.orgD

ownloaded from




INTRODUCTION












ay 5, 2011w

ww

.ajcn.orgD

ownloaded from

SÓ HOUVE MELHORIA DA

INSULINA POSTPRANDIAL NO GRUPO DE

BAIXO IG

Can a Low–Glycemic Index Diet Reduce theNeed for Insulin in Gestational DiabetesMellitus?A randomized trial

ROBERT G. MOSES, MD1

MEGAN BARKER, APD1

MEAGAN WINTER, APD1

PETER PETOCZ, PHD2

JENNIE C. BRAND-MILLER, PHD3

OBJECTIVE — A low–glycemic index diet is effective as a treatment for individuals withdiabetes and has been shown to improve pregnancy outcomes when used from the first trimester.A low–glycemic index diet is commonly advised as treatment for women with gestationaldiabetes mellitus (GDM). However, the efficacy of this advice and associated pregnancy out-comes have not been systematically examined. The purpose of this study was to determinewhether prescribing a low–glycemic index diet for women with GDM could reduce the numberof women requiring insulin without compromise of pregnancy outcomes.

RESEARCHDESIGNANDMETHODS — All women with GDM seen over a 12-monthperiod were considered for inclusion in the study. Women (n ! 63) were randomly assigned toreceive either a low–glycemic index diet or a conventional high-fiber (and higher glycemicindex) diet.

RESULTS — Of the 31 women randomly assigned to a low–glycemic index diet, 9 (29%)required insulin. Of the women randomly assigned to a higher–glycemic index diet, a signifi-cantly higher proportion, 19 of 32 (59%), met the criteria to commence insulin treatment (P !0.023). However, 9 of these 19 women were able to avoid insulin use by changing to a low–glycemic index diet. Key obstetric and fetal outcomes were not significantly different.

CONCLUSIONS — Using a low–glycemic index diet for women with GDM effectivelyhalved the number needing to use insulin, with no compromise of obstetric or fetal outcomes.

Diabetes Care 32:996–1000, 2009

G estational diabetes mellitus (GDM) isdefined as any degree of glucose in-tolerance with onset or first recogni-

tion during pregnancy (1). GDM isassociated with an increase in adversepregnancy outcomes, and the advantagesof treatment on these outcomes have beenidentified (2). All women with GDMshould have medical nutrition therapy(MNT) with the objective of achievingand maintaining blood glucose levels asclose to the normal range as possible (3).MNT needs to be individualized andshould be based on carbohydrate (CHO)

distribution and ideally on the results ofself-monitoring of blood glucose (SMBG).For the purposes of SMBG, a combinationof the fasting and postprandial glucoselevels is desirable.

When MNT alone is unable to keepthe results of SMBG within predeter-mined target ranges, alternative therapiesare required. Although there is some evi-dence that both glyburide (4) and met-formin (5) can be used, the overwhelmingexperience has been with insulin. How-ever, the potential use of insulin can be a

source of both anxiety and of resistance totreatment change.

In normal subjects, mixed mealsbased on low–glycemic index foods leadto a reduction in postprandial glycemia(6). We have previously demonstrated innormal pregnant women that a diet basedon low–glycemic index foods was sus-tainable and resulted in more favorablefetal outcomes (7). The aim of this studywas to examine whether a low–glycemicindex diet used as MNT for women withGDM could result in a reduced need forinsulin use during pregnancy with nocompromise of obstetric and fetaloutcomes.

RESEARCH DESIGN ANDMETHODS — The study was con-ducted in the city of Wollongong, NewSouth Wales, Australia, a coastal city witha population of "280,000 people situ-ated about 50 miles south of Sydney. TheAustralasian Diabetes in Pregnancy Soci-ety (ADIPS) recommends that all preg-nant women should be tested for GDM(8). Unless indicated earlier, women havea 75-g glucose tolerance test at the begin-ning of the third trimester with glucosesamples taken after fasting and at 2 h.GDM is diagnosed if the fasting glucose is!5.5 mmol/l ("100 mg/dl) and/or the2-h glucose is !8.0 mmol/l ("145 mg/dl)(9). Virtually all women with GDM areseen at the Diabetes Center by a diabetesnurse educator and a specialist dietitian.All women seen over a 12-month period(October 2007–September 2008) wereconsidered. There are "3,300 deliverieseach year in the area, including those inboth the public and private hospitals. Theprevalence of GDM is "7%, and there is#90% compliance with universal testing(10).

Inclusion criteria were age 18 – 40years (inclusive), singleton pregnancy, noprevious GDM, nonsmoker, diagnosis ofGDM and seen for the first dietary visitbetween 28 and 32 weeks of gestation,and ability to follow the protocol require-ments. Exclusion criteria included any

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

From the 1Illawarra Diabetes Service, South Eastern Sydney and Illawarra Area Health Service, Wollongong,New South Wales, Australia; the 2Department of Statistics, Macquarie University, Sydney, New SouthWales, Australia; and the 3Human Nutrition Unit, University of Sydney, New South Wales, Australia.

Corresponding author: Professor Robert G. Moses, [email protected] 3 January 2009 and accepted 27 February 2009.Published ahead of print at http://care.diabetesjournals.org on 11 March 2009. DOI: 10.2337/dc09-0007.© 2009 by the American Diabetes Association. Readers may use this article as long as the work is properly

cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be herebymarked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

C l i n i c a l C a r e / E d u c a t i o n / N u t r i t i o n / P s y c h o s o c i a l R e s e a r c hO R I G I N A L A R T I C L E

996 DIABETES CARE, VOLUME 32, NUMBER 6, JUNE 2009



MEGAN BARKER, APD1

MEAGAN WINTER, APD1

PETER PETOCZ, PHD2






Diabetes Care 32:996–1000, 2009









! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !









MEGAN BARKER, APD1

MEAGAN WINTER, APD1

PETER PETOCZ, PHD2






Diabetes Care 32:996–1000, 2009









! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !







REDUZIU AS NECESSIDADES DE INSULINA SEM AFECTAR NEGATIVAMENTE A MÃE E O

FETO

See corresponding editorial on page 949.

Effects of an ad libitum low-glycemic load diet on cardiovasculardisease risk factors in obese young adults1–3

Cara B Ebbeling, Michael M Leidig, Kelly B Sinclair, Linda G Seger-Shippee, Henry A Feldman, and David S Ludwig

ABSTRACTBackground: The optimal nutritional approach for the prevention ofcardiovascular disease among obese persons remains a topic of in-tense controversy. Available approaches range from conventionallow-fat to very-low-carbohydrate diets.Objective: The aim of this pilot study was to evaluate the efficacyof an ad libitum low-glycemic load diet, without strict limitation oncarbohydrate intake, as an alternative to a conventional low-fat diet.Design: A randomized controlled trial compared 2 dietary treat-ments in obese young adults (n ! 23) over 12 mo. The experimentaltreatment emphasized ad libitum consumption of low-glycemic-index foods, with 45–50% of energy from carbohydrates and 30–35% from fat. The conventional treatment was restricted in energy(250–500 kcal/d deficit) and fat ("30% of energy), with 55–60% ofenergy from carbohydrate. We compared changes in study outcomesby repeated-measures analysis of log-transformed data and ex-pressed the results as mean percentage change.Results: Body weight decreased significantly over a 6-mo intensiveintervention in both the experimental and conventional diet groups(#8.4% and #7.8%, respectively) and remained below baseline at12 mo (#7.8% and #6.1%, respectively). The experimental dietgroup showed a significantly greater mean decline in plasma tria-cylglycerols than did the conventional diet group (#37.2% and#19.1%, respectively; P ! 0.005). Mean plasminogen activatorinhibitor 1 concentrations decreased (#39.0%) in the experimentaldiet group but increased (33.1%) in the conventional diet group (P !0.004). Changes in cholesterol concentrations, blood pressure, andinsulin sensitivity did not differ significantly between the groups.Conclusion: An ad libitum low-glycemic load diet may be moreefficacious than a conventional, energy-restricted, low-fat diet inreducing cardiovascular disease risk. Am J Clin Nutr 2005;81:976–82.

KEY WORDS Obesity, glycemic index, glycemic load, di-etary composition, weight-reducing diet, cholesterol, triacylglyc-erol, plasminogen activator inhibitor 1, PAI-1, young adults

INTRODUCTION

The alarming prevalence of obesity and the associated risk ofcardiovascular disease (CVD) have been well documented (1)and extensively publicized in the United States. As a result,millions of obese adults are following weight-loss diets. Re-cently, Atkins-type very-low-carbohydrate diets have rapidlygrown in popularity (2), although low-fat diets remain the cor-nerstone of conventional treatment based on clinical practicerecommendations (3, 4). Whereas a few studies have suggested

that carbohydrate-restricted diets may have significantly greaterbenefits than do low-fat diets in reducing CVD risk (5, 6), thereis widespread concern regarding the safety and long-term effi-cacy of severe carbohydrate restriction (7, 8).

A low-glycemic load (GL) diet, containing unrestrictedamounts of carbohydrate from low-glycemic index (GI) foods,represents an alternative to low-fat diets on the one hand and tolow- carbohydrate diets on the other. The GI is defined as theincremental area under the blood glucose response curve afterconsumption of 50 g of available carbohydrate from a test food,divided by the area under the curve after consumption of 50 g ofcarbohydrate from a reference food (ie, glucose or white bread) (9).The GL is the arithmetic product of the amount of carbohydrateconsumed and the GI (10) and thus describes the overall effects ofboth quantity and source of carbohydrate on postprandial glycemia(11). Risk of CVD has been inversely associated with dietary GI orGL in some (12–15) but not all (16) epidemiologic studies. More-over, whereas several short-term intervention studies have de-scribed beneficial effects of low-GI diets on blood lipids in over-weight adults (17–20) and on the capacity for fibrinolysis in diabeticpatients (21, 22), the long-term efficacy of low-GL diets in reducingCVD risk has not previously been evaluated (23).

The aim of this pilot study was to evaluate the efficacy of anexperimental ad libitum low-GL diet. We hypothesized that theexperimental diet would have a more beneficial effect on CVDrisk factors than would a conventional, energy-restricted, low-fatdiet over a 12-mo intervention.

SUBJECTS AND METHODS

Screening and enrollment

The protocol was approved by the institutional review board atChildren’s Hospital Boston, and written informed consent wasobtained from each subject. Inclusion criteria included: age be-tween 18 and 35 y, body mass index (BMI; in kg/m2) $27, body

1 From the Division of Endocrinology, Department of Medicine, Chil-dren’s Hospital, Boston, MA.

2 Supported by grant no. R01 DK59240 (to DSL) and grant no. K01DK62237 (to CBE) from the National Institute of Diabetes and DigestiveKidney Diseases, the Charles H Hood Foundation (Boston, MA), and grant no.M01 RR02172 from the National Institutes of Health to support the GeneralClinical Research Center at Children’s Hospital Boston (Boston, MA).

3 Reprints not available. Address correspondence to DS Ludwig, Depart-ment of Medicine, Children’s Hospital, 300 Longwood Avenue, Boston, MA02115. E-mail: [email protected].

Received August 18, 2004.Accepted for publication December 3, 2004.

976 Am J Clin Nutr 2005;81:976–82. Printed in USA. © 2005 American Society for Clinical Nutrition


ay 4, 2011w

ww

.ajcn.orgD

ownloaded from






INTRODUCTION














ay 4, 2011w

ww

.ajcn.orgD

ownloaded from






INTRODUCTION













at Lund University Libraries on May 4, 2011

www.ajcn.orgDownloaded from

diet composition on weight loss, although adjustment of otherdietary variables for energy intake may partially correct for un-derreporting. When calculating GL from self-report data, werelied on published GI values (31), many which were derivedfrom studies conducted in countries where foods may differ fromthose consumed in the United States. An attrition rate of 32.4%,although considered problematic in terms of drawing unbiasedconclusions (58), is similar to rates observed in previous long-term dietary intervention studies (5, 6).

In conclusion, a low-GL diet containing moderate amounts ofcarbohydrate from low-GI sources may be more efficacious thana conventional low-fat diet in reducing CVD risk. The greaterbenefits in response to an ad libitum diet, compared with anenergy-restricted diet, are particularly noteworthy. This pilotstudy provides a rationale for conducting long-term, larger-scalestudies comparing the effects of low-GL, low-fat, and very-low-carbohydrate diets on CVD risk among obese persons.

We thank Gary Bradwin for analysis of plasma lipids and plasminogenactivator inhibitor 1, Catherine Murphy for help with data collection, andIrena Clark and Erica Garcia-Lago for technical assistance.

All authors contributed to the interpretation of results. CBE and DSLdesigned the study, provided oversight, and wrote the manuscript. MML wasresponsible for dietary counseling. KBS and LGS conducted the process

evaluations to assess adherence to diet prescriptions. HAF provided consul-tation on statistical analysis of the data. None of the authors had any personalor financial conflict of interest.

REFERENCES1. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM.

Prevalence of overweight and obesity among US children, adolescents,and adults, 1999–2002. JAMA 2004;291:2847–50.

2. Butler D. Science of dieting: slim pickings. Nature 2004;428:252–4.3. Clinical guidelines on the identification, evaluation, and treatment of

overweight and obesity in adults: executive summary. Am J Clin Nutr1998;68:899–917.

4. Klein S, Sheard NF, Pi-Sunyer X, et al. Weight management throughlifestyle modification for the prevention and management of type 2diabetes: rationale and strategies. Diabetes Care 2004;27:2067–73.

5. Foster GD, Wyatt HR, Hill JO, et al. A randomized trial of a low-carbohydrate diet for obesity. N Engl J Med 2003;348:2082–90.

6. Stern L, Iqbal N, Seshadri P, et al. The effects of low-carbohydrate versusconventional weight loss diets in severely obese adults: one-yearfollow-up of a randomized trial. Ann Intern Med 2004;140:778–85.

7. St. Jeor ST, Howard BV, Prewitt TE, Bovee V, Bazzarre T, Eckel RH.Dietary protein and weight reduction: a statement for healthcare profes-sionals from the Nutrition Committee of the Council on Nutrition, Phys-ical Activity, and Metabolism of the American Heart Association. Cir-culation 2001;104:1869–74.

8. Bravata DM, Sanders L, Huang J, et al. Efficacy and safety of low-carbohydrate diets: a systematic review. JAMA 2003;289:1837–50.

TABLE 3Study outcomes1

Variable

Group P2

Experimental diet(n ! 11)

Conventional diet(n ! 12) Group Time

Group " timeinteraction

Weight 0.18 #0.001 0.89Interim3 $8.4 ($11.4, $5.3) $7.8 ($10.7, $4.9)12 mo $7.8 ($13.0, $2.2) $6.1 ($11.2, $0.7)

Total cholesterol 0.90 0.06 0.22Interim $9.9 ($16.7, $2.5) $2.1 ($9.2, 5.5)12 mo $8.5 ($17.4, 1.5) $6.2 ($15.0, 3.5)

LDL cholesterol 0.85 0.17 0.59Interim $9.1 ($18.6, 1.4) $2.6 ($12.3, 8.2)12 mo $9.7 ($21.6, 3.9) $7.4 ($19.1, 6.0)

HDL cholesterol 0.41 0.08 0.20Interim 2.3 ($6.0, 11.3) $0.3 ($8.1, 8.2)12 mo 12.2 (2.9, 22.3) 1.1 ($6.9, 9.8)

Triacylglycerols 0.96 #0.001 0.005Interim $35.4 ($44.6, $24.7) $7.1 ($19.8, 7.6)12 mo $37.2 ($47.7, $24.5) $19.1 ($32.2, $3.6)

PAI-1 0.78 0.11 0.004Interim $58.3 ($74.7, $31.3) 30.4 ($19.2, 110.4)12 mo $39.0 ($70.2, 24.9) 33.1 ($32.9, 164.3)

Systolic blood pressure 0.78 0.81 0.99Interim $0.9 ($5.9, 4.2) $0.5 ($5.3, 4.4)12 mo 0.2 ($4.7, 5.3) 0.6 ($4.1, 5.5)

Diastolic blood pressure 0.84 0.72 0.82Interim $2.0 ($7.2, 3.4) 0.3 ($4.8, 5.6)12 mo $0.3 ($6.2, 6.0) 1.4 ($4.4, 7.6)

Insulin sensitivity index 0.32 #0.001 0.94Interim 6.4 (1.5, 11.5) 5.8 (1.1, 10.7)12 mo 10.4 (3.6, 17.6) 8.7 (2.3, 15.5)

1 Mean change in log-transformed variable at 6 and 12 mo (b), retransformed to percentage change [100% " (exp(b) $1)], with 95% confidence limits.Repeated-measures ANOVA was used to account for within-subject correlations.

2 Testing for overall difference in level between experimental and conventional groups (main effect of group), change over time (main effect of time, 2df), and difference in time course between groups (group " time interaction, 2 df).

3 Data collected at 6 mo.

AD LIBITUM LOW-GLYCEMIC LOAD DIET AND CVD RISK 981


ay 4, 2011w

ww

.ajcn.orgD

ownloaded from

% %

fat diet, and adjustment for change inbody composition did not alter the REEeffect. Of particular significance, differ-ences in dietary protein of the same orgreater magnitude as that used in ourstudy did not result in any differences inREE, either following weight mainte-nance or weight loss, in 5 previous ar-ticles.27-31

A second methodological issue is thatmeasurement of REE was made dur-ing ongoing weight loss. The magni-tude of observed effect could changewith weight stabilization and addi-tional research is needed to assess thispossibility. Nevertheless, the physi-ological adaptations that occur duringactive weight loss may be especially rel-evant to understanding why most obeseindividuals become noncompliant withconventional energy-restricted dietslong before a normal body weight hasbeen reached. Indeed, diet-induced dif-ferences in REE were observed after ourparticipants had lost less than half oftheir excess adiposity, and after just 1week of energy restriction in a previ-ous study.18

The difference in REE is too small toaccount for any significant change inbody composition over the short term.For example, 80 kcal/d over 10 weeks(5600 kcal) would amount to less than1 kg of body weight. Thus, our studydoes not support claims that popular di-ets can cause rapid weight loss by induc-ing major shifts in energy metabolism.

Nevertheless, the REE difference herecould amount to several pounds ofweight change per year, given this effectwould persist over the long term. Forcomparative purposes, an energy bal-ance of –80 kcal/d could be obtained bywalking approximately 1 mile/d or by de-creasing sugar-sweetened soft drink con-sumption 6 oz/d. Indeed, this differ-ence (560 kcal/wk) would explain mostof the mean difference in rate of weightloss between groups (0.09 kg/wk ! 7500kcal/kg = 675 kcal/wk).

A potentially more important ques-tion is whether the magnitude of changein REE during weight loss would pre-dict likelihood of achieving and main-taining clinically significant weight loss.Some studies32,33 but not all34 suggestan inverse relationship between REEand weight gain or regain. An indi-vidual experiencing a larger decline inREE during weight loss may feel morefatigued, cold, and hungry than an in-dividual experiencing a small decline,and these symptoms may make com-pliance with dietary energy restrictionincreasingly difficult over time.

The physiological mechanisms relat-ing dietary composition to REE duringweight loss remain speculative but mayinvolve altered availability of metabolicfuels. Blood glucose and free fatty acidsare reduced in the postabsorptive phasefollowing a high– vs low–glycemic in-dex meal, and this reduction can be suf-ficient in magnitude to trigger release of

stress hormones.12 Low circulating con-centrations of metabolic substrate mightdirectly impair energy metabolism at thecellular level, as occurs with frank hy-poglycemia.35 Alternatively, the de-crease in REE may come from neuroen-docrinological adaptations designed toconserve energy, involving thyroid hor-mone, growth hormone, sex hor-mones, or leptin (an adipocyte-derivedfactor that acts in the hypothalamus)3,5;lack of data on these hormones com-prises a limitation of our study. Inter-estingly, rodents treated with nutrient-controlled high–glycemic index dietscompared with low–glycemic indexdiets demonstrate an increase in meta-bolic efficiency analogous to that ob-served by our participants taking thelow-fat (high–glycemic index) diet.36

Epidemiological analyses have foundassociations between glycemic load andhigh triglycerides, low high-density li-poprotein cholesterol, and elevated C-reactive protein levels.37 In 1 study,38

individuals in the highest vs lowestquintile of glycemic load had double therisk of developing heart disease, aftercontrolling for potentially confound-ing factors. However, these effects havenot previously been examined in inter-ventional studies. We found that dur-ing weight loss, a diet focused on gly-cemic load reduction produced greaterimprovements in several important car-diovascular disease–related and diabe-tes mellitus–related end points than a

Table 5. Cardiovascular Disease Risk Factors Before and After 10% Weight Loss by Dietary Treatment GroupMean (SE)

P Value*

Low-Fat Diet Group (n = 17) Low–Glycemic Load Diet Group (n = 22)

Baseline Posttreatment% Change(Adjusted) Baseline Posttreatment

% Change(Adjusted)

HOMA score 1.45 (0.20) 1.10 (0.13) –15.8 (5.13) 1.50 (0.18) 0.97 (0.11) –33.9 (4.51) .01Triglycerides, mg/dL† 92.4 (9.47) 102.3 (8.11) 16.2 (5.24) 78.3 (8.40) 72.4 (7.19) –3.5 (4.63) .01HDL-C, mg/dL† 49.4 (3.61) 44.1 (2.41) –8.1 (3.49) 46.9 (3.20) 42.2 (2.14) –8.9 (3.09) .87LDL-C, mg/dL† 124.3 (9.86) 104.6 (9.73) –15.0 (4.12) 138.7 (9.75) 115.9 (8.63) –16.1 (3.65) .84C-reactive protein, mg/dL 0.19 (0.06) 0.13 (0.04) –5.1 (13.61) 0.28 (0.06) 0.10 (0.03) –47.7 (11.94) .03Systolic BP, mm Hg 107.5 (2.90) 104.6 (2.35) –3.1 (1.32) 110.4 (2.55) 102.3 (2.06) –6.4 (1.16) .07Diastolic BP, mm Hg 67.8 (2.03) 66.2 (1.80) –2.5 (1.61) 69.2 (1.78) 64.2 (1.58) –6.5 (1.42) .07Mean arterial pressure, mm Hg 94.1 (2.48) 91.7 (2.03) –3.0 (1.27) 96.6 (2.18) 89.5 (1.78) –6.5 (1.12) .04Abbreviations: BP, blood pressure; HDL-C, high-density lipoprotein cholesterol; HOMA, homeostasis model assessment; LDL-C, low-density lipoprotein.SI conversions: To convert HDL-C and LDL-C to mmol/L, multiply by 0.0259; triglycerides to mmol/L, multiply by 0.0113.*Effect of dietary treatment on % change.†For low-fat diet group, n=11; for low–glycemic load diet group, n=14.

LOW–GLYCEMIC LOAD DIET AND RESTING ENERGY EXPENDITURE

©2004 American Medical Association. All rights reserved. (Reprinted) JAMA, November 24, 2004—Vol 292, No. 20 2489

at LUNDS UNIVERSITAET on May 4, 2011jama.ama-assn.orgDownloaded from

fat diet, and adjustment for change inbody composition did not alter the REEeffect. Of particular significance, differ-ences in dietary protein of the same orgreater magnitude as that used in ourstudy did not result in any differences inREE, either following weight mainte-nance or weight loss, in 5 previous ar-ticles.27-31

A second methodological issue is thatmeasurement of REE was made dur-ing ongoing weight loss. The magni-tude of observed effect could changewith weight stabilization and addi-tional research is needed to assess thispossibility. Nevertheless, the physi-ological adaptations that occur duringactive weight loss may be especially rel-evant to understanding why most obeseindividuals become noncompliant withconventional energy-restricted dietslong before a normal body weight hasbeen reached. Indeed, diet-induced dif-ferences in REE were observed after ourparticipants had lost less than half oftheir excess adiposity, and after just 1week of energy restriction in a previ-ous study.18

The difference in REE is too small toaccount for any significant change inbody composition over the short term.For example, 80 kcal/d over 10 weeks(5600 kcal) would amount to less than1 kg of body weight. Thus, our studydoes not support claims that popular di-ets can cause rapid weight loss by induc-ing major shifts in energy metabolism.

Nevertheless, the REE difference herecould amount to several pounds ofweight change per year, given this effectwould persist over the long term. Forcomparative purposes, an energy bal-ance of –80 kcal/d could be obtained bywalking approximately 1 mile/d or by de-creasing sugar-sweetened soft drink con-sumption 6 oz/d. Indeed, this differ-ence (560 kcal/wk) would explain mostof the mean difference in rate of weightloss between groups (0.09 kg/wk ! 7500kcal/kg = 675 kcal/wk).

A potentially more important ques-tion is whether the magnitude of changein REE during weight loss would pre-dict likelihood of achieving and main-taining clinically significant weight loss.Some studies32,33 but not all34 suggestan inverse relationship between REEand weight gain or regain. An indi-vidual experiencing a larger decline inREE during weight loss may feel morefatigued, cold, and hungry than an in-dividual experiencing a small decline,and these symptoms may make com-pliance with dietary energy restrictionincreasingly difficult over time.

The physiological mechanisms relat-ing dietary composition to REE duringweight loss remain speculative but mayinvolve altered availability of metabolicfuels. Blood glucose and free fatty acidsare reduced in the postabsorptive phasefollowing a high– vs low–glycemic in-dex meal, and this reduction can be suf-ficient in magnitude to trigger release of

stress hormones.12 Low circulating con-centrations of metabolic substrate mightdirectly impair energy metabolism at thecellular level, as occurs with frank hy-poglycemia.35 Alternatively, the de-crease in REE may come from neuroen-docrinological adaptations designed toconserve energy, involving thyroid hor-mone, growth hormone, sex hor-mones, or leptin (an adipocyte-derivedfactor that acts in the hypothalamus)3,5;lack of data on these hormones com-prises a limitation of our study. Inter-estingly, rodents treated with nutrient-controlled high–glycemic index dietscompared with low–glycemic indexdiets demonstrate an increase in meta-bolic efficiency analogous to that ob-served by our participants taking thelow-fat (high–glycemic index) diet.36

Epidemiological analyses have foundassociations between glycemic load andhigh triglycerides, low high-density li-poprotein cholesterol, and elevated C-reactive protein levels.37 In 1 study,38

individuals in the highest vs lowestquintile of glycemic load had double therisk of developing heart disease, aftercontrolling for potentially confound-ing factors. However, these effects havenot previously been examined in inter-ventional studies. We found that dur-ing weight loss, a diet focused on gly-cemic load reduction produced greaterimprovements in several important car-diovascular disease–related and diabe-tes mellitus–related end points than a

Table 5. Cardiovascular Disease Risk Factors Before and After 10% Weight Loss by Dietary Treatment GroupMean (SE)

P Value*

Low-Fat Diet Group (n = 17) Low–Glycemic Load Diet Group (n = 22)

Baseline Posttreatment% Change(Adjusted) Baseline Posttreatment

% Change(Adjusted)

HOMA score 1.45 (0.20) 1.10 (0.13) –15.8 (5.13) 1.50 (0.18) 0.97 (0.11) –33.9 (4.51) .01Triglycerides, mg/dL† 92.4 (9.47) 102.3 (8.11) 16.2 (5.24) 78.3 (8.40) 72.4 (7.19) –3.5 (4.63) .01HDL-C, mg/dL† 49.4 (3.61) 44.1 (2.41) –8.1 (3.49) 46.9 (3.20) 42.2 (2.14) –8.9 (3.09) .87LDL-C, mg/dL† 124.3 (9.86) 104.6 (9.73) –15.0 (4.12) 138.7 (9.75) 115.9 (8.63) –16.1 (3.65) .84C-reactive protein, mg/dL 0.19 (0.06) 0.13 (0.04) –5.1 (13.61) 0.28 (0.06) 0.10 (0.03) –47.7 (11.94) .03Systolic BP, mm Hg 107.5 (2.90) 104.6 (2.35) –3.1 (1.32) 110.4 (2.55) 102.3 (2.06) –6.4 (1.16) .07Diastolic BP, mm Hg 67.8 (2.03) 66.2 (1.80) –2.5 (1.61) 69.2 (1.78) 64.2 (1.58) –6.5 (1.42) .07Mean arterial pressure, mm Hg 94.1 (2.48) 91.7 (2.03) –3.0 (1.27) 96.6 (2.18) 89.5 (1.78) –6.5 (1.12) .04Abbreviations: BP, blood pressure; HDL-C, high-density lipoprotein cholesterol; HOMA, homeostasis model assessment; LDL-C, low-density lipoprotein.SI conversions: To convert HDL-C and LDL-C to mmol/L, multiply by 0.0259; triglycerides to mmol/L, multiply by 0.0113.*Effect of dietary treatment on % change.†For low-fat diet group, n=11; for low–glycemic load diet group, n=14.

LOW–GLYCEMIC LOAD DIET AND RESTING ENERGY EXPENDITURE

©2004 American Medical Association. All rights reserved. (Reprinted) JAMA, November 24, 2004—Vol 292, No. 20 2489


PRELIMINARYCOMMUNICATION

Effects of a Low–Glycemic Load Dieton Resting Energy Expenditure andHeart Disease Risk FactorsDuring Weight LossMark A. Pereira, PhDJanis Swain, MS, RDAllison B. Goldfine, MDNader Rifai, PhDDavid S. Ludwig, MD, PhD

THE POOR LONG-TERM EFFICACYof conventional obesity treat-ment has promoted the notionof a body weight set point,1,2

more recently termed settling point. Ac-cording to this concept, deviations inbody weight from baseline elicit physi-ological adaptations that antagonize fur-ther weight change. During energy re-striction, humans and experimentalanimals have increased hunger, de-creased thyroid hormone levels, anddown-regulation of reproductive andgrowth functions,3-5 changes that in-creaseenergy intakeand lowerenergyex-penditure. To examine this phenom-enon, Leibel et al6 underfed or overfedparticipants who were lean or obese toobtain an approximate 10% decrease orincrease in body weight from baseline.Resting energy expenditure (REE) andtotal energy expenditure relative to fat-free body mass declined following weightreduction, whereas total energy expen-diture increased following weight gain.

A decline in REE and associated neu-roendocrine changes have been consis-tently reported during active weight loss,althoughcontroversyexists as towhetherthese adaptations are permanent6-8

or transient.9,10 In any event, defendedbody weight level is evidently not deter-

mined by endogenous mechanisms ex-clusively, as demonstrated by the in-creasing mean body mass index (BMI,calculated as weight in kilograms di-vided by the square of height in meters)among genetically stable populations ob-served in recent years.11 Thus, bodyweight settling point may best be con-ceptualized as representing the inte-grated influences of numerous genetic,behavioral, and environmental factors.1

Previously, the novel dietary factorglycemic load has been proposed to play

a role in body weight regulation basedon experimental and theoreticalgrounds.12,13 Glycemic load (glycemic in-

Author Affiliations: Department of Medicine (DrsPereira and Ludwig) and Department of LaboratoryMedicine (Dr Rifai), Children’s Hospital; General Clini-cal Research Center, Brigham and Women’s Hospital(Ms Swain); and Joslin Diabetes Center (Dr Gold-fine), Boston, Mass. Dr Pereira is now with the Divi-sion of Epidemiology & Community Health, School ofPublic Health, University of Minnesota, Minneapolis.Corresponding Author: David S. Ludwig, MD, PhD,Department of Medicine, Children’s Hospital, 300Longwood Ave, Boston, MA 02115 ([email protected]).

Context Weight loss elicits physiological adaptations relating to energy intake andexpenditure that antagonize ongoing weight loss.

Objective To test whether dietary composition affects the physiological adapta-tions to weight loss, as assessed by resting energy expenditure.

Design, Study, and Participants A randomized parallel-design study of 39 over-weight or obese young adults aged 18 to 40 years who received an energy-restricteddiet, either low–glycemic load or low-fat. Participants were studied in the General Clini-cal Research Centers of the Brigham and Women’s Hospital and the Children’s Hos-pital, Boston, Mass, before and after 10% weight loss. The study was conducted fromJanuary 4, 2001, to May 6, 2003.

Main Outcome Measures Resting energy expenditure measured in the fasting stateby indirect calorimetry, body composition by dual-energy x-ray absorptiometry, car-diovascular disease risk factors, and self-reported hunger.

Results Resting energy expenditure decreased less with the low–glycemic load dietthan with the low-fat diet, expressed in absolute terms (mean [SE], 96 [24] vs 176[27] kcal/d; P=.04) or as a proportion (5.9% [1.5%] vs 10.6% [1.7%]; P=.05). Par-ticipants receiving the low–glycemic load diet reported less hunger than those receiv-ing the low-fat diet (P=.04). Insulin resistance (P=.01), serum triglycerides (P=.01),C-reactive protein (P=.03), and blood pressure (P=.07 for both systolic and diastolic)improved more with the low–glycemic load diet. Changes in body composition (fatand lean mass) in both groups were very similar (P=.85 and P=.45, respectively).

Conclusions Changes in dietary composition within prevailing norms can affect physi-ological adaptations that defend body weight. Reduction in glycemic load may aid inthe prevention or treatment of obesity, cardiovascular disease, and diabetes mellitus.JAMA. 2004;292:2482-2490 www.jama.com

2482 JAMA, November 24, 2004—Vol 292, No. 20 (Reprinted) ©2004 American Medical Association. All rights reserved.


109

0

0,5

1

1,5

2

2,5

1 2 3 4 5

CRP LDL

QUINTIS DE RISCO RELATIVO PARA TODOS OS ACIDENTES CARDIOVASCULARES

Ridker PM et al. N Engl J Med 2002;347:1557-65.


Intake of carbohydrates compared with intake of saturated fatty acidsand risk of myocardial infarction: importance of the glycemic index1–3

Marianne U Jakobsen, Claus Dethlefsen, Albert M Joensen, Jakob Stegger, Anne Tjønneland, Erik B Schmidt,and Kim Overvad

ABSTRACTBackground: Studies have suggested that replacing saturated fattyacids (SFAs) with carbohydrates is modestly associated with a high-er risk of ischemic heart disease, whereas replacing SFAs withpolyunsaturated fatty acids is associated with a lower risk of ische-mic heart disease. The effect of carbohydrates, however, may de-pend on the type consumed.Objectives: By using substitution models, we aimed to investigatethe risk of myocardial infarction (MI) associated with a higher en-ergy intake from carbohydrates and a concomitant lower energyintake from SFAs. Carbohydrates with different glycemic index(GI) values were also investigated.Design: Our prospective cohort study included 53,644 women andmen free of MI at baseline.Results: During a median of 12 y of follow-up, 1943 incident MIcases occurred. There was a nonsignificant inverse association be-tween substitution of carbohydrates with low-GI values for SFAsand risk of MI [hazard ratio (HR) for MI per 5% increment ofenergy intake from carbohydrates: 0.88; 95% CI: 0.72, 1.07). Incontrast, there was a statistically significant positive associationbetween substitution of carbohydrates with high-GI values for SFAsand risk of MI (HR: 1.33; 95% CI: 1.08, 1.64). There was noassociation for carbohydrates with medium-GI values (HR: 0.98;95% CI: 0.80, 1.21). No effect modification by sex was observed.Conclusion: This study suggests that replacing SFAs with carbohy-drates with low-GI values is associated with a lower risk of MI,whereas replacing SFAs with carbohydrates with high-GI values isassociated with a higher risk of MI. Am J Clin Nutr 2010;91:1764–8.

INTRODUCTION

Epidemiologic prospective cohort studies have suggested thatreplacing saturated fatty acids (SFAs) with carbohydrates ismodestly associated with a higher risk of ischemic heart disease(IHD), whereas replacing SFAs with polyunsaturated fatty acidsis associated with a lower risk of IHD (1). The effect of car-bohydrates, however, may depend on the type consumed.

Epidemiologic prospective cohort studies have shown a posi-tive association between dietary glycemic index (GI) and risk ofIHD (2). The dietary GI is an indicator of the average quality of thecarbohydrates consumed in terms of glycemic response. The GI,which was conceived to provide a classification of carbohydrate-containing foods on the basis of their ability to raise blood glu-cose, was introduced by Jenkins et al (3) in 1981. Blood glucose

concentration is tightly regulated by homeostatic regulatorysystems, but the rapid absorption of carbohydrates after con-sumption of a high-GI meal challenges these homeostaticmechanisms (4). A high-GI meal results in a high blood glucoseconcentration and a high insulin-to-glucagon ratio, followed byhypoglycemia, counterregulatory hormone secretion, and ele-vated plasma free fatty acid concentration (4). These events mayaffect the risk of IHD through promoting dyslipidemia, in-flammation, and endothelial dysfunction (4).

The aim of this study was to investigate the risk of myocardialinfarction (MI) with a higher energy intake from carbohydratesand a concomitant lower energy intake from SFAs. Carbohydrateswith different GI values were investigated. Furthermore, potentialeffect modification by sex was investigated because of differencesin the underlying biology such as hormonal differences.


Study population

Between December 1993 and May 1997, 160,725 women andmen were invited by mail to participate in the Danish prospectivecohort study Diet, Cancer, and Health. The criteria for invitationwere as follows: age between 50 and 64 y, born in Denmark, andno previous cancer diagnosis registered in the Danish CancerRegistry. All persons fulfilling these criteria and living in thegreater Copenhagen or Aarhus areas were invited. With the in-

1 From the Department of Clinical Epidemiology Aarhus University Hos-pital, Aalborg, Denmark (MUJ); the Department of Cardiology, Center forCardiovascular Research, Aalborg Hospital, Aarhus University Hospital,Aalborg, Denmark (MUJ, CD, AMJ, JS, EBS, and KO); the Danish CancerSociety, Institute of Cancer Epidemiology, Copenhagen, Denmark (AT); andthe Department of Epidemiology, School of Public Health, Aarhus Univer-sity, Aarhus, Denmark (KO).

2 This work is part of the project Hepatic and Adipose Tissue and Func-tions in the Metabolic Syndrome (HEPADIP; www.hepadip.org), which issupported by the European Commission as an Integrated Project under the6th Framework Programme (contract LSHM-CT-2005-018734), and part ofthe research program of the Danish Obesity Research Centre (DanORC;www.danorc.dk), which is supported by the Danish Council for StrategicResearch (contract 2101-06-0005).

3 Address correspondence to MU Jakobsen, Department of Clinical Epi-demiology, Aarhus University Hospital, Sdr. Skovvej 15, DK-9000 Aalborg,Denmark. E-mail: [email protected].

Received December 17, 2009. Accepted for publication March 16, 2010.First published online April 7, 2010; doi: 10.3945/ajcn.2009.29099.

1764 Am J Clin Nutr 2010;91:1764–8. Printed in USA. ! 2010 American Society for Nutrition

at Lund University Libraries on N

ovember 5, 2010

ww

w.ajcn.org

Dow

nloaded from


Intake of carbohydrates compared with intake of saturated fatty acidsand risk of myocardial infarction: importance of the glycemic index1–3

Marianne U Jakobsen, Claus Dethlefsen, Albert M Joensen, Jakob Stegger, Anne Tjønneland, Erik B Schmidt,and Kim Overvad

ABSTRACTBackground: Studies have suggested that replacing saturated fattyacids (SFAs) with carbohydrates is modestly associated with a high-er risk of ischemic heart disease, whereas replacing SFAs withpolyunsaturated fatty acids is associated with a lower risk of ische-mic heart disease. The effect of carbohydrates, however, may de-pend on the type consumed.Objectives: By using substitution models, we aimed to investigatethe risk of myocardial infarction (MI) associated with a higher en-ergy intake from carbohydrates and a concomitant lower energyintake from SFAs. Carbohydrates with different glycemic index(GI) values were also investigated.Design: Our prospective cohort study included 53,644 women andmen free of MI at baseline.Results: During a median of 12 y of follow-up, 1943 incident MIcases occurred. There was a nonsignificant inverse association be-tween substitution of carbohydrates with low-GI values for SFAsand risk of MI [hazard ratio (HR) for MI per 5% increment ofenergy intake from carbohydrates: 0.88; 95% CI: 0.72, 1.07). Incontrast, there was a statistically significant positive associationbetween substitution of carbohydrates with high-GI values for SFAsand risk of MI (HR: 1.33; 95% CI: 1.08, 1.64). There was noassociation for carbohydrates with medium-GI values (HR: 0.98;95% CI: 0.80, 1.21). No effect modification by sex was observed.Conclusion: This study suggests that replacing SFAs with carbohy-drates with low-GI values is associated with a lower risk of MI,whereas replacing SFAs with carbohydrates with high-GI values isassociated with a higher risk of MI. Am J Clin Nutr 2010;91:1764–8.

INTRODUCTION

Epidemiologic prospective cohort studies have suggested thatreplacing saturated fatty acids (SFAs) with carbohydrates ismodestly associated with a higher risk of ischemic heart disease(IHD), whereas replacing SFAs with polyunsaturated fatty acidsis associated with a lower risk of IHD (1). The effect of car-bohydrates, however, may depend on the type consumed.

Epidemiologic prospective cohort studies have shown a posi-tive association between dietary glycemic index (GI) and risk ofIHD (2). The dietary GI is an indicator of the average quality of thecarbohydrates consumed in terms of glycemic response. The GI,which was conceived to provide a classification of carbohydrate-containing foods on the basis of their ability to raise blood glu-cose, was introduced by Jenkins et al (3) in 1981. Blood glucose

concentration is tightly regulated by homeostatic regulatorysystems, but the rapid absorption of carbohydrates after con-sumption of a high-GI meal challenges these homeostaticmechanisms (4). A high-GI meal results in a high blood glucoseconcentration and a high insulin-to-glucagon ratio, followed byhypoglycemia, counterregulatory hormone secretion, and ele-vated plasma free fatty acid concentration (4). These events mayaffect the risk of IHD through promoting dyslipidemia, in-flammation, and endothelial dysfunction (4).

The aim of this study was to investigate the risk of myocardialinfarction (MI) with a higher energy intake from carbohydratesand a concomitant lower energy intake from SFAs. Carbohydrateswith different GI values were investigated. Furthermore, potentialeffect modification by sex was investigated because of differencesin the underlying biology such as hormonal differences.


Study population

Between December 1993 and May 1997, 160,725 women andmen were invited by mail to participate in the Danish prospectivecohort study Diet, Cancer, and Health. The criteria for invitationwere as follows: age between 50 and 64 y, born in Denmark, andno previous cancer diagnosis registered in the Danish CancerRegistry. All persons fulfilling these criteria and living in thegreater Copenhagen or Aarhus areas were invited. With the in-

1 From the Department of Clinical Epidemiology Aarhus University Hos-pital, Aalborg, Denmark (MUJ); the Department of Cardiology, Center forCardiovascular Research, Aalborg Hospital, Aarhus University Hospital,Aalborg, Denmark (MUJ, CD, AMJ, JS, EBS, and KO); the Danish CancerSociety, Institute of Cancer Epidemiology, Copenhagen, Denmark (AT); andthe Department of Epidemiology, School of Public Health, Aarhus Univer-sity, Aarhus, Denmark (KO).

2 This work is part of the project Hepatic and Adipose Tissue and Func-tions in the Metabolic Syndrome (HEPADIP; www.hepadip.org), which issupported by the European Commission as an Integrated Project under the6th Framework Programme (contract LSHM-CT-2005-018734), and part ofthe research program of the Danish Obesity Research Centre (DanORC;www.danorc.dk), which is supported by the Danish Council for StrategicResearch (contract 2101-06-0005).

3 Address correspondence to MU Jakobsen, Department of Clinical Epi-demiology, Aarhus University Hospital, Sdr. Skovvej 15, DK-9000 Aalborg,Denmark. E-mail: [email protected].

Received December 17, 2009. Accepted for publication March 16, 2010.First published online April 7, 2010; doi: 10.3945/ajcn.2009.29099.

1764 Am J Clin Nutr 2010;91:1764–8. Printed in USA. ! 2010 American Society for Nutrition


ovember 5, 2010

ww

w.ajcn.org

Dow

nloaded from

first tertile of dietary GI: 0.88; 95% CI: 0.72, 1.07) and a sta-tistically significant positive association between substitution ofcarbohydrates with high-GI values for SFAs and risk of MI (HRin the third tertile of dietary GI: 1.33; 95% CI: 1.08, 1.64) (Table3). There was no association for carbohydrates with medium-GIvalues (HR in the second tertile of dietary GI: 0.98; 95% CI:0.80, 1.21) (Table 3). As assessed from the 95% CIs, themeasures of associations for extreme tertiles of GI were statis-tically significantly different. The P value for effect modificationby tertiles of dietary GI was 0.06 in women, 0.29 in men, and0.16 in all participants. The P value for effect modification bysex was 0.86.

DISCUSSION

The findings from this study suggest that the effect of sub-stitution of carbohydrates for SFAs varies depending on the type ofcarbohydrates. There was a nonsignificant inverse association

between substitution of carbohydrates with low-GI values for SFAsand risk of MI but a significant positive association betweensubstitution of carbohydrates with high-GI values for SFAs and riskof MI. In this study, dietary GI was used as an indicator of theaverage quality of carbohydrates consumed, but other classi-fications of carbohydrates may also be relevant, such as the extentof processing, which also reflects the intake of dietary fiber (16, 17).

Selection bias is unlikely to have affected the results. However,if censoring due to death from other causes is associated withintake of carbohydrates and risk of MI, then the true associationsbetween intake of carbohydrates and risk of MI may have beenunderestimated. Random measurement error cannot be excludedfrom having affected the results. A potential source of randommeasurement error arises from dietary self-reporting methods.Generally, random measurement error leads to underestimationof the true risk and to loss of statistical power. However, dietaryintake was determined by using a FFQ, which may reflect thehabitual eating pattern. Information bias is unlikely to have af-fected the results because cases were identified by record linkagewith central Danish registries and diagnoses were establishedindependently of the FFQs of the participants. We included car-bohydrates, proteins, monounsaturated fatty acids, and poly-unsaturated fatty acids expressed as percentages of the total energyintake and the total energy intake in the models because of po-tential confounding and extraneous variation. This also allowed usto estimate the difference in the risk for a higher energy intake fromcarbohydrates and a concomitant lower energy intake from SFAs.Relevant control for established risk factors for IHD did not changethe measures of associations, and thus residual confounding seemsunlikely. However, confounding from other IHD risk factors nottaken into account remains a possible explanation for the observedassociations.

Only 2 epidemiologic studies have investigated the sub-stitution of carbohydrates for SFAs (1, 18). In the prospectivecohort study by Hu et al (18), substitution of carbohydrates forSFAs was nonsignificantly associated with a lower risk of IHD,whereas in the prospective cohort study by Jakobsen et al (1), inwhich data from 11 American and European cohort studies werepooled, substitution of carbohydrates for SFAs was modestly

TABLE 2Hazard ratios for myocardial infarction per 5% increment of energy intakefrom carbohydrates and a concomitant lower energy intake from saturatedfatty acids1

All participants Women Men

Model 12 1.04 (0.93, 1.17) 1.09 (0.88, 1.36) 1.03 (0.90, 1.18)Model 23 1.04 (0.92, 1.17) 1.02 (0.82, 1.28) 1.05 (0.92, 1.21)

1 All values are hazard ratios; 95% CIs in parentheses. n = 53,644 forall participants, n = 28,495 for women, and n = 25,149 for men.

2 Model 1 included intake of glycemic carbohydrates, proteins, mono-unsaturated fatty acids, and polyunsaturated fatty acids expressed as percen-tages of total energy intake, total energy intake (kcal/d), an indicator variablefor alcohol consumption (0 and .0 g/d), and alcohol consumption (g/d).Hazard ratios with 95% CIs for the incidence of myocardial infarction werecalculated by using Cox proportional hazards regression with age as the timemetric. In analyses among all participants, sex was entered into the model.

3 Model 2 included variables in model 1 and BMI (in kg/m2; ,25, 25–29, and !30), education (,8, 8–10, and .10 y), smoking status (never,former, and currently smoking 1–14, 15–24, or !25 g tobacco/d), physicalactivity (,3.5 and !3.5 h/wk), and history of hypertension (yes, no, and donot know).

TABLE 3Hazard ratios (HRs) for myocardial infarction per 5% increment of energy intake from carbohydrates with low–glycemic index (low-GI), medium-GI, orhigh-GI values and a concomitant lower energy intake from saturated fatty acids1

All participants Women Men

Tertiles ofdietary GI2

Median dietary GI(80% central range) HR (95% CI)



Carbohydrates with low-GIvalues (first tertile)

82 (77, 85) 0.88 (0.72, 1.07) 80 (75, 82) 1.17 (0.80, 1.71) 84 (79, 86) 0.83 (0.65, 1.04)

Carbohydrates with medium-GIvalues (second tertile)

88 (86, 90) 0.98 (0.80, 1.21) 85 (84, 87) 0.80 (0.54, 1.18) 89 (87, 91) 1.08 (0.84, 1.38)

Carbohydrates with high-GIvalues (third tertile)

93 (91, 98) 1.33 (1.08, 1.64) 91 (88, 96) 1.10 (0.75, 1.63) 94 (92, 98) 1.34 (1.04, 1.71)

1 All models included intake of glycemic carbohydrates, proteins, monounsaturated fatty acids, and polyunsaturated fatty acids expressed as percentagesof total energy intake, total energy intake (kcal/d), an indicator variable for alcohol consumption (0 and .0 g/d), alcohol consumption (g/d), BMI (in kg/m2;,25, 25–29, and !30), education (,8, 8–10, and .10 y), smoking status (never, former, and currently smoking 1–14, 15–24, or !25 g tobacco/d), physicalactivity (,3.5 and !3.5 h/wk), and history of hypertension (yes, no, and do not know). HRs with 95% CIs for the incidence of myocardial infarction werecalculated by using Cox proportional hazards regression with age as the time metric. In analyses among all participants, sex was entered into the model.

2 Tertiles of dietary GI were based on the distribution of dietary GI among cases. n = 22,144, 17,000, and 14,400 for all participants in the first, second,and third tertiles of dietary GI, respectively; n = 9594, 10,202, and 8699 for women in the first, second, and third tertiles of dietary GI, respectively; and n =8941, 8127, and 8081 for men in the first, second, and third tertiles of dietary GI, respectively.

CARBOHYDRATES AND RISK OF MYOCARDIAL INFARCTION 1767


ovember 5, 2010

ww

w.ajcn.org

Dow

nloaded from

Glycemic index, glycemic load, and chronic disease risk—a meta-analysis of observational studies1,2

Alan W Barclay, Peter Petocz, Joanna McMillan-Price, Victoria M Flood, Tania Prvan,Paul Mitchell, and Jennie C Brand-Miller

ABSTRACTBackground: Inconsistent findings from observational studies haveprolonged the controversy over the effects of dietary glycemic index(GI) and glycemic load (GL) on the risk of certain chronic diseases.Objective: The objective was to evaluate the association betweenGI, GL, and chronic disease risk with the use of meta-analysis tech-niques.Design: A systematic review of published reports identified a totalof 37 prospective cohort studies of GI and GL and chronic diseaserisk. Studies were stratified further according to the validity of thetools used to assess dietary intake. Rate ratios (RRs) were estimatedin a Cox proportional hazards model and combined by using arandom-effects model.Results: From 4 to 20 y of follow-up across studies, a total of 40 129incident cases were identified. For the comparison between the high-est and lowest quantiles of GI and GL, significant positive associa-tions were found in fully adjusted models of validated studies fortype 2 diabetes (GI RR ! 1.40, 95% CI: 1.23, 1.59; GL RR ! 1.27,95% CI: 1.12, 1.45), coronary heart disease (GI RR ! 1.25, 95% CI:1.00, 1.56), gallbladder disease (GI RR ! 1.26, 95% CI: 1.13, 1.40;GL RR ! 1.41, 95% CI: 1.25, 1.60), breast cancer (GI RR ! 1.08,95% CI: 1.02, 1.16), and all diseases combined (GI RR ! 1.14, 95%CI: 1.09, 1.19; GL RR ! 1.09, 95% CI: 1.04, 1.15).Conclusions: Low-GI and/or low-GL diets are independently asso-ciated with a reduced risk of certain chronic diseases. In diabetes andheart disease, the protection is comparable with that seen for wholegrain and high fiber intakes. The findings support the hypothesis thathigher postprandial glycemia is a universal mechanism for diseaseprogression. Am J Clin Nutr 2008;87:627–37.

KEY WORDS Glycemic index, glycemic load, dietary carbo-hydrates, epidemiology

INTRODUCTION

Worldwide, chronic diseases such as diabetes, cardiovasculardisease, stroke, and cancer contribute to "60% of all deaths, andthe proportion is predicted to increase to 75% by the year 2020 (1,2). Habitual diet is the major modifiable risk factor, and theidentification of simple, cost-effective strategies for preventionand management is a matter of urgency.

Although changes in the quantity and quality of fat have re-ceived considerable attention, the role of carbohydrates is lessclear (2). Increases in refined sugar intake have been accompa-nied by more subtle changes in starchy foods, eg, processed

cereal products have replaced more traditionally processedgrains. Because carbohydrate is the main dietary componentaffecting insulin secretion and postprandial glycemia (3), it isimplicated in the etiology of many chronic diseases. Both theamount and type of carbohydrate consumed have an effect onboth insulin secretion and postprandial glycemia, with differ-ences not explained by glucose chain length (4). In 1981, theconcept of the glycemic index (GI) was introduced by Jenkins etal (5) to quantify the glycemic response to carbohydrates indifferent foods. Glycemic load (GL), the mathematical productof the GI of a food and its carbohydrate content, has been pro-posed as a global indicator of the glucose response and insulindemand induced by a serving of food (6).

The results of studies that investigated the association betweenoverall dietary GI, GL, and disease risk have been inconsistent.With respect to diabetes, a positive association was documentedin 6 large cohort studies (6–11), but no association was seen in 2others (12, 13). In cardiovascular disease, 2 studies reported apositive association (14, 15), whereas 1 found no relation (16).Most of the studies that have investigated cancer risk have re-ported no associations (11, 17–29), but there are notable excep-tions (30–37). Two studies that investigated the risk of gallblad-der disease showed positive associations (38, 39). Finally, 2studies (40, 41) reported an association with eye disease, whereasa third found no association (42).

Of concern, 5 (13%) (22, 25, 27, 31, 33) of the 37 prospectivestudies that investigated the relation between dietary carbohy-drates, GI, GL, and chronic disease risk did not validate carbo-hydrate intake, and an additional 5 (13%) (12, 13, 20, 36, 37)showed correlation coefficients for total carbohydrate of #0.5.Another 2 (5%) studies (29, 32) appear to have been validated,but the validation study has not been published, and 2 others (5%)

1 From the Human Nutrition Unit, University of Sydney, Sydney, Austra-lia (AWB, JM-P, VMF, and JCB-M); the Department of Statistics, Macqua-rie University, Sydney, Australia (PP and TP); the Department of Ophthal-mology, Centre for Vision Research, Westmead Millennium Institute,Westmead Hospital, University of Sydney, Sydney, Australia (VMF andPM); and the NSW Centre for Public Health Nutrition, Human NutritionUnit, University of Sydney, Sydney, Australia (VMF).

2 Address reprint requests and correspondence to JC Brand-Miller, HumanNutrition Unit, University of Sydney, Sydney, NSW, Australia 2006. E-mail:[email protected].

Received April 24, 2007.Accepted for publication September 24, 2007.

627Am J Clin Nutr 2008;87:627–37. Printed in USA. © 2008 American Society for Nutrition

by Custodio C

esar on Novem

ber 18, 2008 w

ww

.ajcn.orgD

ownloaded from





INTRODUCTION










by Custodio Cesar on November 18, 2008






INTRODUCTION










by Custodio C

esar on Novem

ber 18, 2008 w

ww

.ajcn.orgD

ownloaded from

foods in a nutrient database is to some extent subjective andmay be unreliable when extrapolating from one country toanother. It is likely in any case that any misclassification of GIor GL would lead to a bias toward the null hypothesis anddiminish the observed effect size. Despite comprehensivemeasurement and adjustment strategies, the uncontrolled orresidual confounding in observational studies of dietary in-take is always a concern. Healthy lifestyle effects associatedwith dietary intake cannot be completely adjusted for in ob-servational studies. Therefore, a meta-analysis of interventionstudies looking at “hard” clinical endpoints, not chronic-disease risk factors, may be warranted, when sufficient datahave accumulated.

Our findings support the hypothesis that postprandial hyper-glycemia, in individuals without diabetes, contributes to chronicdisease. Higher glucose concentrations are thought to play adirect pathogenic role in the disease process. The DECODEstudy, a meta-analysis of 13 studies involving 25 000 individu-als, found an almost 2-fold increased risk of all-cause mortalityin individuals with an elevated 2-h postchallenge blood glucose

(53). Similarly, a meta-analysis of 38 studies involving 194 000individuals found a progressive relation between hyperglycemiaand cardiovascular disease risk (54). Cancer risk is also elevatedin those with preexisting hyperglycemia. Larsson et al (55) re-ported a 30% increase in risk of colorectal cancer in a meta-analyses of persons with type 2 diabetes; similarly, Huxley et al(56) found an 82% increase in risk of pancreatic cancer.

There are plausible mechanisms linking the development ofcertain chronic diseases with high-GI diets. Specifically, 2 majorpathways have been proposed to explain the association withtype 2 diabetes risk (57). First, the same amount of carbohydratefrom high-GI foods produces higher blood glucose concentra-tions and a greater demand for insulin. The chronically increasedinsulin demand may eventually result in pancreatic ! cell failure,and, as a consequence, impaired glucose tolerance. Second, thereis evidence that high-GI diets may directly increase insulin re-sistance through their effect on glycemia, free fatty acids, andcounter-regulatory hormone secretion. High glucose and insulinconcentrations are associated with increased risk profiles forcardiovascular disease, including decreased concentrations of

TABLE 3 (Continued)

Chronic disease Glycemic index rate ratio1 P Glycemic load rate ratio1 P

Eye diseaseSchaumberg et al (41) 20046 1.11 (0.99, 1.25) 0.079 1.01 (0.83, 1.23) 0.921Schaumberg et al (41) 20047 0.95 (0.81, 1.11) 0.523 0.86 (0.65, 1.13) 0.285Chiu et al (42) 20058 1.09 (0.61, 1.94) 0.770 — —Chiu et al (42) 20059 1.15 (0.63, 2.10) 0.649 — —Chiu et al (40) 2006 2.71 (1.24, 5.93) 0.013 — —Overall 1.10 (0.91, 1.31) 0.323 0.96 (0.82, 1.12) 0.590All diseases (6–42)Overall 1.13 (1.08, 1.19) !0.0001 1.10 (1.06, 1.15) !0.0001

1 Final fully adjusted models only.2 White Americans.3 Black Americans.4 BMI ! 25 kg/m2.5 BMI " 25 kg/m2.6 Women.7 Men.8 Cortical opacity.9 Nuclear opacity.

TABLE 4Rate ratios (and 95% CIs) for the comparison of the highest with the lowest quantile for developing chronic disease because of increasing glycemic indexor glycemic load in 27 prospective cohort studies meeting a priori exclusion criteria (correlation between food-frequency questionnaire and weighed foodrecords/24-h dietary recall " 0.5 in representative subgroups)

Chronic diseaseGlycemic index rate

ratio1 PGlycemic load rate

ratio1 P

Type 2 diabetes (6–11) 1.40 (1.23, 1.59) !0.0001 1.27 (1.12, 1.45) !0.0001Heart disease (14, 16) 1.25 (1.00, 1.56) 0.050 1.57 (0.87, 2.84) 0.140Stroke (15) 1.02 (0.86, 1.21) 0.805 1.28 (0.83, 1.98) 0.270Breast cancer (17–19, 21, 30) 1.09 (1.02, 1.16) 0.015 0.99 (0.92, 1.06) 0.797Colorectal cancer (23, 29, 34, 35) 1.11 (0.99, 1.24) 0.059 1.11 (0.88, 1.40) 0.385Pancreatic cancer (11, 24) 0.98 (0.78, 1.25) 0.896 0.96 (0.75, 1.23) 0.733Endometrial cancer (26, 32) 1.13 (0.80, 1.60) 0.489 1.72 (0.75, 3.95) 0.204Gastric cancer (28) 0.77 (0.46, 1.29) 0.320 0.76 (0.46, 1.25) 0.282Gallbladder disease (38, 39) 1.26 (1.13, 1.40) !0.0001 1.41 (1.25, 1.60) !0.0001Eye disease (40–42) 1.10 (0.91, 1.31) 0.323 0.96 (0.82, 1.12) 0.590All diseases (6–11, 14–19, 21, 23, 24, 26, 28–30, 32, 34, 35,

38–42) 1.14 (1.09, 1.19) !0.0001 1.09 (1.04, 1.15) !0.00011 Final fully adjusted models only.

GLYCEMIC INDEX AND LOAD AND CHRONIC DISEASE RISK 633

by Custodio C

esar on Novem

ber 18, 2008 w

ww

.ajcn.orgD

ownloaded from

e-Book

MITO 2





MITO 2





CEREAIS INTEGRAIS


Effect of a Low–Glycemic Indexor a High–Cereal Fiber Diet on Type 2 DiabetesA Randomized TrialDavid J. A. Jenkins, MDCyril W. C. Kendall, PhDGail McKeown-Eyssen, PhDRobert G. Josse, MB, BSJay Silverberg, MDGillian L. Booth, MDEdward Vidgen, BScAndrea R. Josse, MScTri H. Nguyen, MScSorcha Corrigan, BScMonica S. Banach, BScSophie Ares, MA, RD, CDESandy Mitchell, BASc, RDAzadeh Emam, MScLivia S. A. Augustin, MScTina L. Parker, BASc, RDLawrence A. Leiter, MD

THE NEED FOR IMPLEMENTA-tion of effective dietary strate-gies in diabetes prevention andmanagement has been empha-

sized by the success of diet and life-style changes in preventing diabetes inhigh-risk patients.1 There is also con-cern that use of antihyperglycemicmedications to improve glycemic con-trol in type 2 diabetes may not alwayssignificantly improve cardiovascularoutcomes.2-7

One dietary strategy aimed at im-proving both diabetes control and car-diovascular risk factors is the use oflow–glycemic index diets.8-10 These dietshave been reported to benefit the

control of diabetes11; increase high-density lipoprotein cholesterol(HDL-C)12,13; lower serum triglycer-ide, plasminogen activator inhibitor 1,and high-sensitivity C-reactive pro-tein (CRP) concentrations14-16; and re-duce diabetes incidence8,9 and overallcardiovascular events.10 Use of the !-

glucosidase carbohydrate absorption in-hibitor acarbose, which effectivelycreates a low–glycemic index diet by

Author Affiliations are listed at the end of this article.Corresponding Author: David J. A. Jenkins, MD, De-partment of Nutritional Sciences, Faculty of Medi-cine, University of Toronto, 150 College St, Toronto,ON, M5S 3E2, Canada ([email protected]).

Context Clinical trials using antihyperglycemic medications to improve glycemic con-trol have not demonstrated the anticipated cardiovascular benefits. Low–glycemic in-dex diets may improve both glycemic control and cardiovascular risk factors for pa-tients with type 2 diabetes but debate over their effectiveness continues due to triallimitations.

Objective To test the effects of low–glycemic index diets on glycemic control andcardiovascular risk factors in patients with type 2 diabetes.

Design, Setting, and Participants A randomized, parallel study design at a Ca-nadian university hospital research center of 210 participants with type 2 diabetes treatedwith antihyperglycemic medications who were recruited by newspaper advertisementand randomly assigned to receive 1 of 2 diet treatments each for 6 months betweenSeptember 16, 2004, and May 22, 2007.

Intervention High–cereal fiber or low–glycemic index dietary advice.

Main Outcome Measures Absolute change in glycated hemoglobin A1c (HbA1c), withfasting blood glucose and cardiovascular disease risk factors as secondary measures.

Results In the intention-to-treat analysis, HbA1c decreased by !0.18% absolute HbA1c

units (95% confidence interval [CI], !0.29% to !0.07%) in the high–cereal fiber dietcompared with !0.50% absolute HbA1c units (95% CI, !0.61% to !0.39%) in thelow–glycemic index diet (P" .001). There was also an increase of high-density lipo-protein cholesterol in the low–glycemic index diet by 1.7 mg/dL (95% CI, 0.8-2.6 mg/dL) compared with a decrease of high-density lipoprotein cholesterol by !0.2 mg/dL(95% CI, !0.9 to 0.5 mg/dL) in the high–cereal fiber diet (P=.005). The reduction indietary glycemic index related positively to the reduction in HbA1c concentration (r=0.35,P" .001) and negatively to the increase in high-density lipoprotein cholesterol (r=!0.19,P=.009).

Conclusion In patients with type 2 diabetes, 6-month treatment with a low–glycemic index diet resulted in moderately lower HbA1c levels compared with a high–cereal fiber diet.

Trial Registration clinicaltrials.gov identifier: NCT00438698JAMA. 2008;300(23):2742-2753 www.jama.com

2742 JAMA, December 17, 2008—Vol 300, No. 23 (Reprinted) ©2008 American Medical Association. All rights reserved.


studies,41 a substantial number of par-ticipants (11%, n=23) had antihyper-glycemic medication changes during thestudy. Only 6 participants had clear evi-dence of repeated hypoglycemic epi-sodes or low blood glucose levels, butall of these occurred in the low–glycemic index diet group. Further-more, retention of participants withmedication changes in the analysis didnot result in loss of significance in es-tablished treatment differences. Di-etary fiber intakes were not com-pletely balanced between treatmentswith an approximately 4.6 g/d higherfiber intake in the low–glycemic in-dex diet than in the high–cereal fiberdiet at week 24. Viscous fibers or diets

high in fiber from a variety of sourceshave been shown to improve blood lip-ids and markers of glycemic con-trol.24,42 However, with the exceptionof oat and barley fiber, cereal fibers arelargely without metabolic effect. In-creasing dietary intake of wheat fiber,even by as much as 20 g/d, has beenshown not to influence HbA1c or otherbiomarkers of chronic disease in pa-tients with type 2 diabetes.43 In addi-tion, controlling for fiber intake as a co-variate in the analysis of covarianceanalysis did not alter the significanceof the results. Similarly, controlling forfiber in the partial regression analysisdid not alter the significance of the as-sociation of the change in glycemic in-

dex with change in HbA1c. A further po-tential weakness was that our study wasa single-site study, which may be seenas a limitation to its generalizability.

Study strengths include the indepen-dence of the effect of glycemic index onHbA1c from the fiber or carbohydrate in-take and the similarity of the observedHbA1c effect with the magnitude of re-duction in glycated proteins observed inameta-analysis.11 Another strengthofourstudy was the comparison of the low–glycemic index diet with a high–cerealfiber diet representing another treat-ment rather than simply a control.Increased cereal fiber intakes havebeen associated with reduced incidenceof diabetes and CHD in the longer

Figure 3. Mean Study Measurements in Participants With Type 2 Diabetes Completing Either a High–Cereal Fiber Diet or a Low–GlycemicIndex Diet

High–cereal fiber diet (n = 75) Low–glycemic index diet (n = 80)

90

82

84

86

88

80

Time, wk

kg

Body weight

0 4 8 12 16 20 24

7.30

6.50

6.70

6.90

7.10

6.30

Time, wk

%

HbA1c

0 4 8 12 16 20 24

146

122

130

138

114

Time, wk

mg/

dL

Fasting glucose

0 4 8 12 16 20 24

46.0

40.0

42.0

44.0

38.0

Time, wk

mg/

dL

HDL-C

0 4 8 12 16 20 24

140

110

120

130

100

Time, wk

mg/

dL

Triglycerides

0 4 8 12 16 20 24

4.5

3.9

4.1

4.3

3.7

Time, wk

Rat

io

Total cholesterol : HDL-C

0 4 8 12 16 20 24

2.7

2.3

2.5

2.1

Time, wk

Rat

io

LDL-C : HDL-C

0 4 8 12 16 20 24

128

130

120

122

124

126

118

Time, wk

mm

Hg

Systolic BP

0 4 8 12 16 20 24

76

70

72

74

68

Time, wk

mm

Hg

Diastolic BP

0 4 8 12 16 20 24

P <.001 P = .04

P = .01 P = .90 P = .12

P = .09 P = .39 P = .43

P = .052

HbA1c indicates glycated hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; BP, blood pressure. Error bars indi-cate SEM. The P value at the lower left of each panel indicates the comparison between high–cereal fiber diet vs a low–glycemic index diet as change from week 0 toweek 24 for each measurement by intention-to-treat analysis using an analysis of covariance model.

LOW–GLYCEMIC INDEX OR HIGH–CEREAL FIBER DIET AND TYPE 2 DIABETES

©2008 American Medical Association. All rights reserved. (Reprinted) JAMA, December 17, 2008—Vol 300, No. 23 2751








90

82

84

86

88

80

Time, wk

kg

Body weight

0 4 8 12 16 20 24

7.30

6.50

6.70

6.90

7.10

6.30

Time, wk

%

HbA1c

0 4 8 12 16 20 24

146

122

130

138

114

Time, wk

mg/

dL

Fasting glucose

0 4 8 12 16 20 24

46.0

40.0

42.0

44.0

38.0

Time, wk

mg/

dL

HDL-C

0 4 8 12 16 20 24

140

110

120

130

100

Time, wk

mg/

dL

Triglycerides

0 4 8 12 16 20 24

4.5

3.9

4.1

4.3

3.7

Time, wk

Rat

io


0 4 8 12 16 20 24

2.7

2.3

2.5

2.1

Time, wk

Rat

io

LDL-C : HDL-C

0 4 8 12 16 20 24

128

130

120

122

124

126

118

Time, wk

mm

Hg

Systolic BP

0 4 8 12 16 20 24

76

70

72

74

68

Time, wk

mm

Hg

Diastolic BP

0 4 8 12 16 20 24

P <.001 P = .04

P = .01 P = .90 P = .12

P = .09 P = .39 P = .43

P = .052






























90

82

84

86

88

80

Time, wk

kg

Body weight

0 4 8 12 16 20 24

7.30

6.50

6.70

6.90

7.10

6.30

Time, wk

%

HbA1c

0 4 8 12 16 20 24

146

122

130

138

114

Time, wk

mg/

dL

Fasting glucose

0 4 8 12 16 20 24

46.0

40.0

42.0

44.0

38.0

Time, wk

mg/

dL

HDL-C

0 4 8 12 16 20 24

140

110

120

130

100

Time, wk

mg/

dL

Triglycerides

0 4 8 12 16 20 24

4.5

3.9

4.1

4.3

3.7

Time, wk

Rat

io


0 4 8 12 16 20 24

2.7

2.3

2.5

2.1

Time, wk

Rat

io

LDL-C : HDL-C

0 4 8 12 16 20 24

128

130

120

122

124

126

118

Time, wk

mm

Hg

Systolic BP

0 4 8 12 16 20 24

76

70

72

74

68

Time, wk

mm

Hg

Diastolic BP

0 4 8 12 16 20 24

P <.001 P = .04

P = .01 P = .90 P = .12

P = .09 P = .39 P = .43

P = .052






























90

82

84

86

88

80

Time, wk

kg

Body weight

0 4 8 12 16 20 24

7.30

6.50

6.70

6.90

7.10

6.30

Time, wk

%

HbA1c

0 4 8 12 16 20 24

146

122

130

138

114

Time, wk

mg/

dL

Fasting glucose

0 4 8 12 16 20 24

46.0

40.0

42.0

44.0

38.0

Time, wk

mg/

dL

HDL-C

0 4 8 12 16 20 24

140

110

120

130

100

Time, wk

mg/

dL

Triglycerides

0 4 8 12 16 20 24

4.5

3.9

4.1

4.3

3.7

Time, wk

Rat

io


0 4 8 12 16 20 24

2.7

2.3

2.5

2.1

Time, wk

Rat

io

LDL-C : HDL-C

0 4 8 12 16 20 24

128

130

120

122

124

126

118

Time, wk

mm

Hg

Systolic BP

0 4 8 12 16 20 24

76

70

72

74

68

Time, wk

mm

Hg

Diastolic BP

0 4 8 12 16 20 24

P <.001 P = .04

P = .01 P = .90 P = .12

P = .09 P = .39 P = .43

P = .052

















































90

82

84

86

88

80

Time, wk

kg

Body weight

0 4 8 12 16 20 24

7.30

6.50

6.70

6.90

7.10

6.30

Time, wk

%

HbA1c

0 4 8 12 16 20 24

146

122

130

138

114

Time, wk

mg/

dL

Fasting glucose

0 4 8 12 16 20 24

46.0

40.0

42.0

44.0

38.0

Time, wk

mg/

dL

HDL-C

0 4 8 12 16 20 24

140

110

120

130

100

Time, wk

mg/

dL

Triglycerides

0 4 8 12 16 20 24

4.5

3.9

4.1

4.3

3.7

Time, wk

Rat

io


0 4 8 12 16 20 24

2.7

2.3

2.5

2.1

Time, wk

Rat

io

LDL-C : HDL-C

0 4 8 12 16 20 24

128

130

120

122

124

126

118

Time, wk

mm

Hg

Systolic BP

0 4 8 12 16 20 24

76

70

72

74

68

Time, wkm

m H

g

Diastolic BP

0 4 8 12 16 20 24

P <.001 P = .04

P = .01 P = .90 P = .12

P = .09 P = .39 P = .43

P = .052











90

82

84

86

88

80

Time, wk

kg

Body weight

0 4 8 12 16 20 24

7.30

6.50

6.70

6.90

7.10

6.30

Time, wk%

HbA1c

0 4 8 12 16 20 24

146

122

130

138

114

Time, wk

mg/

dL

Fasting glucose

0 4 8 12 16 20 24

46.0

40.0

42.0

44.0

38.0

Time, wk

mg/

dL

HDL-C

0 4 8 12 16 20 24

140

110

120

130

100

Time, wk

mg/

dL

Triglycerides

0 4 8 12 16 20 24

4.5

3.9

4.1

4.3

3.7

Time, wk

Rat

io


0 4 8 12 16 20 24

2.7

2.3

2.5

2.1

Time, wk

Rat

io

LDL-C : HDL-C

0 4 8 12 16 20 24

128

130

120

122

124

126

118

Time, wk

mm

Hg

Systolic BP

0 4 8 12 16 20 24

76

70

72

74

68

Time, wk

mm

Hg

Diastolic BP

0 4 8 12 16 20 24

P <.001 P = .04

P = .01 P = .90 P = .12

P = .09 P = .39 P = .43

P = .052
























The Journal of Nutrition

Nutrition and Disease

Whole-Grain Foods Do Not Affect InsulinSensitivity or Markers of Lipid Peroxidationand Inflammation in Healthy, ModeratelyOverweight Subjects1,2

Agneta Andersson,3* Siv Tengblad,3 Brita Karlstrom,3 Afaf Kamal-Eldin,4 Rikard Landberg,4 Samar Basu,3

Per Aman,4 and Bengt Vessby3

3Clinical Nutrition and Metabolism, Department of Public Health and Caring Sciences, Uppsala University, 751 85 Uppsala, Swedenand 4Department of Food Science, the Swedish University of Agriculture Sciences (SLU), 750 07 Uppsala, Sweden

Abstract

High intakes of whole grain foods are inversely related to the incidence of coronary heart diseases and type 2 diabetes, but

themechanisms remain unclear. Our study aimed to evaluate the effects of a diet rich in whole grains comparedwith a diet

containing the same amount of refined grains on insulin sensitivity and markers of lipid peroxidation and inflammation. In a

randomized crossover study, 22 women and 8 men (BMI 28 6 2) were given either whole-grain or refined-grain products

(3 bread slices, 2 crisp bread slices, 1 portion muesli, and 1 portion pasta) to include in their habitual daily diet for two 6-wk

periods. Peripheral insulin sensitivity was determined by euglycemic hyperinsulinemic clamp tests. 8-Iso-prostaglandin F2a

(8-iso PGF2a), an F2-isoprostane, wasmeasured in the urine as amarker of lipid peroxidation, and highly sensitive C-reactive

protein and IL-6 were analyzed in plasma as markers of inflammation. Peripheral insulin sensitivity [mg glucose ! kg body

wt21 !min21 per unit plasma insulin (mU/L)3 100] did not improvewhen subjects consumedwhole-grain products (6.86 3.0

at baseline and6.562.7 after 6wk) or refinedproducts (6.462.9 and6.963.2, respectively) and therewereno differences

between the 2 periods. Whole-grain consumption also did not affect 8-iso-PGF2a in urine, IL-6 and C-reactive protein in

plasma, blood pressure, or serum lipid concentrations. In conclusion, substitution of whole grains (mainly based on milled

wheat) for refined-grain products in the habitual daily diet of healthy moderately overweight adults for 6-wk did not affect

insulin sensitivity or markers of lipid peroxidation and inflammation. J. Nutr. 137: 1401–1407, 2007.

Introduction

Whole-grain products are reported to have several positive effectson human health (1). An inverse, relatively strong correlationbetween the intake of whole grain foods (2–6) and fiber fromgrains (7–10), based mainly on FFQ and the incidence of coro-nary heart disease, is consistently shown in epidemiological studiesof both men and women. In addition, recent studies have linkedcereal fiber and whole-grain foods to a reduced risk of type 2diabetes (11–16) and of the metabolic syndrome (6,17). Theserelations seem to be most striking among overweight subjects(11,18,19). The scientific evidence is considered sufficient to permithealth claims regarding the cardio-protective effect of whole-

grain products in many countries including the U.S., the U.K.,and Sweden. The claims must, however, be set within the contextof other lifestyle factors such as exercise and healthy eating habitsin general (1).

Despite indications that whole grain foods may beneficiallyinfluence glucose and lipid metabolism, knowledge of howbiological mechanisms contribute to the health effects of wholegrain remain weak. Several bioactive components, such as die-tary fiber, vitamins, minerals, antioxidants, and other phyto-protectants in whole grain may act synergistically to lower therisk of chronic diseases (20,21). Insulin resistance and oxidativestress are both important factors in the pathogenesis of type 2diabetes and cardiovascular diseases (22–25) and may poten-tially be affected by whole-grain intake. Induction of lipid per-oxidation in humans has been associated with impairment ofinsulin sensitivity along with a proportional increase in specificmarkers of lipid peroxidation and inflammation (26). In a studyof patientswith coronary heart disease, the consumption of whole-grain products, in combination with other plant products, re-ducedmarkers of lipid peroxidation (27). A healthy dietary patternthat includes whole grain products lowers serum insulin concen-trations (28), and improved glycemic tolerance was found in

1 Supported by grants from the Swedish Governmental Agency for InnovationSystems (VINNOVA), the Swedish Research Council for Environment,Agricultural Sciences and Spatial Planning (FORMAS), the Swedish ResearchCouncil, and the Swedish Diabetes Association. Supported by food productsfrom Lantmannen Food R&D AB, Wasa Brod AB and ICA AB.2 Author disclosures: A. Andersson, S. Tengblad, B. Karlstrom, A. Kamal-Eldin,R. Landberg, S. Basu, P. Aman, and B. Vessby, no conflicts of interest.* To whom correspondence should be addressed. E-mail: [email protected].

0022-3166/07 $8.00 ª 2007 American Society for Nutrition. 1401Manuscript received 31 January 2007. Initial review completed 8 February 2007. Revision accepted 10 March 2007.

by guest on February 9, 2011jn.nutrition.org

Dow

nloaded from


Nutrition and Disease

Whole-Grain Foods Do Not Affect InsulinSensitivity or Markers of Lipid Peroxidationand Inflammation in Healthy, ModeratelyOverweight Subjects1,2

Agneta Andersson,3* Siv Tengblad,3 Brita Karlstrom,3 Afaf Kamal-Eldin,4 Rikard Landberg,4 Samar Basu,3

Per Aman,4 and Bengt Vessby3

3Clinical Nutrition and Metabolism, Department of Public Health and Caring Sciences, Uppsala University, 751 85 Uppsala, Swedenand 4Department of Food Science, the Swedish University of Agriculture Sciences (SLU), 750 07 Uppsala, Sweden

Abstract

High intakes of whole grain foods are inversely related to the incidence of coronary heart diseases and type 2 diabetes, but

themechanisms remain unclear. Our study aimed to evaluate the effects of a diet rich in whole grains comparedwith a diet

containing the same amount of refined grains on insulin sensitivity and markers of lipid peroxidation and inflammation. In a

randomized crossover study, 22 women and 8 men (BMI 28 6 2) were given either whole-grain or refined-grain products

(3 bread slices, 2 crisp bread slices, 1 portion muesli, and 1 portion pasta) to include in their habitual daily diet for two 6-wk

periods. Peripheral insulin sensitivity was determined by euglycemic hyperinsulinemic clamp tests. 8-Iso-prostaglandin F2a

(8-iso PGF2a), an F2-isoprostane, wasmeasured in the urine as amarker of lipid peroxidation, and highly sensitive C-reactive

protein and IL-6 were analyzed in plasma as markers of inflammation. Peripheral insulin sensitivity [mg glucose ! kg body

wt21 !min21 per unit plasma insulin (mU/L)3 100] did not improvewhen subjects consumedwhole-grain products (6.86 3.0

at baseline and6.562.7 after 6wk) or refinedproducts (6.462.9 and6.963.2, respectively) and therewereno differences

between the 2 periods. Whole-grain consumption also did not affect 8-iso-PGF2a in urine, IL-6 and C-reactive protein in

plasma, blood pressure, or serum lipid concentrations. In conclusion, substitution of whole grains (mainly based on milled

wheat) for refined-grain products in the habitual daily diet of healthy moderately overweight adults for 6-wk did not affect

insulin sensitivity or markers of lipid peroxidation and inflammation. J. Nutr. 137: 1401–1407, 2007.

Introduction

Whole-grain products are reported to have several positive effectson human health (1). An inverse, relatively strong correlationbetween the intake of whole grain foods (2–6) and fiber fromgrains (7–10), based mainly on FFQ and the incidence of coro-nary heart disease, is consistently shown in epidemiological studiesof both men and women. In addition, recent studies have linkedcereal fiber and whole-grain foods to a reduced risk of type 2diabetes (11–16) and of the metabolic syndrome (6,17). Theserelations seem to be most striking among overweight subjects(11,18,19). The scientific evidence is considered sufficient to permithealth claims regarding the cardio-protective effect of whole-

grain products in many countries including the U.S., the U.K.,and Sweden. The claims must, however, be set within the contextof other lifestyle factors such as exercise and healthy eating habitsin general (1).

Despite indications that whole grain foods may beneficiallyinfluence glucose and lipid metabolism, knowledge of howbiological mechanisms contribute to the health effects of wholegrain remain weak. Several bioactive components, such as die-tary fiber, vitamins, minerals, antioxidants, and other phyto-protectants in whole grain may act synergistically to lower therisk of chronic diseases (20,21). Insulin resistance and oxidativestress are both important factors in the pathogenesis of type 2diabetes and cardiovascular diseases (22–25) and may poten-tially be affected by whole-grain intake. Induction of lipid per-oxidation in humans has been associated with impairment ofinsulin sensitivity along with a proportional increase in specificmarkers of lipid peroxidation and inflammation (26). In a studyof patientswith coronary heart disease, the consumption of whole-grain products, in combination with other plant products, re-ducedmarkers of lipid peroxidation (27). A healthy dietary patternthat includes whole grain products lowers serum insulin concen-trations (28), and improved glycemic tolerance was found in

1 Supported by grants from the Swedish Governmental Agency for InnovationSystems (VINNOVA), the Swedish Research Council for Environment,Agricultural Sciences and Spatial Planning (FORMAS), the Swedish ResearchCouncil, and the Swedish Diabetes Association. Supported by food productsfrom Lantmannen Food R&D AB, Wasa Brod AB and ICA AB.2 Author disclosures: A. Andersson, S. Tengblad, B. Karlstrom, A. Kamal-Eldin,R. Landberg, S. Basu, P. Aman, and B. Vessby, no conflicts of interest.* To whom correspondence should be addressed. E-mail: [email protected].

0022-3166/07 $8.00 ª 2007 American Society for Nutrition. 1401Manuscript received 31 January 2007. Initial review completed 8 February 2007. Revision accepted 10 March 2007.


Downloaded from

inflammation are presumed to be central factors involved in theetiology of coronary heart disease and type 2 diabetes (22–25)and are suggested to contribute to the beneficial effects by wholegrain. However, we did not find a significant effect of whole-grain intake on insulin sensitivity, lipid peroxidation, inflamma-tory markers, or on any of the metabolic variables studied.

In the present study, 30 participants were included to allowan 80% power to detect a 10% change in insulin sensitivity. Thispermitted the possibility of detecting a clinically important effectof whole grain on insulin sensitivity and gave a low risk for type2 error. However, there were no indications of any improvementafter subjects consumed a diet enriched with whole-grain foods,suggesting that an effect on insulin sensitivity would be small,if any. An intervention of 6 wk might be too short to achievesignificant effects on insulin sensitivity. However, one earlierdietary intervention study with a similar fiber intake and includ-ing 11 subjects with similar BMI as in the present study, indi-cated improved insulin sensitivity after 6-wk of consuming awhole-grain diet. Their fasting insulin concentrations were low-ered by 10% during the whole-grain diet period and the glucoseinfusion rate increased during the final 30 min of a clamp test(32). Those subjects were hyperinsulinemic, with fasting insulinlevels;3 times that of the present study. Dietary effects are morelikely to occur in individuals with a poor habitual diet and morepronounced metabolic abnormalities.

The whole-grain products in the present study were mostlybased onmilled flour with small particle size in the form of breadand pasta.Wheat, rye, oats, and rice were all included, but wheatclearly dominated. Improved blood glucose and insulin metab-olism following a higher fiber intake was particularly evident forsoluble fiber (29–31). The domination of wheat indicated in the

present study may reduce the effects on glucose and insulinbecause wheat contains less soluble andmore insoluble fiber thanrye, oats, or barley. However, two other intervention studiesshowed no effect of high-fiber cereals (even when high in solublefiber; rye and oats) on insulin sensitivity measured indirectly byintravenous-glucose tolerance test (44,45). Paradoxically, a stron-ger inverse relationship with the intake of insoluble fiber thansoluble fiber on the risk of diabetes is suggested (16).

Other aspects of the cereal products may be of importance.The postprandial insulin response to grain products has beendetermined more by the form of food and botanic structure thanby the amount of fiber or type of cereal (31,46).Whole-grain breadwith a more intact structure has been shown to improve post-prandial glycemic and insulinemic responses than whole-wheatbread made from milled flour (47).

Based on epidemiological studies, positive health effects canbe expected at a level of 3 servings of whole-grain foods per day(48). Our study included 7 servings per day, which is comparableto the dietary intervention by Pereira et al. (32). In that study, allof the food consumed was provided to the subjects on a 6 dmenurotation. In our study, only cereal products were provided as apart of each individual’s habitual diet. The products eaten werecarefully noted in the diaries. The reported nutrient intakes dur-ing both diet periods were similar, except for a higher intake ofseveral minerals, fiber, a-tocopherol, and linoleic acid, 18:2 (n-6)during the whole-grain period, as expected. Diet adherence wastherefore considered good in the present study.

Another study suggested that whole-grain intake, in combi-nation with other plant products, reduced markers of lipid per-oxidation (27). The intervention period in that study was 16 wk.The whole grain was supplied in the form of coarse powder, and

TABLE 5 BMI, blood pressure, and blood chemistry of all participants before and after 6 wk consumingwhole-grain or refined-grain diets1

Whole-grain period Refined-grain period

Before After Before After P-value treatment effect2

n 30 30 30 30BMI, kg/m2 28.5 6 2.4 28.8 6 2.5a 28.4 6 2.1 28.6 6 2.1 0.046Fasting blood glucose, mmol/L 5.2 6 0.8 5.3 6 0.8 5.2 6 0.9 5.2 6 0.8 0.28Fasting insulin, pmol/L 56.2 6 22.9 57.6 6 24.3 60.4 6 30.6 57.6 6 25.7 0.47Insulin sensitivity,3 M 5.9 6 2.1 5.5 6 1.7 5.7 6 1.9 6.0 6 2.0 0.24

M/I 6.8 6 3.0 6.5 6 2.7 6.4 6 2.9 6.9 6 3.2 0.79Total cholesterol, mmol/L 5.5 6 0.7 5.5 6 0.7 5.5 6 0.8 5.5 6 0.7 0.76HDL cholesterol, mmol/L 1.3 6 0.3 1.2 6 0.3 1.2 6 0.2 1.2 6 0.3 0.15LDL cholesterol, mmol/L 3.7 6 0.8 3.7 6 0.7 3.7 6 0.8 3.6 6 0.7 0.40TG cholesterol, mmol/L 1.4 6 0.8 1.5 6 0.8 1.3 6 0.6 1.6 6 1.0c 0.19Free fatty acid, mmol/L 0.56 6 0.19 0.61 6 0.18 0.63 6 0.17 0.62 6 0.18 0.99Systolic blood pressure, mm Hg 130 6 17 129 6 15 130 6 16 130 6 15 0.35*Diastolic blood pressure, mm Hg 81 6 9 81 6 8 80 6 10 81 6 9 0.608-iso-PGF2a, nmol/mmol creatinine 0.43 6 0.14 0.43 6 0.14 0.42 6 0.15 0.44 6 0.21 0.48a-tocopherol, mmol/mmol lipid 4.68 6 0.72 4.78 6 0.61 4.38 6 1.07 4.64 6 0.61 0.08g-tocopherol, mmol/mmol lipid 0.26 6 0.12 0.24 6 0.07 0.26 6 0.10 0.26 6 0.10 0.10CRP, mg/L 2.03 6 1.62 2.38 6 2.29 2.86 6 2.96 2.34 6 1.57 0.55IL-6, ng/L 14.8 6 32.2 15.2 6 33.2 15.9 6 32.4 15.8 6 30.9 0.79PAI-1 activity, kU/L 24.7 6 15.8 26.9 6 20.3 24.8 6 19.9 22.1 6 19.5 0.26

1 Data are means 6 SD.2 P-values (treatment effect) for differences between the whole-grain and refined-grain diet adjusted for changes in BMI. Differences within

groups when compared to baseline: aP , 0.001; bP , 0.01; cP , 0.05. *Parallel group design, only from 1st diet period (because carryover

effect was found).3 M, glucose disposal during clamp in mg glucose ! kg body wt21 ! min21. M/I, glucose disposal rate during clamp in mg glucose ! kg body

wt21 ! min21 per unit plasma insulin ( mU/L ) 3 100. To convert glucose from mg to mmol, divide by 180. To convert insulin from mU/L to

pmol/L, multiply by 6.945.

Whole-grain foods and insulin sensitivity 1405


Dow

nloaded from

J. Nutr. 137: 1401–1407, 2007.

Sem diferenças entre os grupos

Effect of Wheat Bran on Glycemic Controland Risk Factors for CardiovascularDisease in Type 2 DiabetesDAVID J. A. JENKINS, MD

1,2,3,4

CYRIL W. C. KENDALL, PHD1,3

LIVIA S. A. AUGUSTIN, MSC1,3

MARGARET C. MARTINI, PHD5

METTE AXELSEN, PHD6

DOROTHEA FAULKNER, RD1

EDWARD VIDGEN, BSC1,3

TINA PARKER, RD1

HERB LAU, MD7,8

PHILIP W. CONNELLY, PHD2,9,10

JEROME TEITEL, MD7,8

WILLIAM SINGER, MD2

ARTHUR C. VANDENBROUCKE, PHD7,10

LAWRENCE A. LEITER, MD1,2,3,4

ROBERT G. JOSSE, MD1,2,3,4

OBJECTIVE — Cohort studies indicate that cereal fiber reduces the risk of diabetes andcoronary heart disease (CHD). Therefore, we assessed the effect of wheat bran on glycemiccontrol and CHD risk factors in type 2 diabetes.

RESEARCH DESIGN AND METHODS — A total of 23 subjects with type 2 diabetes(16 men and 7 postmenopausal women) completed two 3-month phases of a randomizedcrossover study. In the test phase, bread and breakfast cereals were provided as products high incereal fiber (19 g/day additional cereal fiber). In the control phase, supplements were low in fiber(4 g/day additional cereal fiber).

RESULTS — Between the test and control treatments, no differences were seen in bodyweight, fasting blood glucose, HbA1c, serum lipids, apolipoproteins, blood pressure, serum uricacid, clotting factors, homocysteine, C-reactive protein, magnesium, calcium, iron, or ferritin.LDL oxidation in the test phase was higher than that seen in the control phase (12.1 ! 5.4%, P "0.034). Of the subjects originally recruited, more dropped out of the study for health and foodpreference reasons from the control phase (16 subjects) than the test phase (11 subjects).

CONCLUSIONS — High-fiber cereal foods did not improve conventional markers of glyce-mic control or risk factors for CHD in type 2 diabetes over 3 months. Possibly longer studies arerequired to demonstrate the benefits of cereal fiber. Alternatively, cereal fiber in the diet may bea marker for another component of whole grains that imparts health advantages or a healthylifestyle.

Diabetes Care 25:1522–1528, 2002

There is much interest in the possiblehealth benefits of fiber-containingcereals (1–3). The exact component

or facet of fiber that is responsible has notbeen clearly defined, and there are indi-cations that the whole grain confers met-abolic benefits (4) and reduces the risk ofchronic disease (1,5,6). The results oflarge cohort studies have suggested thatwheat fiber protects against the develop-ment of diabetes (1–3). Many diabetes as-sociations advise increased fiber intake,either to improve glycemic control (7) orto confer general health benefits (8). In-creases in fiber from a variety of dietarysources have been shown to improve gly-cemic control in type 2 diabetes (9). Earlystudies suggested that cereal fiber im-proved both glycemic control in diabetes(10) and glucose tolerance in nondiabeticsubjects (11). The reason for the benefi-cial effects of nonviscous cereal fiber is notclear. Cereal fibers do not reduce the rateof gastric emptying and small intestinalabsorption or flatten the postprandial gly-cemic response to a high-carbohydratetest meal (12). In contrast, viscous fiberssuch as guar and pectin have been shownto reduce the rate of gastric emptying (13)and small intestinal absorption (14),thereby providing a mechanism for po-tential benefits. These fibers have beenshown to reduce postprandial glycemiawhen added to test meals. They also de-crease 24-h urinary glucose losses whenadded to the diets of subjects with type 2diabetes (15).

Furthermore, it is wheat fiber, ratherthan viscous fiber, that for more than twodecades has been shown consistently incohort studies to be associated with a re-duced risk of heart disease (5,6,16,17).These effects are seen despite the fact thatviscous fibers from oats, barley, psyllium,pectins, and guar gum have been shownto lower serum cholesterol and improvethe blood lipid profile, whereas the insol-uble fibers were largely without effect(18,19).

In view of the apparent benefits of ce-real fiber in preventing diabetes and car-diovascular disease and the lack of an

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

From the 1Clinical Nutrition and Risk Factor Modification Center, St. Michael’s Hospital, Toronto, Ontario,Canada; the 2Department of Medicine, Division of Endocrinology and Metabolism, St. Michael’s Hospital,Toronto, Ontario, Canada; the 3Department of Nutritional Sciences, Faculty of Medicine, University ofToronto, Toronto, Ontario, Canada; the 4Department of Medicine, Faculty of Medicine, University ofToronto, Toronto, Ontario, Canada; 5Kraft Foods, Glenview, Illinois; the 6Lundberg Laboratory for DiabeticResearch, Department of Internal Medicine, Sahlgrenska University Hospital, Goteborg, Sweden; the 7De-partment of Laboratory Medicine, Division of Clinical Biochemistry, St. Michael’s Hospital, Toronto, On-tario, Canada; the 8Department of Hematology, St. Michael’s Hospital, Toronto, Ontario, Canada; the9Department of Biochemistry, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada; andthe 10Department of Laboratory Medicine and Pathobiology, Faculty of Medicine, University of Toronto,Toronto, Ontario, Canada.

Address correspondence and reprint requests to David J. A. Jenkins, Clinical Nutrition and Risk FactorModification Center, St. Michael’s Hospital, 61 Queen St. East, Toronto, Ontario, Canada, M5C 2T2. E-mail:[email protected].

Received for publication 12 April 2002 and accepted in revised form 28 May 2002.Abbreviations: CHD, coronary heart disease; NCEP, National Cholesterol Education Program.A table elsewhere in this issue shows conventional and Systeme International (SI) units and conversion

factors for many substances.See accompanying editorial on p. 1652.

C l i n i c a l C a r e / E d u c a t i o n / N u t r i t i o nO R I G I N A L A R T I C L E

1522 DIABETES CARE, VOLUME 25, NUMBER 9, SEPTEMBER 2002


1,2,3,4




METTE AXELSEN, PHD6



TINA PARKER, RD1

HERB LAU, MD7,8



WILLIAM SINGER, MD2








Diabetes Care 25:1522–1528, 2002





! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !








1,2,3,4




METTE AXELSEN, PHD6



TINA PARKER, RD1

HERB LAU, MD7,8



WILLIAM SINGER, MD2








Diabetes Care 25:1522–1528, 2002





! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !








1,2,3,4




METTE AXELSEN, PHD6



TINA PARKER, RD1

HERB LAU, MD7,8



WILLIAM SINGER, MD2








Diabetes Care 25:1522–1528, 2002





! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !







Jenkins D, et al. Diabetes Care 25:1522–1528, 2002

AUMENTO DA INGESTÃO DE FIBRA A PARTIR DE CEREAIS INTEGRAIS DE 9 PARA 17 GRAMAS

AUMENTO NÃO SIGNIFICATIVO DO RISCO RELATIVO EM 18%

Burr ML, Fehily AM, Gilbert JF, et al. Lancet 1989; 2:757-761.

Diet And Reinfarction Trial

ARTICLE

A Palaeolithic diet improves glucose tolerancemore than a Mediterranean-like diet in individualswith ischaemic heart disease

S. Lindeberg & T. Jönsson & Y. Granfeldt &E. Borgstrand & J. Soffman & K. Sjöström & B. Ahrén

Received: 1 May 2007 /Accepted: 4 May 2007# Springer-Verlag 2007

AbstractAims/hypothesis Most studies of diet in glucose intoleranceand type 2 diabetes have focused on intakes of fat, carbo-hydrate, fibre, fruits and vegetables. Instead, we aimed tocompare diets that were available during human evolutionwith more recently introduced ones.Methods Twenty-nine patients with ischaemic heart diseaseplus either glucose intolerance or type 2 diabetes were ran-domised to receive (1) a Palaeolithic (‘Old Stone Age’) diet(n=14), based on lean meat, fish, fruits, vegetables, rootvegetables, eggs and nuts; or (2) a Consensus (Mediterra-nean-like) diet (n=15), based on whole grains, low-fat dairyproducts, vegetables, fruits, fish, oils and margarines. Pri-mary outcome variables were changes in weight, waist cir-cumference and plasma glucose AUC (AUC Glucose0–120)and plasma insulin AUC (AUC Insulin0–120) in OGTTs.Results Over 12 weeks, there was a 26% decrease of AUCGlucose0–120 (p=0.0001) in the Palaeolithic group and a 7%decrease (p=0.08) in the Consensus group. The larger (p=0.001) improvement in the Palaeolithic group was indepen-dent (p=0.0008) of change in waist circumference (!5.6 cmin the Palaeolithic group, !2.9 cm in the Consensus group;

p=0.03). In the study population as a whole, there was norelationship between change in AUC Glucose0–120 andchanges in weight (r=!0.06, p=0.9) or waist circumference(r=0.01, p=1.0). There was a tendency for a larger decreaseof AUC Insulin0–120 in the Palaeolithic group, but because ofthe strong association between change in AUC Insulin0–120and change in waist circumference (r=0.64, p=0.0003), thisdid not remain after multivariate analysis.Conclusions/interpretation A Palaeolithic diet may im-prove glucose tolerance independently of decreased waistcircumference.

Keywords Diet . Evolution . Glucose intolerance .

Ischaemic heart disease . Palaeolithic diet . Type 2 diabetes

AbbreviationsBIA bioelectrical impedance analysisE% percentage of total energy intakeHOMA-IR homeostasis model assessment

of insulin resistanceIFG impaired fasting glucoseIGT impaired glucose toleranceIHD ischaemic heart diseaseNGT normal glucose tolerance

Introduction

Impaired glucose tolerance (IGT) and type 2 diabetes arecommon risk factors for ischaemic heart disease (IHD) [1, 2],which negatively affect the long-term prognosis aftermyocardial infarction [3, 4]. In fact, cross-sectional studieshave found only 35–54% of IHD patients have normalglucose tolerance (NGT) [5–11]. Increased physical activity,

DiabetologiaDOI 10.1007/s00125-007-0716-y

Electronic supplementary material The online version of this article(doi:10.1007/s00125-007-0716-y) contains supplementary material,which is available to authorised users.

S. Lindeberg (*) : T. Jönsson : E. Borgstrand : J. Soffman :K. Sjöström :B. AhrénDepartment of Medicine, Hs 32, University of Lund,SE-221 85 Lund, Swedene-mail: [email protected]

Y. GranfeldtDepartment of Applied Nutrition and Food Chemistry,University of Lund,Lund, Sweden

7% in the Consensus group (p=0.10), and the differencebetween the groups was highly significant. After 12 weeks,all 14 subjects in the Palaeolithic group had normal values,compared with 7 of 15 subjects in the Consensus group (p=0.0007 for group difference; Table 4). At 12 weeks, fivesubjects in the Consensus group still had diabetic values.

There was a decrease of HOMA-IR in both groups withno significant difference between the two groups (Table 4).The QUICKI index of insulin sensitivity [1/(ln fastingplasma insulin+ln fasting plasma glucose)] did not changemore in the Palaeolithic group than in the Consensus group(p=0.23, data not shown). The early phase of post-challenge glucose and insulin responses, as represented by

Incremental AUC Glucose0–30 and Incremental AUCInsulin0–30, did not change significantly during the trial,although a trend towards lowered Incremental AUCInsulin0–30 was seen in both groups (Table 4).

Reported food composition differed between the twogroups such that subjects in the Palaeolithic group had amuch lower intake of dairy products, cereals and oil/margarine, and a higher intake of fruits and nuts (Table 5).The intake of vegetables, meat, meat products or fish didnot differ significantly between the groups. Total fat intakewas low with no difference between the groups (Table 6).Absolute protein intake was identical in the two groupswhile relative protein intake (as a percentage of total energyintake [E%]) was higher in the Palaeolithic group. Absolutecarbohydrate intake was 43% lower in the Palaeolithicgroup, and 23% lower in terms of E%. Glycaemic load was47% lower in the Palaeolithic group and correlated stronglywith cereal intake (r=0.75, p<0.0001).

Energy intake was 25% lower in the Palaeolithic group(p=0.004; Table 6) despite similar quantities of consumedfood (by weight; Table 5). After adjustment for energyintake, the improvement of AUC Glucose0–120 was stilllarger in the Palaeolithic group (p=0.02; SupplementaryTable 2), while the larger waist loss, and the tendency forlarger decrease of AUC Insulin0–120, compared with theConsensus group, disappeared (Supplementary Table 3).

In post hoc analysis among the whole study population, apositive association between intake of cereals and change inwaist circumference explained 42% of waist loss among thewhole study population (p=0.0003; Supplementary Table 6),and 40% in the Consensus group alone (p=0.016). In con-trast, there was a negative correlation between fruit intakeand change in waist circumference, which explained 21% ofwaist loss (p=0.01). Each of these associations remainedsignificant after adjustment for dietary assignment, energyintake, carbohydrate intake or glycaemic load (Supplemen-tary Table 5). Thus, waist loss increased with increasing

a b

100

1,000

0 20 40 60 80 100 120

Time (min)

Pla

sma in

sulin

(pm

ol/l

)

*

100

1,000

0 20 40 60 80 100 120

Time (min)

Pla

sma insu

lin (

pm

ol/l

)

*

**

Fig. 3 Plasma insulin duringOGTTs at study start (baseline,closed circles) and after 12weeks (open circles) in thePalaeolithic (a) and Consensus(b) groups. Values are meansSE. *p<0.05; **p>0.01

0

200

400

600

800

1,000

1,200

1,400

Palaeolithic Consensus

Glu

cose

AU

C0!

120

(mm

ol/l

x m

in)

Fig. 2 Mean glucose AUCs (0–120 min) during OGTTs at study start(baseline, light grey columns) and after 6 weeks (dark grey columns)and 12 weeks (black columns) in the Palaeolithic and Consensusgroups. Error bars denote 95% CIs

Diabetologia

Início

Início

6 sem

6 sem

12 sem

12 sem

ARTICLE










Introduction






Dieta Paleolitica (n=14)

Dieta Med (n=15) P

Início do Estudo 10 9 0.4 6 Semanas 1 3 0.2 12 Semanas 0 5 0.01

Nº INDIVÍDUOS COM DIABETES (GLICEMIA PÓS OGT)

ARTICLE










Introduction






NUTRITION MYTHS

e-Book

MITO 2





MITO 2





Effects of a low-fat dietary intervention on glucose, insulin,and insulin resistance in the Women’s Health Initiative (WHI)Dietary Modification trial1–3

James M Shikany, Karen L Margolis, Mary Pettinger, Rebecca D Jackson, Marian C Limacher, Simin Liu,Lawrence S Phillips, and Lesley F Tinker

ABSTRACTBackground: Glycemic effects of the Women’s Health Initiative(WHI) low-fat dietary intervention are unknown.Objective: Our objective was to analyze the effects of the WHIlow-fat dietary intervention on serum glucose and insulin and in-sulin resistance up to 6 y after random assignment.Design: Postmenopausal WHI Dietary Modification trial interven-tion (DM-I) and comparison (DM-C) participants with blood meas-ures at least at baseline and year 1 (n = 2263) were included.Anthropometric measures, dietary assessments, serum glucose andinsulin concentrations, homeostasis model assessment of insulinresistance (HOMA-IR) measures, and quantitative insulin sensitiv-ity check index (QUICKI) values were obtained at baseline, year 1,year 3, and year 6. Changes in measures were compared betweengroups at years 1, 3, and 6 overall and within stratified analyses.Results: Mean (6SD) differences in changes at year 1 between theDM-I and DM-C groups were as follows: glucose, 21.7 6 17.9 mg/dL; insulin, 20.7 6 5.1 lIU/mL; HOMA-IR, 20.2 6 1.9; andQUICKI, 0.004 6 0.019 (all P , 0.05). Similar findings resultedfrom repeated-measures analyses comparing the intervention andcomparison groups over the 6 y. Whereas normoglycemic womenat baseline had a decrease in glucose at year 1 that was 1.9 617.2 mg/dL greater in the DM-I than in the DM-C group, diabeticwomen had an increase in glucose that was 7.96 20.3 mg/dL greaterin the DM-I than in the DM-C group (P for interaction ,0.001).Conclusions: A low-fat diet was not significantly associated withadverse glycemic effects up to 6 y after random assignment in post-menopausal women. However, diabetic women experienced adverseglycemic effects of the low-fat diet. This trial is registered at clin-icaltrials.gov as NCT00000611. Am J Clin Nutr doi: 10.3945/ajcn.110.010843.

INTRODUCTION

The optimal macronutrient content of the diet for human healthremains a major controversy in nutritional science. Low-fat dietsin general, and the Women’s Health Initiative (WHI) low-fatdietary intervention in particular, have been criticized for theirpotential to substitute unhealthy carbohydrates for fat, potentiallycontributing to hyperglycemia, hyperinsulinemia, and insulinresistance (1).

The WHI Dietary Modification (DM) trial was designed to testthe effects of a dietary pattern low in total fat, along with in-creased vegetables, fruit, and grains, on primarily breast cancer

and colorectal cancer incidence in postmenopausal womenduring a mean follow-up of 8.1 y. Despite the increased intake ofcarbohydrate in the intervention group, and question of associ-ated increased risk of diabetes, no increase in diabetes risk wasobserved. Subgroup analysis suggested that greater decreases inpercentage of energy from total fat reduced diabetes risk (P fortrend = 0.04); however, that finding was not statistically sig-nificant after adjustment for weight loss—a common effect ofeating a low-fat diet (2).

Details of the effects of the WHI diet intervention on glucose,insulin, and insulin resistance have not been reported. The aim ofthis report was to analyze the effect of the overall diet in-tervention, and the specific effects of fiber and whole grainintakes, and dietary glycemic index (GI) and glycemic load (GL)on glucose, insulin, and insulin resistance in the WHI DM trial.


WHI DM trial

Recruitment

Details of the study design and methods were published pre-viously (3). All women provided written informed consent, and

1 From the Division of Preventive Medicine, School of Medicine, Univer-sity of Alabama at Birmingham, Birmingham, AL (JMS); HealthPartnersResearch Foundation, Minneapolis, MN (KLM); the Division of PublicHealth Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA(MP and LFT); the Department of Internal Medicine, Division of Endocri-nology, Diabetes and Metabolism, The Ohio State University, Columbus, OH(RDJ); the Division of Cardiovascular Medicine, University of Florida Col-lege of Medicine, Gainesville, FL (MCL); the Department of Epidemiology,University of California, Los Angeles, Los Angeles, CA (SL); and the At-lanta VA Medical Center and Division of Endocrinology, Department ofMedicine, Emory University School of Medicine, Atlanta, GA (LSP).

2 Supported by the National Heart, Lung, and Blood Institute, NationalInstitutes of Health, US Department of Health and Human Services (contractnumbers N01WH22110, 24152, 32100-2, 32105-6, 32108-9, 32111-13,32115, 32118-32119, 32122, 42107-26, 42129-32, and 44221).

3 Address correspondence to JM Shikany, Division of Preventive Medi-cine, School of Medicine, University of Alabama at Birmingham, 1530 3rdAvenue S, MT 610, Birmingham, AL 35294. E-mail: [email protected].

Received December 21, 2010. Accepted for publication April 15, 2011.doi: 10.3945/ajcn.110.010843.

Am J Clin Nutr doi: 10.3945/ajcn.110.010843. Printed in USA. ! 2011 American Society for Nutrition 1 of 11

AJCN. First published ahead of print May 11, 2011 as doi: 10.3945/ajcn.110.010843.

Copyright (C) 2011 by the American Society for Nutrition




INTRODUCTION






WHI DM trial

Recruitment












INTRODUCTION






WHI DM trial

Recruitment









Shikany JM, et al. Am J Clin Nutr. 2011 Jul;94(1):75-85.




INTRODUCTION






WHI DM trial

Recruitment












INTRODUCTION






WHI DM trial

Recruitment









Shikany JM, et al. Am J Clin Nutr. 2011 Jul;94(1):75-85.

INCIDÊNCIA DE DIABETES

Diabetes Care January 2011 vol. 34 no. 1 14-19

MED C/ AZEITE VIRGEM: 10.1 (5.1–15.1)

MED C/ FRUTOS SECOS: 11.0

(5.9–16.1)

DIETA LOW FAT: 17.9 (11.4–24.4)

Ströhle A, Hahn A. Diets of modern hunter-gatherers vary substantially in their carbohydrate content depending on ecoenvironments: results from an ethnographic analysis. Nutrition Research 2011, em publicação

RESTRINGIR HIDRATOS DE CARBONO É PERIGOSO 209 by hunter-gatherer societies showed that there was a large210 variation in the carbohydrate content of the different diets,211 which ranged from approximately 3% of the total energy to212 approximately 50% of the total energy. Interestingly, most213 hunter-gatherer diets (approximately 85%) were character-214 ized by a relatively low carbohydrate intake (b35% of the215 total energy), which reflected the high reliance on animal-216 based foods of most hunter-gatherer societies. For example,217 of the 229 hunter-gatherer populations examined, Cordain218 et al [47] found that the median subsistence dependence on219 animal food was 66% to 75%. Similarly, an analysis of a220 recent ethnographic compilation of P:A subsistence patterns221 showed that most members of historically studied hunter-222 gatherer societies consumed high amounts of animal-source223 food (median P:A ratio, 35:65) [51]. If we interpreted the224 range of percentages of energy consumed as carbohydrates225 in the diets of hunter-gatherer societies from the perspective226 of contemporary levels (see Table 5), 73% (n = 168) of diets227 would be classified as “low-carb,” and only approximately228 26% (n = 59) would be classified as “moderate-carb” [59].229 Therefore, it is not surprising that the range of percentages of230 energy from carbohydrates in the diets of most hunter-231 gatherer societies is markedly different from the amounts232 recommended for healthy individuals (see Table 6). With233 respect to human health, however, there has been increasing234 evidence from observational and clinical studies suggesting235 that diets with restricted carbohydrates do not cause harm per236 se. Interestingly, carbohydrate-restricted diets have been237 shown to have favorable effects on blood lipids [63],238 coronary heart disease risk [64], and body weight [59].239 Because many hunter-gatherer societies consumed low240 levels of carbohydrates, this fact could explain why these241 populations were relatively free of many chronic and242 nutrition-related degenerative diseases, such as obesity,243 type 2 diabetes mellitus, and coronary heart disease [32,65].244 We must take into account, however, that there was a245 variety of less Westernized humans worldwide (eg, horti-246 culturalists and simple agriculturalists) who ingested high-247 carbohydrate diets (approximately 70% of the total energy)248 without having those diseases [32,66]. In this context,

249250251252253254255256257258259260261262Ludwig and Jenkins [67] may have been correct when they263stated that “humans can probably do well over the long term264by consuming diets that vary widely in macronutrients, as265long as adequate attention is paid to nutrient quality.”266Indeed, with respect to carbohydrate quality, all traditional267diets mentioned above differ from modern Western diets.268Except for honey—available seasonally and consumed in269high amounts (yearly average up to 14%-18% of total daily270energy) by some hunter-gatherers (eg, the Hadza and the271Ache) [68,69]—the plant component of the diets of hunter-272gatherers [70], horticulturalists [71,72], and simple agricul-273turalists [73] has been based on minimally processed plant274foods with low glycemic index (GI) values and high levels275of dietary fiber [55,71-73]. In contrast, foods characterized276by a high GI, such as refined sugars and grains, which277currently supply approximately 40% of the total energy in a278typical US diet [74], have rarely been consumed by non-279Westernized individuals. There has been increasing evi-280dence suggesting that high-fiber, low-GI diets are associated281with a lower risk of coronary heart disease and type 2282diabetes [75]. Therefore, the quality of carbohydrates rather283than the amount of carbohydrates may partly explain the284fact that hunter-gatherers were relatively free of nutrition-285related degenerative diseases. This interpretation is286in accordance with findings from a meta-analysis of 37287prospective observational studies, which showed that “there288were more positive associations of greater magnitude

Table 4t4:1Effect of different ecological environments on the ratio of plant-food energy intake to animal-food energy intake (P:A ratio) and the corresponding dietarycarbohydrate intake (percentage of energy) for 63 hunter-gatherer dietst4:2

t4:3 Characterization of theecological environments

P:A ratio byclass interval (%)

Absolute frequency(no. of societies)

Relative frequency(percentage of societies)

Carbohydrate intake(percentage of energy/d)

t4:4 Tundra, northern areas 6-15:85-94 6 9.5 3-9t4:5 Northern coniferous forest 16-25:75-84 14 22.2 10-15t4:6 Temperate forest,

mostly mountainous36-45:55-64 6 9.5 23-28

t4:7 Desert grasses and shrubs 46-55:45-54 11 17.5 29-34t4:8 Temperate grassland 26-35:65-74 11 17.5 16-22t4:9 Subtropical bush 36-45:55-64 2 3.2 23-28t4:10 Subtropical rainforest 36-45:55-64 4 6.3 23-28t4:11 Tropical grassland 46-55:45-54 4 6.3 29-34t4:12 Monsoon forest 36-45:55-64 2 3.2 23-28t4:13 Tropical rainforest 26-35:65-74 3 4.8 16-22

Table 5 t5:1Definitions for low-carbohydrate diets [59] t5:2

t5:3Absolute carbohydratecontent (g/d)

Relative carbohydratecontent (percentageof energy/d)

t5:4Low-carbohydrate-ketogenic diet

b50 b10

t5:5Low-carbohydratediet

50-b130 10-b25

t5:6Moderatecarbohydratediet

!130 26-45

4 A. Ströhle, A. Hahn / Nutrition Research xx (2011) xxx–xxx

!!!!!

ü DT2!

ü Aumento de peso!

ü DCV!

Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M. Lancet 2002;359:2072– 2077

REDUÇÃO FARMACOLÓGICA DA GLICEMIA PÓS-PRANDIAL DIMINUIU RISCO DE:

Brewer N, et al. Diabetes Care. 2008 Jun;31(6):1144-9

ingin

thisgroup

(i.e.,it

didnot

always

increaseas

thelevel

ofA

1Cincreased);

theproportion

who

diedw

as12.5%

(5of

40)in

thosew

ithA

1Clevels

of5.0

to!

6.0%,

25.4%(18

of71)

inthose

with

levelsof6.0

to!

7.0%,and

14.3%(42

of293)in

thosew

ithlevelsof!

7.0%.O

ver-all,a

priordiabetesdiagnosiswasstrongly

associatedw

ithrisk

ofdeath(3.35

[2.58–

4.34];this

differsfrom

theestim

ateof

5.19given

inTable

2,w

hichinvolved

acom

parisonw

iththose

without

diabetesw

ithan

A1C

of4.0to

!5.0%

,whereasthe

comparison

hereis

with

allthosew

ithoutdiabetes),butthisrisk

reducedsom

ewhat

(2.46[1.86

–3.26])w

henadjusted

forA

1C.Ifthe

adjustmentw

asconductedus-

ingA

1Casa

categoricalratherthana

con-tin

uou

svariab

le(to

allowfor

the

nonlinearassociation

between

A1C

andm

ortality),the

HR

inthose

with

aprior

diabetesdiagnosis

reducedto

2.44(95%

CI1.82–3.26).

CON

CLUSIO

NS

Beforediscussing

thefindings

ofthis

study,several

limitations

shouldbe

ac-know

ledged.The

main

limitation

isthe

lackofanthropom

etricdata

andinform

a-tion

onother

cardiovascularrisk

factors.The

shortfollow

-uptim

eof

thisstudy

meantthatw

ecould

notrestricttheanal-

ysesto

thecases

diagnosedat

leasttw

oyears

afterthe

A1C

testtoelim

inatesub-

jectsw

ithundetected

diseaseat

thetim

eof

theblood

test.Therefore,

we

cannotexclude

thepossibility

thatdiabetesatthetim

eof

theA

1Ctest

might

haveled

toelevated

glucoselevels.

We

conductedseparate

analysesfor

participantsw

itha

previousdiagnosis

ofdiabeteson

theba-

sisofhospitalizations

orgeneralpractice

prescriptionsforinsulin,anoralhypogly-

cemic,ora

home

glucose-monitoring

kit,butthere

stillmay

havebeen

some

partic-ipants

with

diagnoseddiabetes

thatwere

beingtreated

with

lifestyleinterventions

ratherthan

medication,

andthere

may

alsohave

beensom

eparticipantsw

ithun-

diagnoseddiabetes

(particularlythose

with

A1C

levels!

7.0%).A

swith

allstud-ies

ofthistype,there

may

havebeen

mis-

classificationof

specificcauses

ofdeath

becauseof

thew

ell-recognizedtendency

fordiabetes

tobe

underreportedas

acause

ofdeath(19,20).

Thecurrentstudy

was

basedon

par-ticipantsin

anintensive

population-basedhepatitis

Bscreening

program,and

these“volunteers”m

aynotbe

representativeof

thegeneral

population.H

owever,

we

havem

adeinternal

comparisons

(based

onA

1Clevels)w

ithinthe

groupthatpar-

ticipatedin

thescreening

program,and

itishighly

unlikelythatthese

internalcom-

parisonswould

bebiased

dueto

any“vol-

unteereffect.”

On

theother

hand,the

useof

A1C

levelsto

assessblood

glucoseconcentra-

tionsover

time

isone

ofthe

major

strengthsof

thisstudy.

Theselevels

arenot

subjectto

day-to-dayvariations,

asfasting

and2-h

oralglucose

tolerancetests

canbe

(1,2).A

secondstrength

ofthis

studyis

thelarge

numbers

involved,w

ith815

deathscom

paredw

ith521

inthe

studyof

Khaw

etal.

(10).This

hasprovided

relativelygood

power

toexam

-ine

cause-specificm

ortality,and

alsoto

examine

them

ortalityrisks

bysex,

eth-nicity,

andsm

okingstatus.

Afurther

strengthis

thatthe

studyis

comm

unitybased

ratherthan

beingbased

ona

se-lected

patientgroup.

Finally,an

addi-tional

strengthof

thestudy

isthe

likelynear-com

pleteascertainm

entofmortality

inthe

cohortusingnationalN

ewZealand

mortality

data.B

earing

these

limitatio

ns

and

strengthsinm

ind,thefindingsare

ofcon-siderable

interest.A

sexpected,

excessm

ortalityw

asevident

athigh

A1C

con-centrations

(!7.0%

),and

therew

asa

doseresponse

with

increasinglevel

ofA

1Cin

thosew

ithoutdiabetes.The

HRs

steadilyincreased

fromthe

A1C

referencecategory

tothe

highestcategory(!

7.0%;

HR

2.36[95%

CI

1.72–3.25]).This

isconsistent

with

theprevious

findingsof

Khaw

andcolleagues

(9,10),w

hoalso

foundincreasing

risksfor

totalmortality

throughoutthew

holerange

ofconcentra-tions,

includingthose

belowthe

thresh-old

comm

onlyaccepted

fordiabetes.

Inourstudy,a

1%increase

inA

1Clevelw

asassociated

with

a16%

increasein

mortal-

ityin

thosew

ithoutdiabetes,

compared

with

thefigure

of26%estim

atedby

Khaw

etal.(10).A

1Cw

asstrongly

associatedw

ithm

ortalityfrom

“endocrine,nutritional

andm

etabolicand

imm

unitydisorders.”

(The47

deathsin

thiscategory

included38

categorizedas

beingfrom

diabetes.)and

diseasesof

thecirculatory

system,

particularlyischem

icheartdisease.There

were

weaker

associationsw

ithdeaths

fromcancer

andother

andunknow

ncauses.The

weak

associationw

ithcancer

mortality

observedin

thecurrentanalyses

isconsistent

with

ourpreviously

pub-lished

findings(16)

forcancer

incidencein

thesam

ecohort.

A1C

levelshave

alsobeen

associated

Table 3—HRs for the association between A1C levels and mortality in a New Zealand population-based sample by cause of death

Site (ICD-9) n

A1C levels

!4.0% 4.0 to !5.0% 5.0 to !6.0% 6.0 to !7.0% !7.0% Prior diabetes diagnosis

n HR (95% CI)* n HR (95% CI)* n HR (95% CI)* n HR (95% CI)* n HR (95% CI)* n HR (95% CI)*

All deaths 815 3 2.90 (0.91–9.19) 82 1.0† 449 1.33 (1.05–1.70) 129 2.12 (1.58–2.85) 87 2.36 (1.72–3.25) 65 5.19 (3.67–7.35)All cancers (140–239) 262 0 ‡ 26 1.0† 154 1.10 (0.72–1.68) 44 1.50 (0.90–2.48) 23 1.29 (0.72–2.30) 15 2.35 (1.22–4.53)Endocrine, nutritional & metabolic,

and immunity disorders(240–279)

47 0 ‡ 0 ‡ 6 1.0† 2 1.79 (0.35–9.18) 19 27.17 (10.20–72.39) 20 90.36 (33.42–244.33)

Diseases of circulatory system (390–459) 280 1 3.95 (0.53–29.51) 20 1.0† 151 1.43 (0.89–2.30) 54 2.46 (1.44–4.19) 32 2.44 (1.37–4.35) 22 4.75 (2.53–8.92)Ischemic heart disease (410–414) 166 1 6.26 (0.81–48.13) 13 1.0† 91 1.27 (0.71–2.30) 26 1.75 (0.88–3.49) 23 2.55 (1.26–5.16) 12 3.89 (1.72–8.77)

Other and unknown causes 226 2 4.53 (1.09–18.85) 36 1.0† 138 1.41 (0.96–2.06) 29 2.52 (1.47–4.31) 13 1.86 (0.95–3.65) 8 3.88 (1.72–8.74)

*Adjusted for age, sex, ethnicity, and smoking status. †Reference category. ‡Not calculated due to zero deaths.

Brew

erand

Associates

DIA

BET

ES

CA

RE,

VO

LUM

E31,

NU

MBE

R6,JU

NE

20081147

ingin

thisgroup

(i.e.,it

didnot

always

increaseas

thelevel

ofA

1Cincreased);

theproportion

who

diedw

as12.5%

(5of

40)in

thosew

ithA

1Clevels

of5.0

to!

6.0%,

25.4%(18

of71)

inthose

with

levelsof6.0

to!

7.0%,and

14.3%(42

of293)in

thosew

ithlevelsof!

7.0%.O

ver-all,a

priordiabetesdiagnosiswasstrongly

associatedw

ithrisk

ofdeath(3.35

[2.58–

4.34];this

differsfrom

theestim

ateof

5.19given

inTable

2,w

hichinvolved

acom

parisonw

iththose

without

diabetesw

ithan

A1C

of4.0to

!5.0%

,whereasthe

comparison

hereis

with

allthosew

ithoutdiabetes),butthisrisk

reducedsom

ewhat

(2.46[1.86

–3.26])w

henadjusted

forA

1C.Ifthe

adjustmentw

asconductedus-

ingA

1Casa

categoricalratherthana

con-tin

uou

svariab

le(to

allowfor

the


between

A1C

andm

ortality),the

HR

inthose

with

aprior

diabetesdiagnosis

reducedto

2.44(95%

CI1.82–3.26).

CON

CLUSIO

NS

Beforediscussing

thefindings

ofthis

study,several

limitations

shouldbe

ac-know

ledged.The

main

limitation

isthe

lackofanthropom

etricdata

andinform

a-tion

onother

cardiovascularrisk

factors.The

shortfollow

-uptim

eof

thisstudy

meantthatw

ecould

notrestricttheanal-

ysesto

thecases

diagnosedat

leasttw

oyears

afterthe

A1C

testtoelim

inatesub-

jectsw

ithundetected

diseaseat

thetim

eof

theblood

test.Therefore,

we

cannotexclude

thepossibility

thatdiabetesatthetim

eof

theA

1Ctest

might

haveled

toelevated

glucoselevels.

We

conductedseparate

analysesfor

participantsw

itha

previousdiagnosis

ofdiabeteson

theba-

sisofhospitalizations

orgeneralpractice

prescriptionsforinsulin,anoralhypogly-

cemic,ora

home

glucose-monitoring

kit,butthere

stillmay

havebeen

some

partic-ipants

with

diagnoseddiabetes

thatwere

beingtreated

with

lifestyleinterventions

ratherthan

medication,

andthere

may

alsohave

beensom

eparticipantsw

ithun-

diagnoseddiabetes

(particularlythose

with

A1C

levels!

7.0%).A

swith

allstud-ies

ofthistype,there

may

havebeen

mis-

classificationof

specificcauses

ofdeath

becauseof

thew


fordiabetes

tobe

underreportedas

acause

ofdeath(19,20).

Thecurrentstudy

was

basedon

par-ticipantsin

anintensive

population-basedhepatitis

Bscreening

program,and

these“volunteers”m

aynotbe

representativeof

thegeneral

population.H

owever,

we

havem

adeinternal

comparisons

(based

onA

1Clevels)w

ithinthe

groupthatpar-

ticipatedin

thescreening

program,and

itishighly

unlikelythatthese

internalcom-

parisonswould

bebiased

dueto

any“vol-

unteereffect.”

On

theother

hand,the

useof

A1C

levelsto

assessblood

glucoseconcentra-

tionsover

time

isone

ofthe

major

strengthsof

thisstudy.

Theselevels

arenot

subjectto


asfasting

and2-h

oralglucose

tolerancetests

canbe

(1,2).A

secondstrength

ofthis

studyis

thelarge

numbers

involved,w

ith815

deathscom

paredw

ith521

inthe

studyof

Khaw

etal.

(10).This

hasprovided

relativelygood

power

toexam

-ine

cause-specificm

ortality,and

alsoto

examine

them

ortalityrisks

bysex,

eth-nicity,

andsm

okingstatus.

Afurther

strengthis

thatthe

studyis

comm

unitybased

ratherthan

beingbased

ona

se-lected

patientgroup.

Finally,an

addi-tional

strengthof

thestudy

isthe

likelynear-com

pleteascertainm

entofmortality

inthe


ewZealand

mortality

data.B

earing

these

limitatio

ns

and

strengthsinm

ind,thefindingsare

ofcon-siderable

interest.A

sexpected,

excessm

ortalityw

asevident

athigh

A1C

con-centrations

(!7.0%

),and

therew

asa

doseresponse

with

increasinglevel

ofA

1Cin

thosew

ithoutdiabetes.The

HRs

steadilyincreased

fromthe

A1C

referencecategory

tothe

highestcategory(!

7.0%;

HR

2.36[95%

CI

1.72–3.25]).This

isconsistent

with

theprevious

findingsof

Khaw

andcolleagues

(9,10),w

hoalso

foundincreasing

risksfor

totalmortality

throughoutthew

holerange

ofconcentra-tions,

includingthose

belowthe

thresh-old

comm

onlyaccepted

fordiabetes.

Inourstudy,a

1%increase

inA

1Clevelw

asassociated

with

a16%

increasein

mortal-

ityin

thosew

ithoutdiabetes,

compared

with

thefigure

of26%estim

atedby

Khaw

etal.(10).A

1Cw

asstrongly

associatedw

ithm

ortalityfrom

“endocrine,nutritional

andm

etabolicand

imm

unitydisorders.”

(The47

deathsin

thiscategory

included38

categorizedas

beingfrom

diabetes.)and

diseasesof

thecirculatory

system,

particularlyischem

icheartdisease.There

were

weaker

associationsw

ithdeaths

fromcancer

andother

andunknow

ncauses.The

weak

associationw

ithcancer

mortality

observedin

thecurrentanalyses

isconsistent

with

ourpreviously

pub-lished

findings(16)

forcancer

incidencein

thesam

ecohort.

A1C

levelshave

alsobeen

associated


Site (ICD-9) n

A1C levels





47 0 ‡ 0 ‡ 6 1.0† 2 1.79 (0.35–9.18) 19 27.17 (10.20–72.39) 20 90.36 (33.42–244.33)




Brew

erand

Associates

DIA

BET

ES

CA

RE,

VO

LUM

E31,

NU

MBE

R6,JU

NE

20081147

ingin

thisgroup

(i.e.,it

didnot

always

increaseas

thelevel

ofA

1Cincreased);

theproportion

who

diedw

as12.5%(5

of40)

inthose

with

A1C

levelsof

5.0to

!6.0%

,25.4%(18

of71)

inthose

with

levelsof6.0

to!

7.0%,and

14.3%(42

of293)in

thosew

ithlevelsof!

7.0%.O

ver-all,apriordiabetesdiagnosisw

asstronglyassociated

with

riskofdeath

(3.35[2.58

–4.34];

thisdiffers

fromthe

estimate

of5.19

givenin

Table2,w

hichinvolved

acom

parisonw

iththose

without

diabetesw

ithan

A1C

of4.0to

!5.0%

,whereasthe

comparison

hereisw

ithallthose

without

diabetes),butthisriskreduced

somew

hat(2.46

[1.86–3.26])

when

adjustedfor

A1C

.Iftheadjustm

entwasconducted

us-ing

A1C

asacategoricalratherthan

acon-

tinu

ous

variable(to

allowfor

the


between

A1C

andm

ortality),the

HR

inthose

with

aprior

diabetesdiagnosis

reducedto

2.44(95%

CI1.82–3.26).

CON

CLUSIO

NS

Beforediscussing

thefindings

ofthis

study,several

limitations

shouldbe

ac-know

ledged.The

main

limitation

isthe

lackofanthropom

etricdata

andinform

a-tion

onother

cardiovascularrisk

factors.The

shortfollow

-uptim

eof

thisstudy

meantthatw

ecould

notrestricttheanal-

ysesto

thecases

diagnosedat

leasttw

oyears

afterthe

A1C

testtoelim

inatesub-

jectsw

ithundetected

diseaseatthe

time

ofthe

bloodtest.

Therefore,w

ecannot

excludethe

possibilitythatdiabetesatthe

time

ofthe

A1C

testm

ighthave

ledto

elevatedglucose

levels.W

econducted

separateanalyses

forparticipants

with

aprevious

diagnosisofdiabetes

onthe

ba-sis

ofhospitalizationsor

generalpracticeprescriptionsforinsulin,an

oralhypogly-cem

ic,orahom

eglucose-m

onitoringkit,

buttherestillm

ayhave

beensom

epartic-

ipantsw

ithdiagnosed

diabetesthatw

erebeing

treatedw

ithlifestyle

interventionsrather

thanm

edication,and

therem

ayalso

havebeen

some

participantswith

un-diagnosed

diabetes(particularly

thosew

ithA

1Clevels

!7.0%

).Asw

ithallstud-

iesofthis

type,therem

ayhave

beenm

is-classification

ofspecific

causesof

deathbecause

ofthew


fordiabetes

tobe

underreportedas

acause

ofdeath(19,20).

Thecurrentstudy

was

basedon

par-ticipantsin

anintensivepopulation-based

hepatitisB

screeningprogram

,andthese

“volunteers”may

notberepresentative

ofthe

generalpopulation.

How

ever,w

ehave

made

internalcom

parisons(based

onA

1Clevels)w

ithinthe

groupthatpar-

ticipatedin

thescreening

program,and

itishighly

unlikelythatthese

internalcom-

parisonswould

bebiased

dueto

any“vol-

unteereffect.”

On

theother

hand,the

useof

A1C

levelsto

assessblood

glucoseconcentra-

tionsover

time

isone

ofthe

major

strengthsof

thisstudy.

Theselevels

arenot

subjectto


asfasting

and2-h

oralglucose

tolerancetests

canbe

(1,2).A

secondstrength

ofthis

studyis

thelarge

numbers

involved,w

ith815

deathscom

paredw

ith521

inthe

studyof

Khaw

etal.

(10).This

hasprovided

relativelygood

power

toexam

-ine

cause-specificm

ortality,and

alsoto

examine

them

ortalityrisks

bysex,

eth-nicity,

andsm

okingstatus.

Afurther

strengthis

thatthe

studyis

comm

unitybased

ratherthan

beingbased

ona

se-lected

patientgroup.

Finally,an

addi-tional

strengthof

thestudy

isthe

likelynear-com

pleteascertainm

entofmortality

inthe


ewZealand

mortality

data.B

earing

these

limitation

san

dstrengthsin

mind,the

findingsareofcon-

siderableinterest.

As

expected,excess

mortality

was

evidentat

highA

1Ccon-

centrations(!

7.0%),

andthere

was

adose

responsew

ithincreasing

levelof

A1C

inthose

without

diabetes.TheH

Rssteadily

increasedfrom

theA

1Creference

categoryto

thehighestcategory

(!7.0%

;H

R2.36

[95%C

I1.72–3.25]).

Thisis

consistentw

iththe

previousfindings

ofK

hawand

colleagues(9,10),

who

alsofound

increasingrisks

fortotalm

ortalitythroughoutthe

whole

rangeofconcentra-

tions,including

thosebelow

thethresh-

oldcom

monly

acceptedfor

diabetes.In

ourstudy,a1%

increasein

A1C

levelwas

associatedw

itha

16%increase

inm

ortal-ity

inthose

without

diabetes,com

paredw

iththe

figureof26%

estimated

byK

hawetal.(10).

A1C

was

stronglyassociated

with

mortality

from“endocrine,

nutritionaland

metabolic

andim

munity

disorders.”(The

47deaths

inthis

categoryincluded

38categorized

asbeing

fromdiabetes.)

anddiseases

ofthe

circulatorysystem

,particularly

ischemic

heartdisease.Therew

erew

eakerassociations

with

deathsfrom

cancerand

otherand

unknown

causes.Thew

eakassociation

with

cancerm

ortalityobserved

inthe

currentanalysesis

consistentw

ithour

previouslypub-

lishedfindings

(16)for

cancerincidence

inthe

same

cohort.A

1Clevels

havealso

beenassociated


Site (ICD-9) n

A1C levels





47 0 ‡ 0 ‡ 6 1.0† 2 1.79 (0.35–9.18) 19 27.17 (10.20–72.39) 20 90.36 (33.42–244.33)




Brewer

andA

ssociates

DIA

BETES

CA

RE,

VO

LUM

E31,

NU

MBER

6,JUN

E2008

1147

Brewer N, et al. Diabetes Care. 2008 Jun;31(6):1144-9

Carbohydrate Restriction in Diabetes

A recent review summarized the arguments for low-carbohydrate diets for the treatment of diabetes [28••].The major features include the following:

1. Carbohydrate restriction improves glycemic control,the primary target of nutritional therapy and reducesinsulin fluctuations [17, 29••], and references in [28••]and [30–32].

2. Low-carbohydrate diets are at least as effective as andgenerally more effective than low-fat diets for weightloss [19, 25–27, 33–35].

3. Carbohydrate-restriction is the requirement of the dietas described in definitions above. Although moststudies show limited increase in total fat, substitutionof any fat, even saturated fat, for carbohydrate is atleast neutral but generally beneficial for markers of andincidence of CVD.

4. Carbohydrate restriction ameliorates all of the featuresof metabolic syndrome, whereas higher carbohydratediets are associated with postprandial hyperglycemiaand hypertriglyceridemia [30–32].

5. The beneficial effects of carbohydrate restriction do notrequire weight loss [29••].

The review [28••] and a similar narrative review onmetabolic syndrome [31] provided a global perspective and alikely effective mode of treatment. As such, one would haveexpected at least a rebuttal as to why its promise should notbe exploited. Instead, the papers have been almost complete-ly ignored, again following the idea that low-carbohydratediets have their support in the popular media and thescientific literature can be ignored. The net effect is thatoptions available to patients are reduced.

Bottom Line: The Bigger Threat to CVD: Saturated Fator Hemoglobin A1c?

Design of diet for diabetes can probably be boiled down totwo considerations embodied in Figs. 1 and 3: which is thegreater threat to CVD in people with diabetes? Dietary fat,specifically saturated fat? Or glycemic control? Figure 3shows results from the UKPDS 35 (United KingdomProspective Diabetes Study) prospective study on theassociation of hemoglobin A1c (HbA1c) with macrovascularand microvascular complications of type 2 diabetes [36]. Indistinction to the effects of saturated fat shown in Fig. 1 thatare long range and conjectural at best, the effect of HbA1c

correlates strongly with immediate progression of diseaseand long-term outcome (more for microvascular complica-tions than macrovascular but even the latter has good

correlation >7% HbA1c). Recent results also show that thesource of plasma saturated fat, the presumed agent for anydeleterious effects, comes primarily from dietary carbohy-drate not from dietary saturated fat [26, 30].

The underlying mechanism is understood and dependson the anabolic effects of the hormone insulin [20, 30,37••]. As Smith [2] puts it: “In many ways, it is attractive tobelieve that manipulation of macronutrient compositionmight control body weight and improve health.” It isattractive because it is fundamental biochemistry [38, 39]and although metabolism is not considered a fad, it mayhave changed over the years.

Fig. 3 Hazard ratios with 95% confidence intervals as estimate ofassociation between hemoglobin A1c (HbA1c) concentration andmyocardial infarction or microvascular end points. (Modified fromStratton et al. [36])

Curr Diab Rep















Curr Diab Rep

Stratton IM, et al. BMJ 2000, 321(7258):405-412.















Curr Diab Rep

Stratton IM, et al. BMJ 2000, 321(7258):405-412.

BioMed Central

!"#$%&%'(%)!"#$%&'()*%+&',-&.,+&/0-#-0,'&"(+",1%12

Nutrition & Metabolism

Open AccessBrief communicationLow-carbohydrate diet in type 2 diabetes: stable improvement of bodyweight and glycemic control during 44 months follow-upJörgen V Nielsen* and Eva A Joensson

Address: Department of Medicine, Blekingesjukhuset, Karlshamn, 37480 Karlshamn, Sweden

Email: Jörgen V Nielsen* - [email protected]; Eva A Joensson - [email protected]* Corresponding author

AbstractBackground: Low-carbohydrate diets, due to their potent antihyperglycemic effect, are anintuitively attractive approach to the management of obese patients with type 2 diabetes. Wepreviously reported that a 20% carbohydrate diet was significantly superior to a 55–60%carbohydrate diet with regard to bodyweight and glycemic control in 2 groups of obese diabetespatients observed closely over 6 months (intervention group, n = 16; controls, n = 15) and wereported maintenance of these gains after 22 months. The present study documents the degree towhich these changes were preserved in the low-carbohydrate group after 44 months observationtime, without close follow-up. In addition, we assessed the performance of the two thirds ofcontrol patients from the high-carbohydrate diet group that had changed to a low-carbohydratediet after the initial 6 month observation period. We report cardiovascular outcome for the low-carbohydrate group as well as the control patients who did not change to a low-carbohydrate diet.

Method: Retrospective follow-up of previously studied subjects on a low carbohydrate diet.

Results: The mean bodyweight at the start of the initial study was 100.6 ± 14.7 kg. At six monthsit was 89.2 ± 14.3 kg. From 6 to 22 months, mean bodyweight had increased by 2.7 ± 4.2 kg to anaverage of 92.0 ± 14.0 kg. At 44 months average weight has increased from baseline g to 93.1 ±14.5 kg. Of the sixteen patients, five have retained or reduced bodyweight since the 22 month pointand all but one have lower weight at 44 months than at start. The initial mean HbA1c was 8.0 ±1.5%. After 6, 12 and 22 months, HbA1c was 6.1 ± 1.0%, 7.0 ± 1.3% and 6.9 ± 1.1% respectively.After 44 months mean HbA1c is 6.8 ± 1.3%.

Of the 23 patients who have used a low-carbohydrate diet and for whom we have long-term data,two have suffered a cardiovascular event while four of the six controls who never changed diethave suffered several cardiovascular events.

Conclusion: Advice to obese patients with type 2 diabetes to follow a 20% carbohydrate diet withsome caloric restriction has lasting effects on bodyweight and glycemic control.

BackgroundType 2 diabetes reflects a disturbance in the glucose-insu-lin axis of metabolism and has insulin resistance as a

defining feature. As such, it is expected that carbohydraterestriction would be the first line of attack and, in oneform or another, this was the primary approach before the

Published: 22 May 2008

Nutrition & Metabolism 2008, 5:14 doi:10.1186/1743-7075-5-14

Received: 24 January 2008Accepted: 22 May 2008

This article is available from: http://www.nutritionandmetabolism.com/content/5/1/14

© 2008 Nielsen and Joensson; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Nutrition & Metabolism 2008, 5:14 http://www.nutritionandmetabolism.com/content/5/1/14

Page 3 of 6(page number not for citation purposes)

ResultsTable 1 shows the measured parameters from start to 44months. Triglycerides (TG) and HbA1c, both parametersof adherence to the diet, were at their lowest after 3months. The adherence in the group then became lesspronounced, again reflected in triglycerides and HbA1c.

BodyweightThe mean reduction in bodyweight over the first sixmonths was 11.3 ± 4 kg (controls: 1.8 ± 3.8 kg). Tenpatients (62%) but none of the controls lost more than10% of bodyweight.

Mean bodyweight increased from 6 to 22 months by 2.7± 4.3 kg. The total mean increase from month 6 to 44 hasbeen 3.9 ± 5.6 kg. Five of the patients have maintainedbodyweight from 6 to 44 months or reduced it further (seefigure 1). However, five patients have increased meanbodyweight by 10 kg. In 7 patients (43%) the bodyweightis still 10% or more below their original weight.

HbA1cThe initial mean HbA1c in 2003 in the low-carbohydrategroup was 8.0 ± 1.5% (controls: 7.9 ± 1.5%). At the endof the 6 months study period it was 6.6 ± 1.0% (controls:7.3 ± 1.8%), and after 12 months it was 7.0 ± 1.3%. It hassince remained stable and is 6.8 ± 1.3% after 44 months.

The effect of carbohydrate lowering on blood glucose wasrapid. In the first week, mean fasting blood glucosedropped from 11.7 ± 3.3 mmol/l to 7.0 ± 1.4 mmol/lwhich necessitated corresponding reductions in medica-tions.

MedicationsAn important feature of carbohydrate restriction is thatreduction and even elimination of antidiabetic medica-tion normally is required in order to avoid hypoglycemia.

At the start of the study 15 of the 16 used metformin and5 sulfonylurea (SU). Eleven of the patients were treatedwith insulin at a mean daily dosage of 60 ± 33 IU.

All patients with SU reduced or discontinued it. Threepatients of 11 discontinued insulin and the average insu-lin requirement of the last 8 persons was 18 ± 11 IU/dayafter 6 months.

After 22 months 2 patients had resumed insulin treatmentfollowing an increase of carbohydrates. The mean insulin

Table 1: Effect of diet on weight, BMI, HbA1c and fasting lipids. Sixteen obese patients with type 2 diabetes started at month 0 on a diet with the proportions: 20% carbohydrates, 30% protein and 50% fat. The figures shown are means before, 3, 6, 22 and 44 months after the dietary change.

Month 0 3 P* 6 P* 22 44 P*

Weight (kg) 100.6 ± 14.7 91.9 ± 14.7 <0.001 89.2 ± 14.3 <0.001 92.0 ± 14.0 93.1 ± 14.5 <0.001BMI (kg/m2) 36.1 ± 4.2 33.0 ± 4.5 <0.001 32.0 ± 4.3 <0.001 32.9 ± 3.5 33.4 ± 3.9 <0.001HbA1c (%) 8.0 ± 1.5 5.9 ± 0.7 <0.001 6.6 ± 1.0 <0.001 6.9 ± 1.1 6.8 ± 1.3 <0.001Lipids(mmol/l)Tot-Chol. 5.6 ± 1.2 5.8 ± 1.1 0.4 6.1 ± 1.1 0.06 5.7 ± 1.2 5.4 ± 1.0 0.8HDL-Chol. 1.1 ± 0.2 1.2 ± 0.2 <0.002 1.3 ± 0.2 <0.001 1.3 ± 0.3 1.3 ± 0.2 <0.001§Triglycerides 1.4 (1;1.8) 1.2(0.8;1,4) 0.01 1.4(0.9;1.7) 0.4 1.4(1.2;1.9) 1.4(1.2–2) 0.9Chol/HDL 5.4 ± 1.5 5.0 ± 1.5 0.02 5.0 ± 1.7 0.07 4.6 ± 1.6 4.1 ± 0.9 <0.001§TG/HDL 1.4(0.9;1.7) 1.0(0.6;1.2) 0.003 1.0(0.7;1.5) 0.03 1.3(0.8;1.5) 1.1(0.9;1.7) 0.7

§given as medians with 25 and 75 percentiles. * p values are for differences from baseline.

Individual changes in bodyweight in 16 obese patients with type 2 diabetesFigure 1Individual changes in bodyweight in 16 obese patients with type 2 diabetes. The patients at start changed from a high-carbohydrate diet to a diet consisting of 20% carbohy-drates, 30% protein and 50% fat. The dotted red line is the mean weight.



discovery of insulin [1]. In addition, at least anecdotally,some degree of carbohydrate reduction is a component ofmuch clinical treatment. Health agencies have generallybeen reluctant to recommend carbohydrate restrictionalthough the recent American Diabetes Association guide-lines recognize that such diets are at least as effective aslow fat diets for weight loss [2] and, while not recom-mending low carbohydrate diets, recognizes that dietarycarbohydrate is the major factor in controlling blood glu-cose. Short term studies [3-7] in fact, demonstrate dra-matic improvements in glycemic control even in theabsence of weight loss [4].

Experience in our diabetes school showed that advice toreduce fat and increase carbohydrates had a very limitedeffect on long-term weight reduction in our obese diabe-tes patients (unpublished data). We therefore decided totest a different approach in an observational study with acontrol group. We were interested in seeing the effect ofthe diet in compliant patients, who could be expected toadhere to it 3–6 months. This enabled us to measure theactual effect of a carbohydrate-restricted diet with littlecontamination of non-compliant subjects. To this end, allpatients were well-informed of the diet and its rationalebefore they started. We considered a weight reduction of10% of bodyweight to be of clinical significance.

We have previously reported the results of these dietarychanges over 6 months in 16 obese type 2 diabetespatients with a control group. In 2003, the 16 wereadvised to lower their carbohydrate intake to 20% ofenergy. In the course of 6 months, they achieved signifi-cantly better control of hyperglycemia and bodyweightthan a control group of similar patients (n = 15) advisedto follow the official guidelines where 55% carbohydratesis recommended [5]. We have further reported that theimprovements were stable over 22 months [6].

We have now reviewed the clinical charts after 44 monthsand present the data for the intervention group withregard to glycemic control (HbA1c), bodyweight. Bodymass index (BMI = weight/m2) and lipids after 44 months.We also report the results for 7 patients from the controlswho immediately switched to a 20% carbohydrate dietand for whom long-term data is available.

MethodsThe method has previously been described in detail [5,6].In short, the 16 patients, all with BMI>30 kg/m2, free ofthyroid cardiac and renal disease – were advised to followa diet containing initially 1800 kcal for men and 1600kcal for women. The proportions of carbohydrates, fatand protein were 20%, 50% and 30% respectively. Thedaily quantity of carbohydrates was 80–90 g. The recom-mended carbohydrate consumption was limited to vege-

tables and salad. Instead of ordinary bread crisp/hardbread was recommended, each slice containing 3.5 to 8 gcarbohydrates.

Excluded were starch-rich bread, pasta, potatoes, rice andbreakfast cereals. The patients were counselled not to eatbetween meals. It was further recommended that theywalk 30 minutes a day and take a daily multivitamin sup-plement containing extra calcium. There was an introduc-tory meeting lasting most of one day. From day onediabetic medication was reduced by 25–30% to avoidhypoglycaemia. The patients monitored their own bloodglucose 4 times a day and were counselled by telephoneover the first few weeks for further reductions of medica-tions.

The subjects were followed closely for 6 months withgroup follow-ups every second week for the first 3 monthsand once a months for the next 3 months.

The 15 controls were advised on a diet with about thesame caloric content at an introductory meeting., Propor-tions of carbohydrates, fat and protein for this group were55–60%, 25–30% and 15% respectively. In the normaldiabetes diet whole-grain products are recommended.Generous helpings of vegetables and several servings offruits as snacks between meals are also recommended.

As a number of the controls attended our normal diabeteseducational course as introduction to the observationperiod, the control group on average received about 50%more attention – measured in hours – than the low-carbo-hydrate group. The controls were then followed in thesame way as the low-carbohydrate group.

Seven of the 15 controls switched to a 20% carbohydratediet immediately after the 6 months follow-up period. Forthose we have data 32–34 months after the change.

Three more controls sought information and attempted tochange diet later at various dates. The 5 remaining con-trols have not attempted a change of diet despite receivingadditional information.

All the patients were known to us and visited the diabetesnurse regularly after the initial period. The same scalesand laboratory were used for all measurements. Thepresent report is a review of clinical charts at about 44months after the start of the study in 2003. Where a figureis missing at 44 months we have taken the mean from thetwo closest figures. Means are given with standard devia-tions. T-test for dependent samples is used.

BioMed Central

!"#$%&%'(%)!"#$%&'()*%+&',-&.,+&/0-#-0,'&"(+",1%12


Open AccessResearchThe effect of a low-carbohydrate, ketogenic diet versus a low-glycemic index diet on glycemic control in type 2 diabetes mellitusEric C Westman*1, William S Yancy Jr1,2, John C Mavropoulos1, Megan Marquart1 and Jennifer R McDuffie1,2

Address: 1Department of Medicine, Duke University Medical Center, Durham, NC, USA and 2Center for Health Services Research in Primary Care, Department of Veterans' Affairs Medical Center, Durham, NC, USA

Email: Eric C Westman* - [email protected]; William S Yancy - [email protected]; John C Mavropoulos - [email protected]; Megan Marquart - [email protected]; Jennifer R McDuffie - [email protected]* Corresponding author

AbstractObjective: Dietary carbohydrate is the major determinant of postprandial glucose levels, andseveral clinical studies have shown that low-carbohydrate diets improve glycemic control. In thisstudy, we tested the hypothesis that a diet lower in carbohydrate would lead to greaterimprovement in glycemic control over a 24-week period in patients with obesity and type 2diabetes mellitus.

Research design and methods: Eighty-four community volunteers with obesity and type 2diabetes were randomized to either a low-carbohydrate, ketogenic diet (<20 g of carbohydratedaily; LCKD) or a low-glycemic, reduced-calorie diet (500 kcal/day deficit from weight maintenancediet; LGID). Both groups received group meetings, nutritional supplementation, and an exerciserecommendation. The main outcome was glycemic control, measured by hemoglobin A1c.

Results: Forty-nine (58.3%) participants completed the study. Both interventions led toimprovements in hemoglobin A1c, fasting glucose, fasting insulin, and weight loss. The LCKD grouphad greater improvements in hemoglobin A1c (-1.5% vs. -0.5%, p = 0.03), body weight (-11.1 kg vs.-6.9 kg, p = 0.008), and high density lipoprotein cholesterol (+5.6 mg/dL vs. 0 mg/dL, p < 0.001)compared to the LGID group. Diabetes medications were reduced or eliminated in 95.2% of LCKDvs. 62% of LGID participants (p < 0.01).

Conclusion: Dietary modification led to improvements in glycemic control and medicationreduction/elimination in motivated volunteers with type 2 diabetes. The diet lower in carbohydrateled to greater improvements in glycemic control, and more frequent medication reduction/elimination than the low glycemic index diet. Lifestyle modification using low carbohydrateinterventions is effective for improving and reversing type 2 diabetes.

Published: 19 December 2008


Received: 15 July 2008Accepted: 19 December 2008


© 2008 Westman et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.



assigned diet, 1 was unsatisfied with the diet, 2 were lostto follow-up, 3 were too busy, 1 relocated, 1 had difficultyadhering to the diet and 9 cited no reason. The baselinecharacteristics of study participants are shown in Table 1.There were no clinically significant differences betweenthe treatment groups.

Hemoglobin A1cFrom baseline to 24 weeks, the reduction of mean ± SDhemoglobin A1c was greater for the LCKD group (8.8 ±1.8% to 7.3 ± 1.5%, p = 0.009, within group change, n =21) than for the LGID group (8.3 ± 1.9% to 7.8 ± 2.1% p= NS, within group change, n = 29; between groups com-parison p = 0.03) (Table 2). The mean change in hemo-globin A1c for the LCKD group was -1.5% (95% CI: -2.30,-0.71), and for the LGID group was -0.5% (95%CI: -1.04,0.10). Using a theoretical probability matrix comparingthe change in hemoglobin A1c for each individual in onegroup to each individual in the other group, the probabil-ity of having a greater improvement in hemoglobin A1cwas 0.683 for being assigned to the LCKD group, com-pared to 0.300 for being in the LGID group (Figure 1)[26]. Fasting blood glucose and insulin improved simi-larly for both groups over the 24 weeks. In the LOCF anal-ysis, the mean hemoglobin A1c at baseline and week 24was 8.5% and 7.5% for the LCKD group, and 8.3% and8.0% for the LGID group (p = 0.02, between groups com-parison). In a multivariate linear regression model adjust-ing for weight change or BMI change, the between groupcomparison in change in hemoglobin A1c approached sta-tistical significance (p = 0.06). Additionally, there was nocorrelation between change in hemoglobin A1c andchange in weight (Figure 2).

Medication changesAt baseline, 22 (75.9%) of the LGID group were takinghypoglycemic medications (insulin only n = 3, oral agentsonly n = 19), and 20 (95.2%) of the LCKD group were tak-ing hypoglycemic medications (insulin + oral agents n =4, insulin only n = 4, oral agents only n = 12). Twenty of21 (95.2%) LCKD group participants had an elimination

or reduction in medication, compared with 18 of 29(62.1%) LGID group participants (p < 0.01). Table 3shows the changes in medication for those patients whowere taking insulin at baseline. Five individuals (4 in theLCKD group, 1 in the LGID group) who were taking over20 units of insulin at baseline were no longer taking insu-lin at the end of the study.

AdherencePrior to the study intervention, the mean ± SD dietaryintake for both groups was 2128 ± 993 kcal, 245 ± 136 gof carbohydrate (46% of daily energy intake), 86 ± 33 g ofprotein (18% of daily energy intake), 88 ± 57 g of fat(36% of daily energy intake). Over the 24-week durationof the intervention, the LCKD group consumed 1550 ±440 kcal per day, 49 ± 33 g of carbohydrate (13% of dailyenergy intake), 108 ± 33 g of protein (28% of daily energyintake), 101 ± 35 g of fat (59% of daily energy intake). Incomparison, the LGID group consumed 1335 ± 372 kcalper day, 149 ± 46 g of carbohydrate (44% of daily energyintake), 67 ± 20 g of protein (20% of daily energy intake),55 ± 23 g of fat (36% of daily energy intake). There was nodifference in self-reported exercise between the groups:the mean number of exercise sessions per week increasedfrom 2.0 ± 2.0 to 3.0 ± 2.0 for the LCKD group and from2.2 ± 2.2 to 3.8 ± 2.9 for the LGID group (p = 0.39 forcomparison).

Vital signsThere was significantly greater weight loss for the LCKDthan the LGID group over the 24 weeks: body weightdecreased from 108.4 ± 20.5 kg to 97.3 ± 17.6 kg for theLCKD group, and from 105.2 ± 19.8 to 98.3 ± 20.3 kg forthe LGID group (Table 2). Both groups had reductions insystolic blood pressure and diastolic blood pressure(Table 4).

Other metabolic effectsFor fasting lipid profiles, the LCKD group had an increasein HDL cholesterol (+12.7%), while the LGID group hadno change over the 24 weeks. All 7 parameters associated

Table 1: Baseline participant characteristics*

Characteristic Low -glycemic, reduced-calorie diet Low-carbohydrate, ketogenic dietEnrollees(n = 46)

Completers(n = 29)

Non-completers(n = 17)

Enrollees(n = 38)

Completers(n = 21)

Non-completers(n = 17)

Age, years 51.8 ± 7.8 50.0 ± 8.4 54.9 ± 5.7 51.8 ± 7.3 51.2 ± 6.1 52.4 ± 8.7Female gender, % 80.4 79.3 82.3 76.3 66.7 88.2White race, % 45.7 44.8 47.1 57.9 66.7 47.1African-American race, % 50 51.7 47.1 36.8 23.8 52.9College degree, % 58.7 68.9 41.2 57.9 61.9 52.9Body weight, kg 106.3 ± 20.1 105.2 ± 19.8 108.1 ± 20.9 105.5 ± 19.5 108.4 ± 20.5 101.9 ± 18.1Body mass index, kg/m2 38.5 ± 5.6 37.9 ± 6.0 39.4 ± 5.0 37.7 ± 6.1 37.8 ± 6.7 37.6 ± 5.3

* Values with plus/minus signs are means ± SD.



with the metabolic syndrome showed improvement forthe LCKD group; 5 of 7 improved for the LGID group(Table 4).

In terms of renal function, serum creatinine and calcu-lated GFR did not change significantly over the 24 weeksfor either group. There was a greater reduction in 24-hoururine protein for the LCKD group (baseline = 445 ± 1175mg/24 hour, week 24 = 296 ± 750 mg/24 hours, n = 18),

as compared with the LGID group (baseline = 276 ± 705mg/24 hour, week 24 = 223 ± 623 mg/24 hours, n = 24, p= 0.007 for between-groups comparison).

Adverse effectsThere were no statistically significant differences betweengroups in reported symptomatic adverse effects. The mostcommon symptoms experienced at any point during thestudy were headache (LCKD: 53.1%, LGID: 46.3%), con-

Table 2: Effect of diet programs on indices of glycemic control and body weight

Week 0 Week 12 Week 24 Week 0 to 24 Between Groups Between Groups Adjusted*mean ± sd mean ± sd mean ± sd mean change p value p value

LGID n = 29 n = 29 n = 29Hemoglobin A1c, % 8.3 ± 1.9 7.5 ± 1.7 7.8 ± 2.1 -0.5 0.03 0.06Fasting glucose, mg/dL 166.8 ± 63.7 140.7 ± 39.9 150.8 ± 47.4 -16.0** 0.67 0.76Fasting insulin, �U/mL 14.8 ± 6.9 13.9 ± 9.9 12.6 ± 6.5 -2.2** 0.10 0.84Body mass index, kg/m2 37.9 ± 6.0 36.5 ± 5.7 35.2 ± 6.1 -2.7** 0.05 0.10Body weight, kg 105.2 ± 19.8 101.0 ± 16.9 98.3 ± 20.3 -6.9** 0.008 0.01

LCKD n = 21 n = 21 n = 21Hemoglobin A1c, % 8.8 ± 1.8 7.2 ± 1.2 7.3 ± 1.5 -1.5**Fasting glucose, mg/dL 178.1 ± 72.9 156.4 ± 50.7 158.2 ± 50.0 -19.9**Fasting insulin, uU/mL 20.4 ± 9.3 14.3 ± 8.3 14.4 ± 6.9 -6.0**Body mass index, kg/m2 37.8 ± 6.7 34.4 ± 5.6 33.9 ± 5.8 -3.9**Body weight, kg 108.4 ± 20.5 100.1 ± 17.8 97.3 ± 17.6 -11.1**

LGID = low-glycemic, reduced-calorie diet; LCKD = low-carbohydrate, ketogenic diet* Adjusted for baseline values.** p < 0.05 for within-group change from Baseline to Week 24.

Payoff matrix for dietary comparisonsFigure 1Payoff matrix for dietary comparisons. Matrices show the theoretical paired comparison between the change in hemo-globin A1c for each individual in the LGI group compared with each individual in the LCKD group. In rank order across the top of the matrix, the change in hemoglobin A1c from baseline to week 24 is shown for the LCKD group; down the matrix side is shown the LGI group. Each matrix element shows the difference between the value for the LGI (row) and the LCKD (column) individual (LGI-LCKD). Positive values indicate greater reduction in hemoglobin A1c for LCKD, negative values indicate greater reduction in hemoglobin A1c for LGI. At the right of the Figure, the number of matrix elements in each category are divided by the total number of matrix elements (paired differences). LGI = Low glycemic index group, LCKD = Low carbohydrate ketogenic diet group, Prob = Probability.

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

!67 /-+, +$"( +$") +!"* +!"* +!") +!"$ !"& !") !"# !", !", $ $ $") $") $"( )"$ )"& )", *"& *"# -8/83+- &&# !")'%/-+4 +$"% +$"* +!"# +!"# +!"* +!"& !"$ !"& !"* !"' !"' !"% !"% $"& $"& $", ) )"$ )"' *"$ *"*/-+4 +$"% +$"* +!"# +!"# +!"* +!"& !"$ !"& !"* !"' !"' !"% !"% $"& $"& $", ) )"$ )"' *"$ *"* -+- $! !"!$'/-+4 +$"% +$"* +!"# +!"# +!"* +!"& !"$ !"& !"* !"' !"' !"% !"% $"& $"& $", ) )"$ )"' *"$ *"*/-+2 +&"$ +$"' +!", +!", +!"' +!"* +!"$ ! !"& !"* !"* !", !", $ $ $"# &"( &"% )"* )"% *"& /83+-8D<8- $)& !"&$,/-+2 +&"$ +$"' +!", +!", +!"' +!"* +!"$ ! !"& !"* !"* !", !", $ $ $"# &"( &"% )"* )"% *"& -+0/-+5 +&"& +$", +!"( +!"( +!", +!"# +!"& +!"$ !"$ !") !") !"' !"' !"% !"% $"* &", &"( )") )"( *"$ /3+-8D<8/,+- )% !"!'*/-+. +&") +$"( +!"% +!"% +!"( +!"' +!") +!"& ! !"& !"& !"# !"# !"( !"( $") &"' &", )"& )", */-+. +&") +$"( +!"% +!"% +!"( +!"' +!") +!"& ! !"& !"& !"# !"# !"( !"( $") &"' &", )"& )", * E8/8,+- $& !"!&/*+* +&"# +& +$"$ +$"$ +$ +!"( +!"# +!"* +!"& ! ! !") !") !"' !"' $"$ &"* &"# ) )"# )"(/*+1 +) +&"# +$"' +$"' +$"# +$") +$ +!"% +!", +!"# +!"# +!"& +!"& !"$ !"$ !"' $"% & &"# ) )")/*+2 +)"$ +&"' +$", +$", +$"' +$"* +$"$ +$ +!"( +!"' +!"' +!") +!") ! ! !"# $"( $"% &"* &"% )"&/0 +*"* +)"% +) +) +&"% +&", +&"* +&") +&"$ +$"% +$"% +$"' +$"' +$") +$") +!"( !"# !"' $"$ $"' $"%/0+5 +#"& +*", +)"( +)"( +)", +)"# +)"& +)"$ +&"% +&", +&", +&"* +&"* +&"$ +&"$ +$"' +!") +!"& !") !"( $"$/4 +'"* +#"% +# +# +*"% +*", +*"* +*") +*"$ +)"% +)"% +)"' +)"' +)") +)") +&"( +$"# +$"* +!"% +!"* +!"$

%&'( /-+,1.

BioMed Central

!"#$%&%'(%)!"#$%&'()*%+&',-&.,+&/0-#-0,'&"(+",1%12


Open AccessResearchThe effect of a low-carbohydrate, ketogenic diet versus a low-glycemic index diet on glycemic control in type 2 diabetes mellitusEric C Westman*1, William S Yancy Jr1,2, John C Mavropoulos1, Megan Marquart1 and Jennifer R McDuffie1,2

Address: 1Department of Medicine, Duke University Medical Center, Durham, NC, USA and 2Center for Health Services Research in Primary Care, Department of Veterans' Affairs Medical Center, Durham, NC, USA

Email: Eric C Westman* - [email protected]; William S Yancy - [email protected]; John C Mavropoulos - [email protected]; Megan Marquart - [email protected]; Jennifer R McDuffie - [email protected]* Corresponding author

AbstractObjective: Dietary carbohydrate is the major determinant of postprandial glucose levels, andseveral clinical studies have shown that low-carbohydrate diets improve glycemic control. In thisstudy, we tested the hypothesis that a diet lower in carbohydrate would lead to greaterimprovement in glycemic control over a 24-week period in patients with obesity and type 2diabetes mellitus.

Research design and methods: Eighty-four community volunteers with obesity and type 2diabetes were randomized to either a low-carbohydrate, ketogenic diet (<20 g of carbohydratedaily; LCKD) or a low-glycemic, reduced-calorie diet (500 kcal/day deficit from weight maintenancediet; LGID). Both groups received group meetings, nutritional supplementation, and an exerciserecommendation. The main outcome was glycemic control, measured by hemoglobin A1c.

Results: Forty-nine (58.3%) participants completed the study. Both interventions led toimprovements in hemoglobin A1c, fasting glucose, fasting insulin, and weight loss. The LCKD grouphad greater improvements in hemoglobin A1c (-1.5% vs. -0.5%, p = 0.03), body weight (-11.1 kg vs.-6.9 kg, p = 0.008), and high density lipoprotein cholesterol (+5.6 mg/dL vs. 0 mg/dL, p < 0.001)compared to the LGID group. Diabetes medications were reduced or eliminated in 95.2% of LCKDvs. 62% of LGID participants (p < 0.01).

Conclusion: Dietary modification led to improvements in glycemic control and medicationreduction/elimination in motivated volunteers with type 2 diabetes. The diet lower in carbohydrateled to greater improvements in glycemic control, and more frequent medication reduction/elimination than the low glycemic index diet. Lifestyle modification using low carbohydrateinterventions is effective for improving and reversing type 2 diabetes.



Received: 15 July 2008Accepted: 19 December 2008


© 2008 Westman et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.



on diabetes [27,28]. A reduced-glycemic index diet with-out weight loss can also lead to improvement in diabeticcontrol, with the magnitude of effect of a 0.43% reductionin hemoglobin A1c, when compared with higher-glycemicdiets of similar carbohydrate content [4]. The greater effectof the low-carbohydrate, ketogenic diet in this studyappeared to be due to the lower carbohydrate intake,because statistical significance remained after adjustmentfor weight loss. Because "low-glycemic" diets in previousstudies typically contain from 40–60% of calories fromcarbohydrate, it is possible that the beneficial effect of"low-glycemic" diets could be augmented by furtherreduction of the absolute amount of carbohydrate, or bya reduction in caloric content.

While this study was a treatment trial of individuals withtype 2 diabetes, lifestyle modification has been shown toprevent type 2 diabetes in the Diabetes Prevention Pro-gram (DPP). The intensive lifestyle modification arm ofthe DPP included a calorie- and fat-restricted diet with anenergy intake of 1380 kcal/day for women and 1583 kcal/day for men, and a percentage of energy from carbohy-drate of 54% [29]. While the effect was stronger than med-ication, the intensive lifestyle group developed diabetes ata rate of 20% after 4 years. Future research should includethe use of lower-carbohydrate diets for the treatment andprevention of type 2 diabetes.

Like previous studies, we found that the LCKD led toweight reduction, improvement in glycemic control, andelevation in HDL-cholesterol, but no deterioration in fast-ing lipid parameters. Extending these findings, weobserved that all metabolic syndrome components wereimproved by the LCKD [30]. It is interesting to note thatthe LGID group reported consuming fewer calories thanthe LCKD group, yet had less weight loss. This may reflectproblems with the diet data as collected, issues with differ-ential physical activity, or metabolic inefficiency (leadingto increased energy expenditure) which may occur duringthe consumption of a carbohydrate-restricted diet.

Limitations of this study include the lack of blinding ofphysicians and outcome assessors to treatment group, andthe use of food records. The study participants were com-munity volunteers, and predominantly women, whichmay limit generalization of these findings to clinical pop-ulations and men. The analysis and presentation of onlydetailed food records may bias the estimate of foodintake. We chose the "completer analysis" as the primaryoutcome because we were interested in answering thequestion of what might be expected from patients whocan adhere to the intervention. The LOCF analysis mightgeneralize better to a population of patients who have dif-ferent food preferences from their assigned diet, who lose/lack motivation, or who experience other barriers to die-tary change. Another possible limitation is the baselineimbalance in the primary outcome, HgA1c, which

Table 4: Effect of diet programs on metabolic syndrome parameters and fasting lipid profiles

Low glycemic, reduced-calorie diet group (n = 29) Low carbohydrate, ketogenic diet group (n = 21)

Test Week 0 Week 24 Week 0 to 24 Week 0 Week 24 Week 0 to 24

mean ± sd mean ± sd mean change mean ± sd mean ± sd mean change

Fasting glucose, mg/dL 166.8 ± 63.7 150.8 ± 47.4 -16.0 * 178.1 ± 72.9 158.2 ± 50.0 -19.9*Waist circumference, inches 47.0 ± 5.1 42.4 ± 5.5 -4.6 * 47.1 ± 5.5 41.8 ± 5.3 -5.3 *Triglycerides, mg/dL 167.1 ± 125.7 147.8 ± 128.5 -19.3 210.4 ± 10.3 142.9 ± 76.9 -67.5 *HDL cholesterol, mg/dL 48.7 ± 11.8 48.7 ± 10.1 -0 † 44.0 ± 8.7 49.6 ± 11.7 +5.6 * †Systolic blood pressure, mmHg 140.8 ± 15.7 130.1 ± 17.1 -10.7 * 144.4 ± 15.0 127.8 ± 13.4 -16.6 *Diastolic blood pressure, mmHg 84.1 ± 11.0 78.5 ± 8.7 -5.6 * 83.9 ± 10.3 75.8 ± 10.9 -8.1 *Body mass index, kg/m2 37.9 ± 6.0 35.2 ± 6.1 -2.7 * † 37.8 ± 6.7 33.9 ± 5.8 -3.9 * †Total cholesterol, mg/dL 190.6 ± 43.8 184.8 ± 45.6 -5.8 191.4 ± 32.0 187.0 ± 35.8 -4.4LDL cholesterol, mg/dL 113.8 ± 40.9 111.0 ± 42.2 -2.8 105.8 ± 25.7 107.1 ± 26.3 +1.3VLDL cholesterol, mg/dL 27.7 ± 13.2 24.4 ± 12.3 -3.3* 37.3 ± 14.9 27.3 ± 15.2 -10.0*Total cholesterol/HDL cholesterol ratio 4.1 ± 1.3 3.9 ± 1.2 -0.2 4.5 ± 1.1 4.1 ± 4.1 -0.4Triglyceride/HDL cholesterol ratio 3.9 ± 3.7 3.3 ± 3.1 -0.6 5.2 ± 3.4 3.4 ± 3.0 -1.8*

These changes were observed with a reduction or elimination of diabetic medication as shown in Table 3.HDL = high-density lipoprotein; LDL = low-density lipoprotein; VLDL = very-low-density lipoprotein* p < 0.05 for within-group change from Baseline to Week 24.† p < 0.05, for between-groups comparison of changes from Baseline to Week 24.P values with adjustment for baseline values: fasting glucose: 0.76, waist circumference: 0.43, triglycerides: 0.17, HDL: 0.09, systolic blood pressure: 0.89, diastolic blood pressure: 0.48, body mass index: 0.10, total cholesterol: 0.85, LDL: 0.79, VLDL: 0.24, total cholesterol/HDL ratio: 0.92, triglyceride/HDL ratio: 0.54.



with the metabolic syndrome showed improvement forthe LCKD group; 5 of 7 improved for the LGID group(Table 4).

In terms of renal function, serum creatinine and calcu-lated GFR did not change significantly over the 24 weeksfor either group. There was a greater reduction in 24-hoururine protein for the LCKD group (baseline = 445 ± 1175mg/24 hour, week 24 = 296 ± 750 mg/24 hours, n = 18),

as compared with the LGID group (baseline = 276 ± 705mg/24 hour, week 24 = 223 ± 623 mg/24 hours, n = 24, p= 0.007 for between-groups comparison).

Adverse effectsThere were no statistically significant differences betweengroups in reported symptomatic adverse effects. The mostcommon symptoms experienced at any point during thestudy were headache (LCKD: 53.1%, LGID: 46.3%), con-

Table 2: Effect of diet programs on indices of glycemic control and body weight

Week 0 Week 12 Week 24 Week 0 to 24 Between Groups Between Groups Adjusted*mean ± sd mean ± sd mean ± sd mean change p value p value

LGID n = 29 n = 29 n = 29Hemoglobin A1c, % 8.3 ± 1.9 7.5 ± 1.7 7.8 ± 2.1 -0.5 0.03 0.06Fasting glucose, mg/dL 166.8 ± 63.7 140.7 ± 39.9 150.8 ± 47.4 -16.0** 0.67 0.76Fasting insulin, �U/mL 14.8 ± 6.9 13.9 ± 9.9 12.6 ± 6.5 -2.2** 0.10 0.84Body mass index, kg/m2 37.9 ± 6.0 36.5 ± 5.7 35.2 ± 6.1 -2.7** 0.05 0.10Body weight, kg 105.2 ± 19.8 101.0 ± 16.9 98.3 ± 20.3 -6.9** 0.008 0.01

LCKD n = 21 n = 21 n = 21Hemoglobin A1c, % 8.8 ± 1.8 7.2 ± 1.2 7.3 ± 1.5 -1.5**Fasting glucose, mg/dL 178.1 ± 72.9 156.4 ± 50.7 158.2 ± 50.0 -19.9**Fasting insulin, uU/mL 20.4 ± 9.3 14.3 ± 8.3 14.4 ± 6.9 -6.0**Body mass index, kg/m2 37.8 ± 6.7 34.4 ± 5.6 33.9 ± 5.8 -3.9**Body weight, kg 108.4 ± 20.5 100.1 ± 17.8 97.3 ± 17.6 -11.1**

LGID = low-glycemic, reduced-calorie diet; LCKD = low-carbohydrate, ketogenic diet* Adjusted for baseline values.** p < 0.05 for within-group change from Baseline to Week 24.

Payoff matrix for dietary comparisonsFigure 1Payoff matrix for dietary comparisons. Matrices show the theoretical paired comparison between the change in hemo-globin A1c for each individual in the LGI group compared with each individual in the LCKD group. In rank order across the top of the matrix, the change in hemoglobin A1c from baseline to week 24 is shown for the LCKD group; down the matrix side is shown the LGI group. Each matrix element shows the difference between the value for the LGI (row) and the LCKD (column) individual (LGI-LCKD). Positive values indicate greater reduction in hemoglobin A1c for LCKD, negative values indicate greater reduction in hemoglobin A1c for LGI. At the right of the Figure, the number of matrix elements in each category are divided by the total number of matrix elements (paired differences). LGI = Low glycemic index group, LCKD = Low carbohydrate ketogenic diet group, Prob = Probability.

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

!67 /-+, +$"( +$") +!"* +!"* +!") +!"$ !"& !") !"# !", !", $ $ $") $") $"( )"$ )"& )", *"& *"# -8/83+- &&# !")'%/-+4 +$"% +$"* +!"# +!"# +!"* +!"& !"$ !"& !"* !"' !"' !"% !"% $"& $"& $", ) )"$ )"' *"$ *"*/-+4 +$"% +$"* +!"# +!"# +!"* +!"& !"$ !"& !"* !"' !"' !"% !"% $"& $"& $", ) )"$ )"' *"$ *"* -+- $! !"!$'/-+4 +$"% +$"* +!"# +!"# +!"* +!"& !"$ !"& !"* !"' !"' !"% !"% $"& $"& $", ) )"$ )"' *"$ *"*/-+2 +&"$ +$"' +!", +!", +!"' +!"* +!"$ ! !"& !"* !"* !", !", $ $ $"# &"( &"% )"* )"% *"& /83+-8D<8- $)& !"&$,/-+2 +&"$ +$"' +!", +!", +!"' +!"* +!"$ ! !"& !"* !"* !", !", $ $ $"# &"( &"% )"* )"% *"& -+0/-+5 +&"& +$", +!"( +!"( +!", +!"# +!"& +!"$ !"$ !") !") !"' !"' !"% !"% $"* &", &"( )") )"( *"$ /3+-8D<8/,+- )% !"!'*/-+. +&") +$"( +!"% +!"% +!"( +!"' +!") +!"& ! !"& !"& !"# !"# !"( !"( $") &"' &", )"& )", */-+. +&") +$"( +!"% +!"% +!"( +!"' +!") +!"& ! !"& !"& !"# !"# !"( !"( $") &"' &", )"& )", * E8/8,+- $& !"!&/*+* +&"# +& +$"$ +$"$ +$ +!"( +!"# +!"* +!"& ! ! !") !") !"' !"' $"$ &"* &"# ) )"# )"(/*+1 +) +&"# +$"' +$"' +$"# +$") +$ +!"% +!", +!"# +!"# +!"& +!"& !"$ !"$ !"' $"% & &"# ) )")/*+2 +)"$ +&"' +$", +$", +$"' +$"* +$"$ +$ +!"( +!"' +!"' +!") +!") ! ! !"# $"( $"% &"* &"% )"&/0 +*"* +)"% +) +) +&"% +&", +&"* +&") +&"$ +$"% +$"% +$"' +$"' +$") +$") +!"( !"# !"' $"$ $"' $"%/0+5 +#"& +*", +)"( +)"( +)", +)"# +)"& +)"$ +&"% +&", +&", +&"* +&"* +&"$ +&"$ +$"' +!") +!"& !") !"( $"$/4 +'"* +#"% +# +# +*"% +*", +*"* +*") +*"$ +)"% +)"% +)"' +)"' +)") +)") +&"( +$"# +$"* +!"% +!"* +!"$

%&'( /-+,1.

pared a low GI diet with a standard high cereal diet. At thesame time, Westman et al. [17] published a comparison of alow GI diet with a true low-carbohydrate diet. Thegenerally superior performance of the low-carbohydratediet to the cereal diet and both versions of the low GI dietare evident in Fig. 2.

The problem may be historical. Carbohydrate restric-tion was the established treatment for diabetes in thepreinsulin era [18]. Because the predominant form ofdiabetes in this era was type 1, it is likely that thediscovery of insulin cast diabetes as a deficiency diseasethat had to be treated with administered insulin or otherdrugs. The persistence of this view of diabetes may haverepressed the understanding of diabetes as a metabolicdisease, which might be more relevant to type 2 diabetes,the prevalent form today. The latter approach wouldpredict the value of carbohydrate reduction. The evolutionof the low-fat diet–heart hypothesis, however, solidifiedthe high-carbohydrate fad.

Fig. 1 Associations betweensaturated fat intake in relationto coronary heart disease andstroke. IV—inverse variance;SAT—saturated fat intake.(Modified from Fung et al. [6])

Fig. 2 Comparison of cereal-based diets and low-GI diets of Jenkins etal. [16] and comparison of low-GI diets with low-carbohydrate diets ofWestman et al. [17]. Changes are shown as percentage for the indicatedparameters. C—cholesterol; CHO—carbohydrate; GI—glycemic index;HbA1c—hemoglobin A1c; HDL—high-density lipoprotein; LDL—low-density lipoprotein; TG—triglyceride

Curr Diab Rep

Feinman RD. Fad diets in the treatment of diabetes. Curr Diab Rep. 2011 Apr;11(2):128-35




which was anticipated to be deleterious was, in fact, ben-eficial compared to fat reduction. Studies of CR continueto show improvements in atherogenic dyslipidemia[5,8,9] and there is an evolving picture that the effects ofCR, notably in lowering insulin and thereby changingmetabolic regulation of lipid, may be more importantthan the total amount of lipid substrate.

Atherogenic dyslipidemiaThe level of LDL cholesterol is generally considered themost clinically useful marker for cardiovascular disease(CVD) and there seems little question that reduction inLDL, especially if effected by administration of statins, isaccompanied by substantial reduction in the risk of cardi-ovascular events. The importance of LDL as a primarymarker, however, must be tempered by observations thatLDL is not homogeneous and atherogenicity appears to bea function of particle size: small dense LDL particles aremore atherogenic [10]. In addition, other risk factors,high triglyceride (TAG), low HDL, and insulin resistanceare frequently increased under conditions that lower LDL[8]. The atherogenicity of a greater number of small LDLparticles is reflected in an increase in apolipoprotein B(apoB) since each atherogenic lipoprotein particle con-tains one molecule of apoB and total LDL would biasresults towards lower risk. Barter et al. [11] summarizedevidence that apoB is a more reliable indicator of risk thanLDL and further that the apoB/apoA-I ratio is superior toconventional cholesterol ratios. Finally, circulating TAG isof considerable importance in this discussion because it ismechanistically linked to the formation of atherogenicparticles, and is highly sensitive to dietary manipulation.

Study synopsisDetails of the dietary intervention of Krauss et al. [1] areshown in Table 1. Diets were consumed over 3 sequentialperiods in which energy was prescribed to achieve weightmaintenance (3 wk), weight loss (-1000 kcal/day; 5 wk),and weight stabilization (4 wk). Blood measures werereported at baseline, after weight maintenance and afterweight loss + stabilization. For comparison, the protocolfrom Sharman, et al. [12] for six weeks on a very low car-bohydrate ketogenic diet (VLCKD) designed for weightmaintenance are also shown in Table 1.

Figures 1, 2, 3, 4, 5 show the results of sequential changein macronutrient composition and caloric restriction.

Effect of dietary interventions on reduction in TAGFigure 1Effect of dietary interventions on reduction in TAG. Solid Lines : Data from Reference [1] was converted from reported log values in their Table 2 and per cent of baseline calculated. At week 3, a 1000 kcal reduction in energy was implemented and at week 9, dieters were put on mainte-nance diet. Combined effect of calorie reduction and mainte-nance are reported at week 12. Dashed line: Data from reference [12]: A eucaloric ketogenic diet was instituted for six weeks (no weight loss phase). Points were recorded at week 3 and 6.

Table 1: Macronutrient composition of diet. Diet composition from the indicated references.

Group % CHO % FAT % Protein % SF

Krauss – Ref. [1]LF 54 30 16 7CR (39 % CHO) 39 29 29 8CR (26 % CHO) 26 46 29 9CR (26 % CHO + SF) 26 45 29 15

Sharman – Ref [12]Low Carbohydrate Ketogenic (LCKD)

8 61 30 11

LF (Control) 59 25 15 8

BioMed Central



Open AccessPerspectiveLow carbohydrate diets improve atherogenic dyslipidemia even in the absence of weight lossRichard D Feinman*1 and Jeff S Volek2

Address: 1Department of Biochemistry, SUNY Downstate Medical Center, Brooklyn, NY 11203, USA and 2Human Performance Laboratory, Department of Kinesiology, University of Connecticut, Storrs, CT 06269-1110, USA

Email: Richard D Feinman* - [email protected]; Jeff S Volek - [email protected]* Corresponding author

AbstractBecause of its effect on insulin, carbohydrate restriction is one of the obvious dietary choices forweight reduction and diabetes. Such interventions generally lead to higher levels of dietary fat thanofficial recommendations and have long been criticized because of potential effects oncardiovascular risk although many literature reports have shown that they are actually protectiveeven in the absence of weight loss. A recent report of Krauss et al. (AJCN, 2006) separates theeffects of weight loss and carbohydrate restriction. They clearly confirm that carbohydraterestriction leads to an improvement in atherogenic lipid states in the absence of weight loss or inthe presence of higher saturated fat. In distinction, low fat diets seem to require weight loss foreffective improvement in atherogenic dyslipidemia.

BackgroundThe recent report of Krauss et al. [1] highlights the ratherdramatic differences in the effects of carbohydraterestricted (CR) and low fat (LF) diets on the lipid changesthat may predispose to atherosclerosis. By first imple-menting weight maintenance diets of different composi-tions followed by calorie reduction, the authors show thatsignificant improvement in atherogenic lipid states (lowerTAG, higher HDL, lower apoB/apoA-1) can be broughtabout by CR even in the absence of weight loss or in thepresence of higher saturated fat. When weight loss was fur-ther implemented in the CR groups, there was little fur-ther improvement in most markers although HDLcontinued to increase on calorie reduction. The LF diet, indistinction, required weight loss for effective improve-ment in the lipid profile, and the additive outcome of dietchange and calorie reduction were not as effective as in theCR diets. These results have obvious implications forchoice of diets and represents a philosophical reversal of

the practical implications of macronutrient composition.Criticism of the use of CR for weight loss has traditionallyfocused on the potential effect on risk of CVD because ofthe substitution of fat for carbohydrate. It is now clear thatsuch a change is beneficial and the demonstration thatactual weight loss is not required for the benefits of CRsuggests the need for reevaluation of current guidelines.

In general, reports on the effects of CR diets continue todefy conventional wisdom. While most official agenciesrecommend limiting dietary fat or at least saturated fat,experimental data show that replacement of fat, even sat-urated fat with carbohydrate is deleterious to markers foratherogenic dyslipidemia [2]. Although several studies inthe literature had pointed to the value of replacing dietarycarbohydrate with fat or protein [3-6], an historical turn-ing point in the reappraisal of CR diets might be consid-ered the study of Foster et al. [7] comparing the Atkins dietwith a LF diet. The results were unexpected in that CR,

Published: 21 June 2006


Received: 09 June 2006Accepted: 21 June 2006


© 2006 Feinman and Volek; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

BioMed Central















BioMed Central

















Overall, the data show a significantly greater improve-ment in lipid profile as carbohydrate is reduced even ifthis change is not accompanied by caloric reduction andeven in the presence of relatively high saturated fat, inagreement with results of Sharman, et al. [12]. Data fromSharman's study is indicated by dotted lines in Figures 2and 5. When caloric restriction is introduced after theseweight stable changes, there is little additional change inmost parameters on the CR diets, but the additive effectsof the CR and caloric reduction are greater than the addi-tive effects of fat reduction followed by caloric restriction(e.g., compare week Wk 12 values for the 56% CHO to the26% CHO + SF diets in Figures 1, 2, 3, 4, 5).

Mechanisms and the separation of carbohydrate restriction and weight lossThe striking data tabulated by Krauss, et al. [1] and pre-sented in graphic form in Figures 1, 2, 3, 4, 5 are some-what at odds with their stated conclusion: "Moderatecarbohydrate restriction and weight loss provide equiva-lent but non-additive approaches to improving athero-genic dyslipidemia" and "the beneficial lipid changesresulting from a reduced carbohydrate intake were not sig-nificant after weight loss."

It is not clear what is meant by additive since the com-bined effects due to macronutrient change and caloricrestriction are not compared to experiments where theyare implemented together which are known to providelarge positive effects. In any case, at least HDL values doshow significant increases in both phases of the experi-ment, positive for low carbohydrates, initially negativeand then positive for low fat and there is a pronouncedadditive effect (Figure 2).

Compared to weight loss on a LF diet, the high saturatedfat CR diet with no weight loss resulted in better improve-ments in LDL peak size, TAG, HDL, and the ratios totalcholesterol/HDL and apoB/ApoA-1, that is, the effects arenot equivalent; CR is significantly better than weight lossin the presence of LF for atherogenic dyslipidemia. Thefact that these effects are not equivalent is further shownin Figures 1, 2, 3, 4, 5, where the combined (weight stableand weight loss) effects for LF are not as great as for CR.The results suggest that if, at week 12, a 26 % CR with sat-urated fat were instituted for the LF group, furtherimprovement in lipid profile would be brought about.

Krauss et al. [1] make the underlying assumption thatweight loss is the same whether caused by caloric restric-tion in the presence of low carbohydrate or low fat buttheir data show that this is not so. It seems that they also

Change in the ratio of total cholesterol to HDL with dietFigure 3Change in the ratio of total cholesterol to HDL with diet. The effect is largely due to HDL increases (Figure 2) since the total changes in LDL were -11.5, -1.8, -6.9 for decreasing carbohydrate. LDL for 26% CHO + SF changed little (+0.4).

Change in HDL with dietFigure 2Change in HDL with diet. Data from reference [1]. Extrapolated lines are drawn to indicate that there is a greater change during the weight loss phase on low carbohy-drate diets with or without saturated fat than on the low fat diet, that is, carbohydrate restriction improves HDL during the macronutrient change and also additively during weight loss.

BioMed Central















BioMed Central

















assume that weight loss is a cause not an effect (due to cal-orie reduction). Whereas this may play a role, it is reason-able to assume that improvement in dyslipidemia andweight loss are parallel responses to calorie reductionwhich is the major physiologic stimulus. It is obvious thatthe reason lipid changes are not brought about in Krauss'sexperiment by weight loss in the low carbohydrate arms isthat the lipid markers have already changed drastically.Finally, it should be pointed out that the reduction in cal-ories that is effective in the weight loss phase of the low fatarm included a substantial reduction in carbohydrate asone component.

SummaryAlthough some effort is required to disentangle the dataand interpretation, the recent publication from Krauss etal. [1] should be recognized as a breakthrough. Their find-ings, presented in Figures 1, 2, 3, 4, 5 make it clear that thesalutary effects of CR on dyslipidemia do not requireweight loss, a benefit that is not a feature of strategiesbased on fat reduction. As such, Krauss et al. [1] providesone of the strongest arguments to date for CR as a funda-mental approach to diet, especially for treating athero-genic dyslipidemia.

Abbreviationsapo: apolipoprotein, CR: carbohydrate restricted diet,CVD: cardiovascular disease, HDL: high density lipopro-tein cholesterol, LDL: low density lipoprotein cholesterol,LF: low fat diet, SF: saturated fat, TAG: triacylglycerol (trig-lyceride)

Competing interestsThe author(s) declare that they have no competing inter-ests.

References1. Krauss RM, Blanche PJ, Rawlings RS, Fernstrom HS, Williams PT:

Separate effects of reduced carbohydrate intake and weightloss on atherogenic dyslipidemia. Am J Clin Nutr 2006,83(5):1025-1031.

2. Mensink RP, Zock PL, Kester AD, Katan MB: Effects of dietaryfatty acids and carbohydrates on the ratio of serum total toHDL cholesterol and on serum lipids and apolipoproteins: ameta-analysis of 60 controlled trials. Am J Clin Nutr 2003,77(5):1146-1155.

3. Volek JS, Gomez AL, Kraemer WJ: Fasting lipoprotein and post-prandial triacylglycerol responses to a low-carbohydrate dietsupplemented with n-3 fatty acids. J Am Coll Nutr 2000,19(3):383-391.

4. Volek JS, Westman EC: Very-low-carbohydrate weight-lossdiets revisited. Cleve Clin J Med 2002, 69(11):849-853.

5. Westman EC, Mavropoulos J, Yancy WS, Volek JS: A review of low-carbohydrate ketogenic diets. Curr Atheroscler Rep 2003,5(6):476-483.

6. Westman EC, Yancy WS, Edman JS, Tomlin KF, Perkins CE: Effect of6-month adherence to a very low carbohydrate diet pro-gram. Am J Med 2002, 113(1):30-36.

7. Foster GD, Wyatt HR, Hill JO, McGuckin BG, Brill C, Mohammed BS,Szapary PO, Rader DJ, Edman JS, Klein S: A randomized trial of alow-carbohydrate diet for obesity. N Engl J Med 2003,348(21):2082-2090.

8. Volek JS, Feinman RD: Carbohydrate restriction improves thefeatures of Metabolic Syndrome. Metabolic Syndrome maybe defined by the response to carbohydrate restriction. NutrMetab (Lond) 2005, 2:31.

9. Volek JS, Sharman MJ, Forsythe CE: Modification of lipoproteinsby very low-carbohydrate diets. J Nutr 2005, 135(6):1339-1342.

10. Krauss RM: Heterogeneity of plasma low-density lipoproteinsand atherosclerosis risk. Curr Opin Lipidol 1994, 5(5):339-349.

Change in peak diameter of LDL particlesFigure 5Change in peak diameter of LDL particles. Dotted line are data from reference [12] for the effect of LCKD or the subset of effect of LCKD for the population of subjects with high levels of pattern B (small dense LDL particles).

Change in the ratio of apoB:apoA-1 with dietFigure 4Change in the ratio of apoB:apoA-1 with diet.

BioMed Central















v. 47 – no.2 – abr./jun. 2010 Arq Gastroenterol 165

ARTIG

O ORIG

INAL /

ORIGI

NAL A

RTICL

E ARQGA/1482

INSULIN RESISTANCE INDEX (HOMA-IR) IN THE DIFFERENTIATION OF PATIENTS WITH NON-ALCOHOLIC FATTY LIVER DISEASE AND HEALTHY INDIVIDUALS

Ana Lúcia Farias de Azevedo SALGADO1, Luciana de CARVALHO1, Ana Claudia OLIVEIRA1, Virgínia Nascimento dos SANTOS1, Jose Gilberto VIEIRA2 and Edison Roberto PARISE1,

ABSTRACT – Context - Due to its good correlation to glycemic clamp, HOMA-IR has been widely utilized as insulin resistance index in clinical and epidemiological studies involving non-alcoholic fatty liver disease carriers. However, values used for this parameter have shown large variability. Objective – To identify the HOMA-IR cut value that best distinguishes non-diabetic non-alcoholic fatty liver disease patients from a control group. Methods - One hundred sixteen non-alcoholic fatty liver disease patients were studied, diagnosed by clinical, biochemical, and liver image or biopsy criteria, and 88 healthy individuals, without any liver disease and testing for oral glucose tolerance within normality. These groups did not differ in age and gender. All were submitted to oral glucose tolerance test and blood samples were collected for glucose and insulin measurements by immuno!uorometric method. HOMA-IR was calculated according to the formula: fasting insulin (�U/L) x fasting glucose (nmol/L)/22.5. Results - NAFLD patients showed higher insulin, glycemia, and HOMA-IR values than control group, even when excluding glucose intolerant and diabetes mellitus patients by their glycemic curves. HOMA-IR 75th percentile for control group was 1.78 and the best area under the curve index was obtained for HOMA-IR values of 2.0 [AUC= 0.840 (0.781–0.899 CI 95%), sensitivity (Se): 85%, speci"city (Sp): 83%] while value 2.5 showed best speci"city without important loss in sensitivity [AUC=0,831 (0.773-0.888) Se = 72%, Sp = 94%]. Conclusion: HOMA-IR values above or equal to 2.0 or 2.5 show enhanced diagnostic value in distinguishing non-alcoholic fatty liver disease carriers from control group individuals.

HEADINGS - Insulin resistance. Fatty liver.

INTRODUCTION

Non-alcoholic fatty liver disease (NAFLD) has been pointed out as the most prevalent hepatic disease throughout the world. In the liver biopsy of these patients, pure steatosis, either associated to in!ammation or not, and even steatohepatitis with or without "brosis or cirrhosis, can be found(21). Insulin resistance has a central role in both steatosis installation and in its progression to more advanced forms of the disease as non-alcoholic steatohepatitis (NASH), what makes it the main pathogenic mechanism of NAFLD(9, 24).

Several methods have been used for diagnosing insulin resistance in humans. Glycemic clamp continues to be the gold standard procedure; however, its complexity limits its application in daily medical practice(16). Several methods using glycemia and insulinemia measurements, both during fasting or after oral or endovenous glucose overload, have been proposed(26, 28). Due to the simplicity of its determination and calculation, insulin resistance assessment by the homeostatic

assay (HOMA-IR) has been the most frequently employed technique both in clinical practice and in epidemiological studies. HOMA-IR, as proposed by Matthews et al.(19), shows signi"cant correlation to glycemic clamp in non-diabetic patients and has been widely utilized in NAFLD clinical studies(2, 13, 14). In these studies, however, cut values of HOMA-IR to identify IR have been arbitrarily set and show great variety among authors. Moreover, these values were obtained in case-control studies or in trials performed with a small number of control subjects(11, 12, 25, 27).

The purpose of this study was to identify the best HOMA-IR cut value to differentiate non-diabetic NAFLD patients from a control group of non-obese subjects, without any known liver disease and with oral glucose overload within the normal range.

METHODS

For this study, 116 NAFLD patients were selected, diagnosed by liver biopsy or ultrasound detection of

Departamento de Medicina, 1 Disciplina de Gastroenterologia and 2 Endocrinologia, Universidade Federal de São Paulo, SP, BrazilCorrespondence: Dr. Edison Roberto Parise - Rua Botucatu, 740 - Vila Clementino - 04023-900 - São Paulo, SP, Brazil. E-mail: [email protected]

Salgado ALFA, Carvalho L, Oliveira AC, Santos VN, Vieira JG, Parise ER. Insulin resistance index (HOMA-IR) in the differentiation of patients with non-alcoholic fatty liver disease and healthy individuals

Arq Gastroenterol166 v. 47 – no.2 – abr./jun. 2010

steatosis and increased liver enzymes, excluding patients with fasting glucose level above 125 mg/dL, positive hepatitis virus B or C serology (third generation ELISA detection method), alcohol consumption >20 g/day, other associated liver diseases, or use of medication with hepatotoxic potential. Control group constituted of 88 subjects without any detectable liver disease, body mass index <25, normal GTT, and of comparable age and gender with the group of evaluated patients.

Histological criteria: in patients submitted to percutaneous liver biopsy, histological analysis was conducted according to the criteria established by Mateonni et al.(18) and Brunt et al.(7).

A 2-hour oral glucose tolerance test (GTT) was performed after a 12-hour fasting period, following glucose overload with 75 g of dextrosol, diluted in 300 mL of water. Blood samples were collected in order to measure glycemia and insulinemia at fasting and 30, 60, 90, and 120 minutes after glucose overload.

AST, ALT, GGT, alkaline phosphatase, and glucose values were assessed by automatic kinetic methods. Insulinemia values were obtained by immuno!uorometric assay (Perkin Elmer BR-CS).

HOMA-IR was calculated using the formula: HOMA-IR = [glucose (nmol/L) * insulin (�U/mL)/22.5], using fasting values(19).

According to the criteria adopted by the American Diabetes Association for fasting and post-oral overload glycemic values(4), patients were classi"ed as intolerant or pre-diabetic when fasting glycemia was between 100 and 125 mg/dL, or when the 120-minute glycemia reached between 140 and 199 mg/dL. Patients with 120-minute GTT glucose >200 mg/dL were considered diabetic.

The study protocol was approved by the Human Ethics Committee of Hospital de São Paulo, Universidade Federal de São Paulo SP, Brazil. Written informed consent was obtained from all participant subjects.

Statistical analysis: values were expressed as mean ± standard error of the mean (M ± SEM). Student t test and �2 test were employed for comparisons among the studied groups and the ROC curve was used to evaluate diagnostic sensitivity. Values of P<0.05 were considered statistically signi"cant.

RESULTS

One hundred sixteen NAFLD carriers were included in this study, amounting to 86 (74%) males with mean age of 41 years. Control group constituted of 88 subjects, 54 (68%) males with mean age of 42 years (Table 1).

Mean values of analyzed parameters (insulin, glycemia, HOMA-IR) for the studied groups are shown in Table 1.

Among NAFLD patients, 49% were biopsied, of which 32% were classi"ed as carriers of NAFLD types 1 and 2 (steatosis with or without in!ammation) and 68% as types 3 and 4 (non-alcoholic steatohepatitis with or without "brosis). Liver cirrhosis was diagnosed in 4.5% of the cases.

Control group presented HOMA-IR mean value of 1.27 ± 0.63 (median = 1.10; 75th percentile = 1.78). All control group subjects presented glycemic curves within normal range.

Among the studied patients, 33.6% (39/116) were classi"ed as intolerant, while 10.4% (12/116) showed a diabetic curve. In 28.4% (33/116) of the patients, fasting glycemia was found above 99 mg%.

Signi"cant differences between patients and control group subjects were observed in all studied points of glycemia and insulinemia concentrations (Figures 1A and 1B).

TABLE 1. Characteristics of the studied groupsCharacteristics Control group NAFLD group Pn 88 116

Age (years) 42.3 ± 11.7 41.2 ± 11.0 0.488

Male gender (%) 54 (61%) 86 (74%) 0.075

BMI (kg/height2) 20.41 ± 0.31 30.05 ± 0.51 <0.001

Glucose (mg/dL) 84.7 ± 6.8 94.5 ± 9.9 <0.001

Insulin (µUi/mL) 6.04 ± 2.8 15.7 ± 7.6 <0.001

HOMA-IR 1.2 ± 0.6 3.9 ± 2.8 <0.001

FIGURE 1A. Comparative values for glucose levels during oral tolerance test in control and NAFLD groups

NAFLDControl

250

200

150

100

500 30 60 90 120

TIME

gluc

ose

mg/

dL

FIGURE 1B. Comparative values for insulin levels during oral tolerance test in control and NAFLD groups

200

150

100

50

00 30 60 90 120

NAFLD

Control

Tim

insu

lin μ

U/L

Salgado AL, Carvalho L, Oliveira AC, Santos VN, Vieira JG, Parise ER. Arq Gastroenterol. 2010 Jun;47(2):165-9


See corresponding CME exam on page 268.

Dietary glycemic index and liver steatosis1–3

Silvia Valtuena, Nicoletta Pellegrini, Diego Ardigo, Daniele Del Rio, Filippo Numeroso, Francesca Scazzina,Lucilla Monti, Ivana Zavaroni, and Furio Brighenti

ABSTRACTBackground: Insulin resistance (IR) and liver steatosis (LS) areinterlinked metabolic derangements whose prevalence is rapidlyincreasing, but the effect of dietary carbohydrate quality on LS isunknown.Objective: The objective was to describe the relation of IR and LSto total carbohydrate, total dietary fiber, and the glycemic index (GI)and glycemic load of the diet.Design: The study was a cross-sectional evaluation of 247 appar-ently healthy subjects who had no evidence of viral, toxic, or auto-immune hepatitis and who were unselected for alcohol intake. Thehomeostasis model assessment index was used as a surrogate mea-sure of IR, and a liver echography was used as a proxy for LS grading.Dietary data were collected by using 3-d food records. Total carbo-hydrate intake, total dietary fiber, GI, and glycemic load were cal-culated by using a semiquantitative food-frequency questionnaireconcerning the dietary sources of carbohydrates.Results: The prevalence of high-grade LS (HG-LS) increased sig-nificantly across quartiles of dietary GI (P for trend ! 0.034):HG-LS in the 4th quartile (high GI) was twice that in the first 3quartiles (low to medium GIs), whereas no relation was observedwith total carbohydrates, total dietary fiber, or glycemic load. Ininsulin-sensitive subjects (first 3 quartiles of homeostasis modelassessment index of IR), the prevalence of HG-LS did not differsignificantly between GI groups, but, in insulin-resistant subjects(4th quartile of homeostasis model assessment index of IR), it wastwice as high in those with high GI as in those with low to mediumGIs (P " 0.005).Conclusions: High-GI dietary habits are associated with HG-LS,particularly in insulin-resistant subjects. Dietary advice on the qual-ity of carbohydrate sources therefore may be a complementary toolfor preventing or treating LS of metabolic origin. Am J ClinNutr 2006;84:136–42.

KEY WORDS Diet, glycemic index, insulin resistance, liversteatosis, metabolic syndrome

INTRODUCTION

Nonalcoholic fatty liver disease (NAFLD), a recognized riskfactor for nonalcoholic steatohepatitis (NASH) and liver cirrho-sis, is the term used to describe an abnormal accumulation oftriacylglycerols in the hepatocyte that is commonly observed inobese and insulin-resistant persons (1). Given the increasingworldwide prevalence of insulin resistance (IR) and associated

syndromes, the development of dietary strategies to prevent theoccurrence of NAFLD and its progression to NASH is of majorinterest. Weight loss and supplementation with dietary antioxi-dants have been proposed for potential use in the prevention ofNAFLD and NASH, but little or no attention has been paid to theeffect of carbohydrate quality on fat accretion in the liver (2-4).

It is well documented that both excess body fat and resistanceto the action of insulin impair the suppression of circulatingnonesterified fatty acids (NEFAs) in the postprandial state,which favors a greater NEFA influx into the hepatocyte andsubsequent synthesis of triacylglycerols. In addition, the meta-bolic pathway controlling mitochondrial fat oxidation is down-regulated in the presence of IR, which further contributes tointracellular fat accretion (5). Because weight loss and insulin-sensitizing medications have shown some efficacy in the treat-ment of NAFLD, we hypothesize that diets that can modulateeither body weight or the metabolic consequences of IR couldalso have an effect on liver steatosis (LS).

Epidemiologic data indicate that a low dietary glycemic index(GI) is associated with lower food intake and body weight (6, 7).In addition, low dietary GI and glycemic load (GL) seem to belinked to favorable lipid profiles and lower concentrations ofC-reactive protein only in overweight and obese persons, whichsuggests that the metabolic effects of dietary carbohydrates maybe particularly important in insulin-resistant persons (8–10). In-deed, randomized controlled intervention trials in insulin-resistant subjects and persons with type 2 diabetes show that lowdietary GI can improve the metabolic derangements associatedwith IR—namely, glucose intolerance, hyperinsulinemia, andthe increase in circulating postprandial NEFA—but whetherdietary GI or GL has an effect on fat accretion in the liver remainsto be established (11–13). The aim of this cross-sectional studywas to investigate the relation between dietary carbohydrates,

1 From the Departments of Public Health (NP, DDR, FS, and FB) andInternal Medicine and Biomedical Sciences (SV, DA, FN, and IZ), Univer-sity of Parma, Parma, Italy, and the Core Lab, the Diabetology, Endocrinol-ogy, and Metabolic Disease Unit, Division of Medicine, Hospital San Raf-faele, Milan, Italy (LM).

2 Supported by IST-2001-33204, “Healthy Market,” from the EuropeanCommunity; COFIN 2001 from the Italian Ministry of University and Re-search; and CU01.009.23.CT26 from the National Research Council.

3 Reprints not available. Address correspondence to F Brighenti, HumanNutrition Unit, Department of Public Health, University of Parma, Via Vol-turno 39, 43100 Parma, Italy. E-mail: [email protected].

Received September 15, 2005.Accepted for publication January 31, 2006.

136 Am J Clin Nutr 2006;84:136–42. Printed in USA. © 2006 American Society for Nutrition


ay 5, 2011w

ww

.ajcn.orgD

ownloaded from

effect of high-GI foods may exacerbate liver fat depositionthrough multiple mechanisms related to the reciprocal control ofglucose and fatty acid metabolism in the liver and other tissues.In insulin-resistant persons consuming high-GI foods, the liver issimultaneously exposed to hyperglycemia (deriving from higherglucose availability) and hyperinsulinemia (deriving from bothhyperglycemia and IR). This combined exposure may up-regulate de novo lipogenesis (DNL) and inhibit NEFA oxidationthrough the effect of malonyl coenzyme A on carnitine palmitoyltransferase-1–mediated mitochondrial transport of fatty acids(32). In effect, there is evidence that a liquid diet administered asa meal increases hepatic lipogenesis much more than does thesame diet administered via continuous feeding, which suggests a

role not only of the total amount of glucose consumed but also ofthe glucose delivery rate (33). However, the availability of fattyacids for conversion into triacylglycerols remains critical to drivefat deposition in the liver and cannot be entirely accounted for byan increase in carbohydrate-stimulated DNL. Donnelly et al (34)recently showed that, in hyperinsulinemic NAFLD patients,DNL contributes up to 26% of liver triacylglycerols, whereas thecontribution of peripheral NEFA is !59%, and that of dietarytriacylglycerols is 15%; their findings were discussed in an ac-companying commentary (35). Lack of prolonged suppression ofcirculating NEFA deriving from lack of inhibition of thehormone-sensitive lipase in adipose tissue due to peripheral IRmay explain the enhanced availability of NEFA for hepatic tri-acylglycerol synthesis and fat deposition (35), and this too maybe exacerbated by high-GI diets (11, 30).

Two limitations of this study merit further comment. One is theuse of ultrasonography to assess liver fat content. As comparedwith the gold standard—histologic analysis—which was ethi-cally unacceptable for most of our “healthy” volunteers, thesensitivity (80%–90%) and specificity (85%–95%) of ultra-sonography to show increased fat in the liver are very good, andits accuracy approaches 100% for subjects with echographicfeatures of moderate or severe steatosis; thus, we are confidentabout the stratification of our subjects into subgroups of LS (36,37). The second limitation is that subjects were unselected foralcohol intake. It is well known that alcohol consumption"20–30 g/d has a steatogenic effect, but this variable did notseem to contribute significantly to LS in this population of mild-to-moderate drinkers independently of other known risk factorsfor LS (ie, sex, waist circumference, and IR). Moreover, alcoholintake did not differ significantly between dietary GI groups, andthere is no reason to believe in an undetected association betweenalcohol intake and GI of the diet.

Finally, it cannot be excluded that other components present inthe foods identified as having low GI or some dietary habitsassociated with a low- or high-GI diet could have played a role in

TABLE 4Percentage prevalence, unadjusted and adjusted odds ratios (ORs), and 95% CIs for high-grade liver steatosis in relation to sex, waist circumference,homeostasis model assessment index for insulin resistance (HOMA-IR), and dietary glycemic index1

Risk factors Prevalence Unadjusted ORs (95% CIs) P Adjusted ORs (95% CIs)2 P

% (n/total)Sex

Females 15.9 (17/107) 1 1Males 23.9 (32/134) 1.66 (0.87, 3.19) 0.1283 5.76 (2.21, 15.01) # 0.001

Waist circumferenceLow 9.9 (14/141) 1 1High 35.0 (35/100) 4.89 (2.46, 9.72) # 0.001 9.29 (3.54, 24.43) # 0.001

HOMA-IRLow 10.5 (19/181) 1 1High 50.0 (30/60) 7.83 (3.93, 15.6) # 0.001 6.17 (2.83, 13.44) # 0.001

Glycemic indexLow 16.5 (30/182) 1 1High 32.2 (19/59) 2.41 (1.23, 4.71) 0.010 3.15 (1.34, 7.39) 0.009

1 In the logistic regression models, values for waist circumference were $88 and $102 cm for females and males, respectively (low), and "88 and "102cm for females and males, respectively (high); values for HOMA-IR and glycemic index are first 3 quartiles (low) and 4th quartile (high).

2 Adjusted for sex, waist circumference, HOMA-IR, and glycemic index. Additional terms for plasma concentrations of triacylglycerols and HDLcholesterol, alcohol intake, intakes of carbohydrates and fiber, and dietary glycemic load (first 3 quartiles compared with the 4th quartile) did not modify anyrisk estimate. Glycemic index was calculated on a scale in which glucose % 100.

3 The sex & waist circumference interaction was significant, P % 0.008.

TABLE 5Percentage contribution to total dietary glycemic load (GL) from differentfood sources of carbohydrates according to the average glycemic index(GI) of the diet

Food group

Low tomedium GI1

(n % 182)High GI2

(n % 59) P

Breads, pizza, and savory snacks(% of GL)

45.9 (11.9)3 57.8 (11.4) # 0.001

Pasta, rice, and corn (% of GL) 21.6 (9.7) 16.7 (8.6) # 0.001Pulses and potatoes (% of GL) 1.3 (1.5) 1.3 (1.1) 0.610Milk and dairy products (% of GL) 2.4 (2.9) 1.2 (2.8) # 0.010Juices and soft drinks (% of GL) 0.8 (2.3) 0.7 (3.8) 0.760Fresh and dry fruit and nuts (% of

GL)8.0 (5.9) 3.9 (2.7) # 0.001

Sugar, sweets, cakes, and chocolate(% of GL)

16.5 (9.8) 13.6 (9.0) 0.077

Fresh vegetables (% of GL) 0.1 (0.2) 0.0 (0.1) 0.4081 The 1st to 3rd quartiles.2 The 4th quartile.3 Median; interquartile range in parentheses (all such values). P by

Kolmogorov-Smirnov z test.

GLYCEMIC INDEX AND LIVER STEATOSIS 141


ay 5, 2011w

ww

.ajcn.orgD

ownloaded from

Valtuena S, Pellegrini N, Ardigo D, et al (2006). Am J Clin Nutr 84:136–142

ü MELHORA RI

ü REDUÇÃO DA INGESTÃO CALÓRICA (> SACIEDADE)

ü  6 A 12 MESES:

ü  > PERDA DE PESO

ü  > DIMINUIÇÃO TRIGLICÉRIDOS

ü  > AUMENTO HDL

ü MELHORA VÁRIOS MARCADORES DE DCV, MESMO NA AUSÊNCIA DE PERDA DE PESO

NUTRITION MYTHS

HIDRATOS DE CARBONO SÃO TODOS MAUS

A Low–Glycemic Load Diet FacilitatesGreater Weight Loss in Overweight AdultsWith High Insulin Secretion but Not inOverweight Adults With Low InsulinSecretion in the CALERIE TrialANASTASSIOS G. PITTAS, MD

1

SAI KRUPA DAS, PHD2

CHERYL L. HAJDUK, MS, MPH, RD2

JULIE GOLDEN, MD2

EDWARD SALTZMAN, MD2

PAUL C. STARK, SCD3

ANDREW S. GREENBERG, MD1,2

SUSAN B. ROBERTS, PHD2

L ifestyle changes, in particular reduc-ing energy intake, are the corner-stone of current approaches to

weight loss and prevention of type 2 dia-betes. However, there is currently no con-sensus that one dietary regimen is moreeffective than another for weight loss (1)or whether particular diets work better foridentifiable groups of individuals. Thereis evidence, however, to suggest that bothinsulin resistance and insulin secretionplay a role in body weight regulation (2–12). Therefore, dietary factors such as thedietary glycemic load (glycemic load !glycemic index [GI] " available carbohy-drate amount) that influence these pa-rameters may theoretically interact withsubject-specific characteristics of glucose-insulin dynamics to influence the effect ofdifferent hypocaloric diets on weight lossor maintenance (13,14). Weight lossstudies using the concept of the dietaryglycemic index or glycemic load haveshown conflicting results for heteroge-neous groups of individuals (15–21).

In a 6-month controlled feeding trialin healthy overweight adults with normalglucose tolerance, we tested the hypothe-sis that individuals with higher insulinsecretion lose more weight when ran-

domized to a low– glycemic load dietcompared with a high– glycemic loaddiet.

RESEARCH DESIGN ANDMETHODS — This study was per-formed as part of the ComprehensiveAssessment of Long-term Effects of Re-ducing Intake of Energy (CALERIE) trialat the Human Nutrition Research Centeron Aging at Tufts University with ap-proval from the Tufts-New EnglandMedical Center Human Investigation Re-view Committee. Written informed con-sent was obtained from all participants.Healthy women and men aged 24–42years with a BMI of 25–29.9 kg/m2 andfasting plasma glucose #100 mg/dl wererecruited. After a 6-week baseline periodwhen usual energy requirements forweight stability were assessed using thedoubly labeled water method (22), partic-ipants were randomized for 24 weeks toeither a high–glycemic load diet (60%carbohydrate, 20% protein, 20% fat, 15 gfiber/1,000 kcal, mean estimated dailyglycemic index of 86, and glycemic loadof 116 g/1,000 kcal) or a low–glycemicload diet (40% carbohydrate, 30% pro-tein, 30% fat, 15 g giber/1,000 kcal, mean

estimated daily glycemic index of 53, andglycemic load of 45 g/1,000 kcals) at 30%calorie restriction compared with baselineindividual energy needs. The glycemic in-dex and glycemic load of the diets weredetermined using the International Ta-bles of Glycemic Index and GlycemicLoad (23) and the Nutrition Data Systemfor Research (version 4.05_33) developedby the Nutrition Coordinating Center,University of Minnesota, Food and Nutri-ent Database 33, released in 2002 (24).During the 6-month intervention period,all food was provided by the research cen-ter, and participants were requested toconsume only this food and report addi-tional foods if they were eaten. To maxi-mize adherence to the study diet, regularbehavioral group meetings and individualsessions with a dietitian were held. Fromparticipants’ reports of leftover food andextra items, actual daily nutrient intakeduring the intervention period was calcu-lated (24).

Height ($0.1 cm) was measured atbaseline, and body weight ($100 g) wasmeasured weekly. Insulin secretion wasalso estimated at baseline, as was the in-sulin value at 30 min (INS-30) after a 75-goral glucose tolerance test (25). Insulinsensitivity in the fasting state was esti-mated at baseline by the homeostasismodel assessment of insulin resistance(HOMA-IR) (26). Glucose was measuredby the hexokinase method, and insulinwas measured by radioimmunoassay. Toexamine change in weight at 6 months,we used general linear models adjustingfor baseline weight, HOMA-IR and INS-30, and the interaction between diet "HOMA-IR and diet " INS-30 (PROCGLM procedure in SAS software, version8.2; SAS Institute, Cary, NC).

RESULTS — A total of 32 (25 womenand 7 men) of 34 enrolled participantscompleted the 6-month intervention pe-riod and necessary measurements. Atbaseline, their mean age was 34.6 years,

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

From the 1Division of Endocrinology, Diabetes, and Metabolism, Tufts-New England Medical Center,Boston, Massachusetts; the 2Energy Metabolism Laboratory, Jean Mayer USDA Human Nutrition ResearchCenter on Aging, Tufts University, Boston, Massachusetts; and the 3Institute for Clinical Research and HealthPolicy Studies, Tufts-New England Medical Center, Boston, Massachusetts.

Address correspondence and reprint requests to Anastassios G. Pittas, M.D., Division of Endocrinology,Diabetes, and Metabolism, Tufts-New England Medical Center, 750 Washington St., #268, Boston, MA02111. E-mail: [email protected].

Received for publication 28 July 2005 and accepted in revised form 5 September 2005.Abbreviations: HOMA-IR, homeostasis model assessment of insulin resistance; INS-30, insulin value at

30 min.A table elsewhere in this issue shows conventional and Systeme International (SI) units and conversion

factors for many substances.© 2005 by the American Diabetes Association.The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby

marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

C l i n i c a l C a r e / E d u c a t i o n / N u t r i t i o nB R I E F R E P O R T

DIABETES CARE, VOLUME 28, NUMBER 12, DECEMBER 2005 2939

BMI was 27.5 kg/m2, and fasting insulinwas 82.5 pmol/l. The mean target energyintake was 1,966 kcal/day, and the meanreported daily energy intake during theintervention did not differ between thetwo groups (2,017 kcal in the high–glycemic load diet vs. 1,972 kcal in thelow–glycemic load diet, P ! 0.70).

We examined whether baselineINS-30 predicted change in weight overthe 6-month intervention period. Wefound a diet " INS-30 interaction (P !0.02) in the multivariate predictionmodel, and the weight data were stratifiedinto two groups separated by the medianINS-30 value (Fig. 1). Participants withhigh baseline INS-30 lost more weight ifrandomized to the low– glycemic loaddiet compared with the high–glycemicload diet (P # 0.05). The reverse was ob-served in the low–INS-30 group, namely,low–INS-30 participants in the high–glycemic load diet lost more weight thanthose in the low–glycemic load diet, butthe difference was not statistically signifi-cant (P ! 0.25). We also examinedwhether baseline HOMA-R predictedweight change, and we found no diet "HOMA-R interaction.

CONCLUSIONS — The main find-ing from this pilot study was that healthyoverweight women and men with rela-tively greater insulin secretion in response

to a standard oral glucose tolerance testlost more weight when assigned to a low–glycemic load hypocaloric diet than to ahigh–glycemic load diet, but there was nodifferential effect of the two diets onweight loss in individuals who had rela-tively lower insulin secretion.

Data from human studies on whetherinsulin secretion predicts future weightgain (4,6,27–29) or affects the ability ofoverweight individuals to lose weight inresponse to a hypocaloric diet (7,30) arecontroversial. However, the influence ofpostchallenge hyperinsulinemia onweight loss may be particularly importantfor specific dietary compositions, in par-ticular diets that differ in glycemic load orglycemic index, as suggested by animalstudies (31–33). High–glycemic load di-ets increase postprandial hyperinsulin-emia, which favors fatty acid uptake,inhibition of lipolysis, and energy storageleading to weight gain (34). High–glycemic load diets also lead to otherpostprandial metabolic changes, includ-ing a lower glucose nadir and increase incounterregulatory hormones that may ex-plain increased hunger and increased en-ergy intake in the postabsorptive period,presumably leading to weight gain overtime (32). All of these mechanisms maybe exacerbated in individuals with highinsulin secretory capacity at baseline,which makes them more susceptible to

weight gain upon exposure to a high–glycemic load diet (35). These individualscan be hypothesized to do best on low–glycemic load diets, a hypothesis sup-ported by our present findings.

Our results require confirmation infurther studies with larger numbers ofsubjects, but nevertheless offer the firstevidence that simple indexes of insulinsecretion may help enhance weight losssuccess in overweight individualsthrough the use of targeted dietary recom-mendations specific for insulin secretionstatus.

Acknowledgments— This research was sup-ported by National Institutes of Health GrantsK23-DK61506 (to A.G.P.), U01-AG20480 (toS.B.R.), and U.S. Department of Agriculturecooperative agreement no. 58-1950-4-401 (toS.B.R. and A.S.G.).

We thank the study participants and staff ofthe MRU at the USDA Human Nutrition Re-search Center on Aging.

References1. Dansinger ML, Gleason JA, Griffith JL,

Selker HP, Schaefer EJ: Comparison of theAtkins, Ornish, Weight Watchers, andZone diets for weight loss and heart dis-ease risk reduction: a randomized trial.JAMA 293:43–53, 2005

2. Wedick NM, Mayer-Davis EJ, WingardDL, Addy CL, Barrett-Connor E: Insulinresistance precedes weight loss in adultswithout diabetes: the Rancho BernardoStudy. Am J Epidemiol 153:1199–1205,2001

3. Odeleye OE, de Courten M, Pettitt DJ, Ra-vussin E: Fasting hyperinsulinemia is apredictor of increased body weight gainand obesity in Pima Indian children. Dia-betes 46:1341–1345, 1997

4. Sigal RJ, El-Hashimy M, Martin BC,Soeldner JS, Krolewski AS, Warram JH:Acute postchallenge hyperinsulinemiapredicts weight gain: a prospective study.Diabetes 46:1025–1029, 1997

5. Swinburn BA, Nyomba BL, Saad MF,Zurlo F, Raz I, Knowler WC, Lillioja S,Bogardus C, Ravussin E: Insulin resis-tance associated with lower rates ofweight gain in Pima Indians. J Clin Invest88:168–173, 1991

6. Schwartz MW, Boyko EJ, Kahn SE, Ravus-sin E, Bogardus C: Reduced insulin secre-tion: an independent predictor of bodyweight gain. J Clin Endocrinol Metab 80:1571–1576, 1995

7. McLaughlin T, Abbasi F, Carantoni M,Schaaf P, Reaven G: Differences in insulinresistance do not predict weight loss inresponse to hypocaloric diets in healthyobese women. J Clin Endocrinol Metab 84:578–581, 1999

Figure 1—Mean ($SEM) weight change during a 6-month feeding study of a high– (HG) vs.low– (LG) glycemic load diet in overweight adults stratified by baseline insulin secretion based onserum insulin at 30 min after a 75-g oral glucose tolerance test (low INS-30 # 473 pmol/l [66mU/l] # high INS-30). P values are adjusted for baseline weight. *P # 0.005 for within-groupchange in weight from baseline.

Glycemic load and weight loss

2940 DIABETES CARE, VOLUME 28, NUMBER 12, DECEMBER 2005

Pittas AG, et al. Diabetes Care. 2005 Dec;28(12):2939-41.

NÃO FOI SIGNIFICATIVO

FOI SIGNIFICATIVO

Human Physiology

Insulin Sensitivity Determines the Effectivenessof Dietary Macronutrient Composition onWeight Loss in Obese WomenMarc-Andre Cornier,*†† W. Troy Donahoo,*†† Rocio Pereira,* Inga Gurevich,¶ Rickard Westergren,**Sven Enerback,** Peter J. Eckel,‡ Marc L. Goalstone,*¶ James O. Hill,†§ Robert H. Eckel,*‡ andBoris Draznin*¶

AbstractCORNIER, MARC-ANDRE, W. TROY DONAHOO,ROCIO PEREIRA, INGA GUREVICH, RICKARDWESTERGREN, SVEN ENERBACK, PETER J. ECKEL,MARC L. GOALSTONE, JAMES O. HILL, ROBERT H.ECKEL, AND BORIS DRAZNIN. Insulin sensitivitydetermines the effectiveness of dietary macronutrientcomposition on weight loss in obese women. Obes Res.2005;13:703–709.Objective: To determine whether macronutrient composi-tion of a hypocaloric diet can enhance its effectiveness andwhether insulin sensitivity (Si) affects the response to hy-pocaloric diets.Research Methods and Procedures: Obese nondiabetic in-sulin-sensitive (fasting insulin ! 10 !U/mL; n " 12) andobese nondiabetic insulin-resistant (fasting insulin # 15!U/mL; n " 9) women (23 to 53 years old) were random-ized to either a high carbohydrate (CHO) (HC)/low fat (LF)(60% CHO, 20% fat) or low CHO (LC)/high fat (HF) (40%CHO, 40% fat) hypocaloric diet. Primary outcome measuresafter a 16-week dietary intervention were: changes in bodyweight (BW), Si, resting metabolic rate, and fasting lipids.Results: Insulin-sensitive women on the HC/LF diet lost13.5 $ 1.2% (p ! 0.001) of their initial BW, whereas those

on the LC/HF diet lost 6.8 $ 1.2% (p ! 0.001; p ! 0.002between the groups). In contrast, among the insulin-resistantwomen, those on the LC/HF diet lost 13.4 $ 1.3% (p !0.001) of their initial BW as compared with 8.5 $ 1.4% (p! 0.001) lost by those on the HC/LF diet (p ! 0.04 betweentwo groups). These differences could not be explained bychanges in resting metabolic rate, activity, or intake. Over-all, changes in Si were associated with the degree of weightloss (r " %0.57, p ! 0.05).Discussion: The state of Si determines the effectiveness ofmacronutrient composition of hypocaloric diets in obesewomen. For maximal benefit, the macronutrient composi-tion of a hypocaloric diet may need to be adjusted tocorrespond to the state of Si.

Key words: CHO, fat, insulin resistance

IntroductionSuccessful dietary interventions are based on a significant

reduction in caloric intake, relative to energy expenditure(1). The question of whether the macronutrient compositionof hypocaloric diets has an impact on the effectiveness ofthese diets, however, has gained substantial interest with thepopularization of low-carbohydrate (CHO)1 (LC), hypoca-loric dietary regimens (2–11).

Total body insulin sensitivity (Si) is an overall measure ofthe ability of insulin to regulate glucose uptake and metab-olism (12,13). Insulin-resistant (IR) individuals requirehigher than normal levels of insulinemia to maintain normalglycemia. Thus, either fasting or postprandial hyperinsulin-emia prevents the development of impaired glucose toler-

Received for review June 7, 2004.Accepted in final form January 20, 2005.The costs of publication of this article were defrayed, in part, by the payment of pagecharges. This article must, therefore, be hereby marked “advertisement” in accordance with18 U.S.C. Section 1734 solely to indicate this fact.Departments of *Medicine and †Pediatrics, ‡Adult General Clinical Research Center, andthe §Center for Human Nutrition, University of Colorado Health Sciences Center, ¶ResearchService of the Denver Veterans Administration Medical Center, Denver, Colorado; and**Medical Genetics, Department of Medical Biochemistry, Gothenburg University, Gote-borg, Sweden.††These authors contributed equally to the design and implementation of this study.Address correspondence to Boris Draznin, Veterans Administration Medical Center, (151)1055 Clermont Street Denver, CO 80220.E-mail: [email protected] © 2005 NAASO

1 Nonstandard abbreviations: CHO, carbohydrate; LC, low CHO; Si, insulin sensitivity; IR,insulin resistant; IS, insulin sensitive; HC, high CHO; LF, low fat; HF, high fat; BW, bodyweight; GCRC, General Clinical Research Center; RMR, resting metabolic rate; RQ,respiratory quotient; FFA, free fatty acid; HDL, high-density lipoprotein; LDL, low-densitylipoprotein.

OBESITY RESEARCH Vol. 13 No. 4 April 2005 703

Human Physiology











Human Physiology











Cornier MA, et al. Obes Res. 2005 Apr;13(4):703-9.

Human Physiology











diet lost significantly more weight if they were identified atbaseline to be IS as opposed to IR (p ! 0.01).

Impact of Weight Loss on Metabolic ProfileFasting insulinemia improved in both IS and IR groups

(IS, 7.08 " 1.26 to 5.42 " 0.82 !U/mL, p ! 0.008; IR,19.57 " 0.89 to 8.84 " 0.95 !U/mL, p # 0.001) with asignificantly greater improvement in the IR group (p #0.001). Similarly, although the mean Si did not change inthe IS patients (5.12 " 0.83 to 4.16 " 0.95, p ! 0.35), thisparameter improved substantially in the IR groups (3.04 "

1.44 to 3.87 " 1.20, p ! 0.028; p ! 0.005 for differencebetween Si groups), suggesting that weight loss in the IRpatients improves their Si. Overall, the change in Si corre-lated with the degree of weight loss (r ! $0.57, p # 0.05).Weight loss at the end of the 16-week intervention periodhad a favorable but not significant effect on total choles-terol, low-density lipoprotein (LDL) cholesterol, and HDLcholesterol in all groups (Table 2). Triglycerides also im-proved with weight reduction in all groups except for the IRgroup randomized to the HC/LF diet (p ! 0.003). Thisgroup, in fact, demonstrated an increase in their triglycerideconcentrations (from 124 " 15 to 157 " 10 mg/dl, p #0.05). As predicted, leptin concentrations decreased in allgroups with weight loss regardless of the macronutrientcomposition of the hypocaloric diet.

Measures of Energy BalanceIS subjects in the LC/HF group and IR subjects in the

HC/LF group lost the expected amount of weight for thecaloric deficit imposed (24). In contrast, IS individuals onHC/LF and IR patients on LC/HF hypocaloric diets lostalmost twice the expected amount of weight (Figure 1). Allsubjects received all of their food from the GCRC. Carefuland frequent dietary recalls revealed no detectable differ-ences between the groups that lost the expected amount ofweight and those that lost more weight. Thus, although wecannot definitively rule out differences in energy intake asthe etiology for the differences in weight loss, differences inenergy expenditure seem to be a more plausible explanation.We addressed the energy expenditure side of the equationonly through self-report and RMR. Overall, when all sub-jects were pooled, RMR was found to be decreased (1304 "36 to 1221 " 43 kcal/d, p ! 0.03); however, there were nosignificant changes in RMR in any of the four groups whenanalyzed separately, and no group or diet interactions werefound. Therefore, changes in RMR could not account for theweight loss differences observed.

DiscussionThe salient feature of this investigation is that the state of

Si profoundly influenced the response to a distinct macro-nutrient composition of hypocaloric diet. Moderately obesewomen who were IS at baseline responded better to anHC/LF hypocaloric diet than to an LC/HF hypocaloric diet.On the other hand, equally moderately obese women whowere more IR at baseline responded better to an LC/HFhypocaloric diet than to an HC/LF one.

The most important point of this discussion is why the ISgroup on an HC/LF diet and the IR group on an LC/HF dietlost almost twice the amount of weight as their counterpartson the opposite diets. All subjects lost at least the expectedamount of weight on a hypocaloric diet (daily deficit of 400

Figure 1: Absolute (A) and percentage (B) change in BW in IS andIR women randomized to 16 weeks of hypocaloric HC/LF orLC/HF diet. (*) p # 0.01 for diet effect within IS group. (†) p #0.05 for Si effect within HC/LF diet. (‡) p # 0.05 for diet effectwithin IR group. (§) p # 0.01 for Si effect within LC/HF diet.

Effects of Macronutrient Composition on Weight Loss, Cornier et al.

706 OBESITY RESEARCH Vol. 13 No. 4 April 2005

Cornier MA, et al. Obes Res. 2005 Apr;13(4):703-9.


Effects of a Low–Glycemic Loadvs Low-Fat Diet in Obese Young AdultsA Randomized TrialCara B. Ebbeling, PhDMichael M. Leidig, RDHenry A. Feldman, PhDMargaret M. Lovesky, RDDavid S. Ludwig, MD, PhD

WITH PREVALENCE AP-proaching one third ofthe population, obesityis among the most im-

portant medical problems in the UnitedStates1 and identification of effective di-etary treatment has become a majorpublic health priority.2 Three populardiets—low fat, low carbohydrate, andlow glycemic load—have recently re-ceived much attention. However, clini-cal trials have produced inconsistentfindings, with some suggesting that onediet is superior for weight loss3-8 andothers indicating no difference be-tween diets.9-11 This inconsistency mayarise from methodological problemsboth within and between trials, such asdifferent treatment intensity betweengroups, inadequate attention to treat-ment fidelity, variable nutrition edu-cation and dietary counseling strate-gies, and confounding by dietary andnondietary factors. An alternative ex-planation for this inconsistency re-lates to inherent physiological differ-ences among study participants.

One physiological mechanism thatmight relate weight loss to dietary com-position is individual differences in in-sulin secretion. Diets with a high gly-cemic load (the mathematical product

of the glycemic index and the carbo-hydrate amount12) result in higher post-prandial insulin concentration, calo-

Author Affiliations: Department of Medicine, Chil-dren’s Hospital Boston, Boston, Mass.Corresponding Author: David S. Ludwig, MD, PhD, Chil-dren’s Hospital Boston, 300 Longwood Ave, Boston, MA02115 ([email protected]).

Context The results of clinical trials involving diet in the treatment of obesity have beeninconsistent, possibly due to inherent physiological differences among study participants.

Objective To determine whether insulin secretion affects weight loss with 2 popu-lar diets.

Design, Setting, and Participants Randomized trial of obese young adults (aged18-35 years; n=73) conducted from September 2004 to December 2006 in Boston,Mass, and consisting of a 6-month intensive intervention period and a 12-month fol-low-up period. Serum insulin concentration at 30 minutes after a 75-g dose of oralglucose was determined at baseline as a measure of insulin secretion. Outcomes wereassessed at 6, 12, and 18 months. Missing data were imputed conservatively.

Interventions A low–glycemic load (40% carbohydrate and 35% fat) vs low-fat(55% carbohydrate and 20% fat) diet.

Main Outcome Measures Body weight, body fat percentage determined by dual-energy x-ray absorptiometry, and cardiovascular disease risk factors.

Results Change in body weight and body fat percentage did not differ between thediet groups overall. However, insulin concentration at 30 minutes after a dose of oral glu-cose was an effect modifier (group!time! insulin concentration at 30 minutes: P=.02for body weight and P=.01 for body fat percentage). For those with insulin concentra-tion at 30 minutes above the median (57.5 µIU/mL; n=28), the low–glycemic load dietproduced a greater decrease in weight (–5.8 vs –1.2 kg; P=.004) and body fat percent-age (–2.6% vs –0.9%; P=.03) than the low-fat diet at 18 months. There were no sig-nificant differences in these end points between diet groups for those with insulin con-centration at 30 minutes below the median level (n=28). Insulin concentration at 30 minutesafter a dose of oral glucose was not a significant effect modifier for cardiovascular dis-ease risk factors. In the full cohort, plasma high-density lipoprotein cholesterol and tri-glyceride concentrations improved more on the low–glycemic load diet, whereas low-density lipoprotein cholesterol concentration improved more on the low-fat diet.

Conclusions Variability in dietary weight loss trials may be partially attributable todifferences in hormonal response. Reducing glycemic load may be especially impor-tant to achieve weight loss among individuals with high insulin secretion. Regardlessof insulin secretion, a low–glycemic load diet has beneficial effects on high-densitylipoprotein cholesterol and triglyceride concentrations but not on low-density lipopro-tein cholesterol concentration.

Trial Registration clinicaltrials.gov Identifier: NCT00130299JAMA. 2007;297:2092-2102 www.jama.com

2092 JAMA, May 16, 2007—Vol 297, No. 19 (Reprinted) ©2007 American Medical Association. All rights reserved.

by guest on April 16, 2009 www.jama.comDownloaded from



































Ebbeling CB, et al. JAMA. 2007 May 16;297(19):2092-102


















Ebbeling CB, et al. JAMA. 2007 May 16;297(19):2092-102

saturated fat content of the low-fat dietprovides a plausible explanation for thisfinding.43 When controlling for mac-ronutrient content, some clinical trialsshow benefits of a low–glycemic in-dex diet on LDL cholesterol concen-tration.11,14 Therefore, we speculate thata low–glycemic load diet in which satu-rated fat is kept low (eg, by substitut-ing monounsaturated or polyunsatu-rated fat from vegetable sources for

Table 2. Participant Satisfaction*Mean (SE)

PValue†

Low–GlycemicLoad Diet(n = 32)

Low-Fat Diet(n = 34)

How satisfied are you with this diet? 7.0 (0.4) 6.9 (0.3) .80

How easy has this diet been? 5.3 (0.3) 5.1 (0.4) .68

How tasty have the foods been? 6.9 (0.3) 6.8 (0.3) .92

How satisfied are you with your weight loss to date? 4.1 (0.4) 4.7 (0.5) .37*On 10-cm visual analog scale with 0 indicating not at all and 10, extremely.†Testing for group difference using the t test.

Figure 3. Change in Body Weight

Time, mo

5

0

–5

–10

Cha

nge

From

Bas

elin

e, k

g

P = .99

Time, mo

P = .90

Time, mo

AllInsulin Concentration !57.5 µIU/mL at

30 min After 75-g Dose of Oral GlucoseInsulin Concentration >57.5 µIU/mL at

30 min After 75-g Dose of Oral Glucose

0 6 12 18 0 6 12 18 0 6 12 18

P = .02

Low–Glycemic Load Diet Low-Fat Diet

In the full cohort (left panel), weight loss did not differ significantly between participants assigned to the low–glycemic load diet vs the low-fat diet. Among participantsfor whom data were available at baseline (right 2 panels), insulin concentration at 30 minutes after a 75-g dose of oral glucose was a significant effect modifier (P=.02).Error bars indicate 95% confidence intervals. Data based on repeated-measures analysis, accounting for within-subject correlation and between-subject variability. TheP value at the lower left of each panel tests the group! time interaction. Missing data were imputed conservatively.

Table 3. Changes in Adiposity and Cardiovascular Disease Risk Factors*6-mo Follow-up, Mean (SE)

PValue

18-mo Follow-up, Mean (SE)

PValue

Low–GlycemicLoad Diet Low-Fat Diet

Low–GlycemicLoad Diet Low-Fat Diet

Body fat percentage†All !1.3 (0.4) !1.4 (0.3) .94 !1.5 (0.4) !1.1 (0.3) .50

Insulin concentration "57.5 µIU/mLat 30 min‡

!0.9 (0.5) !2.2 (0.6) .11 !0.9 (0.5) !1.4 (0.6) .56

Insulin concentration #57.5 µIU/mLat 30 min‡

!2.0 (0.6) !0.4 (0.5) .04 !2.6 (0.6) !0.9 (0.5) .03

LipidsCholesterol, mg/dL

Low-density lipoprotein !5.8 (3.4) !16.3 (3.3) .03 !0.3 (3.4) !10.6 (3.3) .03

High-density lipoprotein 1.6 (1.4) !4.4 (1.3) .002 !3.7 (1.5) !8.2 (1.5) .03

Triglycerides,%§ !21.2 (4.7) !4.0 (5.6) .02 !9.0 (5.4) 2.0 (6.0) .18

Blood pressure, mm HgSystolic !5.1 (2.3) !4.8 (2.3) .93 !3.2 (2.3) 1.1 (2.3) .18

Diastolic !2.4 (1.7) !2.0 (1.7) .88 0 (1.7) 2.9 (1.7) .22

Glucose homeostasisFasting glucose, mg/dL 1.6 (1.3) !0.3 (1.3) .31 2.1 (1.3) 1.4 (1.3) .73

Fasting insulin, µIU/mL !2.1 (0.8) !0.9 (0.8) .28 !0.8 (0.8) 0 (0.8) .49SI conversion factors: To convert low-density and high-density lipoprotein cholesterol to mmol/L, multiply by 0.0259; glucose to mmol/L, multiply by 0.0555; triglycerides to mmol/L,

multiply by 0.0113.*The means are from repeated-measures analysis of conservatively imputed data. The P values are for significant difference between diet groups.†Effect modification by insulin concentration at 30 minutes, P = .01. Effect modification was nonsignificant for all other listed variables, P#.10.‡After a 75-g dose of oral glucose.§Measurements in mg/dL were log-transformed for analysis to reduce skew; mean change is expressed as percentage, calculated as 100% ! (exp[mean log change]!1).

LOW–GLYCEMIC LOAD VS LOW-FAT DIET



NÃO FOI SIGNIFICATIVO NÃO FOI SIGNIFICATIVO FOI SIGNIFICATIVO

Hypothesis: Could Excessive Fructose Intake and UricAcid Cause Type 2 Diabetes?

Richard J. Johnson, Santos E. Perez-Pozo, Yuri Y. Sautin, Jacek Manitius, Laura Gabriela Sanchez-Lozada,Daniel I. Feig, Mohamed Shafiu, Mark Segal, Richard J. Glassock, Michiko Shimada, Carlos Roncal,and Takahiko Nakagawa

Division of Nephrology (R.J.J., Y.Y.S., L.G.S.-L., M.Sha., M.Se., R.J.G., M.Shi., C.R., T.N.), University of Florida,Gainesville, Florida 32620-0224; Division of Nephrology (S.E.P.-P.), Mateo Orfila Hospital, 07703 Minorca (BalearicIslands), Spain; Division of Nephrology (J.M.), The Ludwig Rydygier Medical University, 85-067 Bydgoszcz, Poland;Department of Nephrology (L.G.S.-L.), Instituto Nacional de Cardiología Ignacio Chavez, 14080 Mexico City, Mexico;Division of Pediatric Nephrology (D.I.F.), Baylor College of Medicine, Houston, Texas 77030; and Retired Professor(R.J.G.), Torrance, California

We propose that excessive fructose intake (>50 g/d) may beone of the underlying etiologies of metabolic syndrome andtype 2 diabetes. The primary sources of fructose are sugar(sucrose) and high fructose corn syrup. First, fructose intakecorrelates closely with the rate of diabetes worldwide. Sec-ond, unlike other sugars, the ingestion of excessive fructoseinduces features of metabolic syndrome in both laboratoryanimals and humans. Third, fructose appears to mediate themetabolic syndrome in part by raising uric acid, and there arenow extensive experimental and clinical data supporting

uric acid in the pathogenesis of metabolic syndrome.Fourth, environmental and genetic considerations providea potential explanation of why certain groups might bemore susceptible to developing diabetes. Finally, we discussthe counterarguments associated with the hypothesis and apotential explanation for these findings. If diabetes mightresult from excessive intake of fructose, then simple publichealth measures could have a major impact on improvingthe overall health of our populace. (Endocrine Reviews 30:96–116, 2009)

I. IntroductionII. Unique Characteristics of Fructose Metabolism

III. Fructose Causes Metabolic Syndrome in AnimalsIV. Mechanism(s) for Fructose-Induced Insulin ResistanceV. Mechanism(s) by Which Fructose Induces Other Features

of the Metabolic Syndrome: Role of Uric AcidVI. Human Studies with Fructose

VII. Epidemiological Studies: Sugar Intake and Type 2Diabetes

VIII. Epidemiological Studies: Uric Acid and Type 2 DiabetesIX. Do Other Conditions That Modify Uric Acid Levels Affect

the Development of Metabolic Syndrome or Diabetes?X. Twelve Countering Arguments and Caveats

XI. The Thrifty Gene RevisitedXII. What Research Should Be Done to Prove Our Hypothesis?

I. Introduction

Although diabetes was described by Aretaeus, Galen,and Paracelsus, by the mid to late 1800s William Prout

(1) and others recognized that diabetes could have two pre-sentations: one manifesting as a rapidly progressive andwasting condition in a thin and feeble individual (likely type1 diabetes), and a slower and more progressive disease in an

overweight or obese subject (likely type 2 diabetes) (1, 2).Both conditions were rare; indeed, Osler (3) projected a prev-alence of approximately two or three cases per 100,000 pop-ulation in Europe and North America. By the early 1900s,however, a remarkable rise in the prevalence of the secondtype of diabetes was observed in Europe and the UnitedStates (4). Similarly, a dramatic increase in diabetes wasobserved in a number of tropical countries (5). In these earlyreports, the type of subject developing diabetes was oftenwealthy, overweight, and living in an urban environment (4,5). However, over the last 50 yr there has been a transitionsuch that diabetes is now increasing most rapidly among thepoor and minorities (6). Although some of the increase indiabetes prevalence may be due to the increasing longevityof the population, an increase in the rate of type 2 diabetesis also being observed among the young, suggesting that anactive process is driving the epidemic. Today diabetes ispresent in over 217 million individuals worldwide. Approx-imately 7% of the U.S. adult population has type 2 diabetesthat carries a yearly financial burden of over $130,000,000,000(7). Over the next few decades a remarkable increase indiabetes is projected, especially in Asia and India (8). By 2030,over 350 million people are projected to suffer from thiscondition, making it one of the most serious diseases ofhumankind (7, 8).

Identifying the etiology of type 2 diabetes is key to pre-vention. The frequent association of diabetes with obesity hasled many investigators to propose that obesity may be re-sponsible for up to 90% of type 2 diabetes (9). Obesity, andin particular intraabdominal fat accumulation, has been

First Published Online January 16, 2009Abbreviations: Glut, Glucose transporter; HDL, high-density lipopro-

tein; HFCS, high fructose corn syrup; KHK, ketohexokinase; MCP-1,monocyte chemoattractant protein-1; NO, nitric oxide.Endocrine Reviews is published by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.

0163-769X/09/$20.00/0 Endocrine Reviews 30(1):96–116Printed in U.S.A. Copyright © 2009 by The Endocrine Society

doi: 10.1210/er.2008-0033

96

at Colorado State Univ Libraries on September 28, 2009 edrv.endojournals.orgDownloaded from

scription factors (nuclear factor-!B and activator protein-1),resulting in platelet-derived growth factor-dependent pro-liferation, cyclooxygenase-2-dependent thromboxane pro-duction, MCP-1 and C-reactive protein synthesis, and stim-ulation of angiotensin II (101, 103, 116–120). Uric acid alsoinhibits endothelial cell proliferation and migration (101).Finally, uric acid has potent effects on proximal tubular cells(stimulating MCP-1 production) as well as adipocytes (in-ducing oxidative stress, stimulating oxidized lipids, and low-ering NO levels) (44, 104).

More recently, uric acid has been implicated in the patho-genesis of hypertension (reviewed in Ref. 121). An elevateduric acid has been found to be an independent risk factor forhypertension in multiple studies (121). Uric acid is also com-monly elevated in subjects with essential hypertension, par-ticularly in newly diagnosed hypertension (122). Further-more, lowering uric acid has been found to normalize bloodpressure in 66% of adolescents with essential hypertensionand asymptomatic hyperuricemia compared with 3% in theplacebo-treated controls (123).

The mechanism by which uric acid could raise blood pres-sure has been studied in the rat. Rats normally have a lowuric acid due to the presence of uricase, an enzyme in the liverthat degrades uric acid to allantoin. In contrast, humans haveno functional uricase due to a mutation that occurred in theMiocene epoch (124). To study the effects of hyperuricemia,it was necessary to provide a uricase inhibitor (oxonicacid) in the diet. When uric acid levels were raised, thelaboratory animals developed the clinical, histological,and hemodynamic characteristics of essential hyperten-sion (125, 126). The hypertension was shown to be initiallymediated by oxidative stress, activation of the renin an-giotensin system, and endothelial dysfunction, and itcould be reversed by treating with antioxidants, l-argi-nine, or inhibitors of the renin angiotensin system (116,125, 127, 128). However, as renal microvascular diseasedevelops, the blood pressure switches from being uricacid-dependent to one that is salt-sensitive and kidneydependent (119), similar to many other models of salt-sensitive hypertension (129).

FIG. 3. Potential mechanisms by which fructose and uric acid may induce insulin resistance. Fructose enters cell via a transporter (primarilyGlut 5) where it is acted on by fructokinase (KHK). As part of this metabolism, ATP depletion may occur, generating uric acid with systemiceffects that block insulin-dependent NO-mediated vascular dilation as well as direct cellular effects on the adipocyte. Fructose also causes denovo lipogenesis that can lead to intracellular triglycerides that can also induce insulin resistance. DAG, Diacylglycerol; PKC, protein kinaseC; VLDL, very low-density lipoprotein.

Johnson et al. • Fructose as a Cause of Type 2 Diabetes Endocrine Reviews, February 2009, 30(1):96–116 101


adipocyte. There is evidence that insulin resistance is medi-ated in part by inflammation and oxidative stress within theadipocyte (109). Sautin et al. (104) have recently shown thaturic acid induces this phenotype in cultured adipocytes. Inaddition, Cheung et al. (110) reported that xanthine oxi-doreductase knockout mice fail to become fat due to a defectin adipogenesis. These studies therefore implicate xanthineoxidase and uric acid in metabolic syndrome.

The observation that fructose-induced insulin resistancecan occur independent of weight gain or differences in en-ergy intake (37, 78) does not negate the possibility that theinsulin resistance could still be mediated by the adipocyte.We have found, for example, that fructose-glucose or sucrosecombinations result in increased intraabdominal fat accu-mulation compared with starch-fed rats given equivalentenergy intake and with similar weight gain (R.J. Johnson, C.Roncal, Y.Y. Sautin, T. Nakagawa, L.G. Sánchez-Lozada, un-published observations). Finally, it is likely that insulin re-sistance will continue to manifest once an animal developsextensive fat stores via classical mechanisms (10), so thatcontinued insulin resistance might be expected even if fruc-tose intake was reduced. Hence, fructose induced insulinresistance might be considered an initiator of the insulinresistance syndrome, with obesity-based mechanisms per-petuating the condition.

V. Mechanism(s) by Which Fructose Induces OtherFeatures of the Metabolic Syndrome: Role of

Uric Acid

There is increasing evidence that intracellular ATP deple-tion and uric acid generation may have important roles in the

ability of fructose to induce features of the metabolic syn-drome. As mentioned, lowering uric acid was found to ame-liorate a number of features of metabolic syndrome in fructose-fed rats, including hypertension, hypertriglyceridemia,hyperinsulinemia, insulin resistance, renal vasoconstriction,glomerular hypertension, and renal microvascular disease(37, 78, 83). Allopurinol can also reduce fructose-inducedmonocyte chemoattractant protein-1 (MCP-1) production inhuman proximal tubular cells (44).

The finding that uric acid might have a role in metabolicsyndrome is surprising because uric acid is considered oneof the major antioxidants in the circulation (111). Further-more, whereas uric acid is commonly elevated in subjectswith metabolic syndrome (112), it has been thought to beelevated secondary to the hyperinsulinemia (113) that occursin these subjects.

Nevertheless, the evidence that uric acid may be a truemediator of cardiovascular disease is mounting. Uric acid,whereas an antioxidant in the extracellular environment, caninduce oxidative stress in a variety of cells including vascularsmooth muscle cells and murine adipocytes (103, 104, 114).The mechanism appears to involve stimulation of nicotin-amide adenine dinucleotide phosphate oxidase (104). Uricacid also reduces NO bioavailability in endothelial cells, adi-pocytes, and vascular smooth muscle cells (100, 102–104),and the mechanism is mediated by oxidative stress (103, 104,114), the stimulation of arginase (115), and the direct scav-enging of NO by uric acid (106). Uric acid also stimulatesvascular smooth muscle cells by entering cells via a specificorganic anion transport pathway with the stimulation ofmitogen-activated kinases (p38 and ERK) and nuclear tran-

FIG. 2. Effect of fructose on various organ systems. Table sugar, HFCS, and natural sources provide fructose, which in excess has numerous effectson the brain, liver, vasculature, kidney, and adipocyte. The net effects induce all features of the metabolic syndrome and ultimately type 2 diabetes.

100 Endocrine Reviews, February 2009, 30(1):96–116 Johnson et al. • Fructose as a Cause of Type 2 Diabetes


ALIMENTO! DOSE (GRS)! CARGA GLICÉMICA! FRUTOSE!Maçã! 120" 5,5! 11,2!

Alperces! 120" 5,2! 4,0!Banana crua! 120" 12,4! 7,2!

Cerejas! 120" 2,7! 4,3!Tâmaras! 60! 41,6! 13,4!Figos! 60" 15,7! 1,8!Toranja! 120" 2,7! 3,5!Uvas! 120" 8,2! 9,1!Kiwi! 120" 6,2! 5,9!

Manga! 120" 8,5! 9,5!Laranja! 120" 4,6! 5,5!Papaia! 120" 10,2! 4,3!Pêssegos! 120" 4,6! 4,9!Pêras! 120" 4,2! 8,8!Ananás! 120" 7,4! 4,4!Ameixa! 120" 4,8! 4,0!Passas! 60! 28,5! 20,3!Meloa! 120" 3,7! 5,4!

Morangos! 120" 1,3! 3,6!Melancia! 120" 4,3! 6,1!

The effect of two energy-restricted diets, a low-fructose dietversus a moderate natural fructose diet, on weight loss andmetabolic syndrome parameters: a randomized controlled trial

Magdalena Maderoa,⁎, Julio C. Arriagaa, Diana Jalalb, Christopher Rivardb, Kim McFannb,Oscar Pérez-Méndeza, Armando Vázqueza, Arturo Ruiza, Miguel A. Lanaspab,Carlos Roncal Jimenezb, Richard J. Johnsonb, Laura-Gabriela Sánchez Lozadaa,b

a Division of Nephrology, Department of Medicine, Instituto Nacional de Cardiología Ignacio Chávez, Juan Badiano no 1 Col Seccion XVITlalpan, México City, 14080, Méxicob Division of Renal Diseases and Hypertension, University of Colorado, Denver, CO, USA

A R T I C L E I N F O A B S T R A C T

Article history:Received 20 December 2010Accepted 4 April 2011

One of the proposed causes of obesity and metabolic syndrome is the excessive intake ofproducts containing added sugars, in particular, fructose. Although the ability of excessiveintakeof fructose to inducemetabolic syndrome ismounting, to date, no studyhas addressedwhether a diet specifically lowering fructose but not total carbohydrates can reduce featuresof metabolic syndrome. A total of 131 patients were randomized to compare the short-termeffects of 2 energy-restricted diets—a low-fructose diet vs amoderate natural fructose diet—on weight loss and metabolic syndrome parameters. Patients were randomized to receive1500, 1800, or 2000 cal diets according to sex, age, and height. Because natural fructosemightbe differently absorbed compared with fructose from added sugars, we randomized obesesubjects to either a low-fructose diet (<20 g/d) or a moderate-fructose diet with natural fruitsupplements (50-70 g/d) and compared the effects of both diets on the primary outcome ofweight loss in a 6-week follow-up period. Blood pressure, lipid profile, serum glucose, insulinresistance, uric acid, soluble intercellular adhesionmolecule–1, andquality of life scoreswereincluded as secondary outcomes. One hundred two (78%) of the 131 participants werewomen, mean age was 38.8 ± 8.8 years, and the mean bodymass index was 32.4 ± 4.5 kg/m2.Each intervention diet was associated with significant weight loss compared with baseline.Weight loss was higher in the moderate natural fructose group (4.19 ± 0.30 kg) than the low-fructose group (2.83 ± 0.29 kg) (P = .0016). Comparedwith baseline, each intervention diet wasassociated with significant improvement in secondary outcomes. Reduction of energy andadded fructose intakemay represent an important therapeutic target to reduce the frequencyof obesity and diabetes. For weight loss achievement, an energy-restricted moderate naturalfructose diet was superior to a low-fructose diet.

© 2011 Elsevier Inc. All rights reserved.

M E T A B O L I S M C L I N I C A L A N D E X P E R I M E N T A L X X ( 2 0 1 1 ) X X X – X X X

Author contributions: Magdalena Madero, Richard Johnson, Diana Jalal, Miguel Lanaspa, Carlos Roncal Jimenez, Armando Vazquez,and Laura-Gabriela Sanchez Lozada were involved in the study design. Julio Arriaga was responsible for recruiting patients and patientfollow-up. KimMcFann was the statistician. Christopher Rivard and Arturo Ruiz were involved with the laboratory determinations. OscarPerez Mendez supported the study with the lipid profile determinations.

ClinicalTrials.gov NCT0086873.⁎ Corresponding author. Tel.: +52 55 73 69 02; fax: +52 55 73 77 16.

E-mail address: [email protected] (M. Madero).

0026-0495/$ – see front matter © 2011 Elsevier Inc. All rights reserved.doi:10.1016/j.metabol.2011.04.001

ava i l ab l e a t www.sc i enced i r ec t . com

www.metabo l i sm jou rna l . com

Please cite this article as: Madero M, et al, The effect of two energy-restricted diets, a low-fructose diet versus a moderatenatural fructose diet, on weight loss and metabol..., Metabolism (2011), doi:10.1016/j.metabol.2011.04.001

Table 2 – Differences in diet compositions at baseline and among intervention groups

Baselineenergy intake(kcal) total

cohort

% Low-fructose

diet1500 kcal

% Moderate-fructose

diet1500 kcal

% Low-fructose

diet1800 kcal

% Moderate-fructose

diet1800 kcal

% Low-fructose

diet2000 kcal

% Moderate-fructose

diet2000 kcal

%

Pyramid food groupsProtein 416.4 14 264 18 227 15 308 17 272 15 344 17 296 15Fat 764.3 26 427.5 29 441 29 508.5 28 504 28 571.5 29 625.5 32Carbohydrate 1818.9 61 796 54 832 56 971 54 1014 57 1062 54 1104 56Major food groupsDairy products 363.9 13 220 15 220 15 220 12 110 6 220 11 220 11Fruits 528.1 17 60 4 480 32 60 3 540 30 60 3 540 27Animal products 228.3 7 225 15 225 15 225 13 300 17 300 15 300 15Vegetables 178.5 6 100 7 100 7 100 6 50 3 87.5 4 75 4Leguminous products 99.9 3 60 4 60 4 120 7 240 13 120 6 180 9.1Cereals 899.2 30 665 45 280 19 770 43 280 16 875 44 350 18Fat 154.6 6 157.5 11 135 9 292.5 16 270 15 315 16 315 16Candies 71.7 2 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0Juices and soft drinks 263.8 9 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0Snacks 154.8 4 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0Added sugars 56.8 3 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0 0 0.0Average caloric intake 2999.7 100 1487.5 100 1500.0 100 1787.5 100 1790.0 100 1977.5 100 1980.0 100

“%” refers to the percentage of calories from each food group.

5M

ET

ABO

LISM

CLIN

IC

AL

AN

DEX

PER

IM

EN

TA

LX

X(2011)

XX

X–X

XX

Pleasecite

thisarticle

as:Madero

M,et

al,Theeffect

oftw

oenergy-restricted

diets,alow

-fructosediet

versusamoderate

naturalfructosediet,on

weight

lossand

metabol...,

Metabolism

(2011),doi:10.1016/j.metabol.2011.04.001

compared with baseline. Because subjects were assigned todifferent caloric intakes, ANCOVA was then performed,adjusting for baseline and caloric intake. Data from ANCOVAare reported as adjusted least square means ± SE. Paired ttests were used to compare within-group changes frombaseline to 6 weeks. The data were analyzed first forcompleters of the study because those who dropped outtended to drop out after the baseline visit and second asintention to treat. The study was powered to detect an effectsize of 0.5 kg (difference in 1 kg between groups as an effectof the fructose content on each diet with a 2-kg standard

deviation). The number of patients required to have a powerof 80% was 64 patients in each arm.

3. Results

3.1. Participants

A total of 107 (82%) of the 131 subjects recruited into the studycompleted the trial, with baseline characteristics shown inTable 1. Nine patients from the low-fructose arm and 13

Table 3 –Within- and between-group changes in the low-fructose group and themoderate-fructose groupwith natural fruitsupplements

Δ = final ! baseline Low fructose Moderate natural fructose Comparison betweenintervention groups

Δ Mean ± SD P value Δ Mean ± SD P value P value

Weight (kg) !2.94 ± 2.18 <.0001 !4.07 ± 2.39 <.0001 .002Systolic BP (mm/Hg) !9.46 ± 7.77 <.0001 !7.85 ± 8.73 <.0001 .09Diastolic BP (mm/Hg) !5.17 ± 4.69 <.0001 !6.04 ± 5.40 <.0001 .57Fat (%) !2.09 ± 6.32 .02 !2.89 ± 6.33 .002 .10Waist to hip ratio !0.03 ± 0.02 <.0001 !0.18 ± 1.04 .21 .41BMI (kg/m2) !1.18 ± 0.82 <.0001 !1.57 ± 1.08 <.0001 .02Uric acid (mg/dL) !0.24 ± 0.60 .004 !0.22 ± 0.56 .01 .90sICAM (ng/dL) !0.28 ± 0.78 .01 !0.42 ± 0.67 <.0001 .19Urine microalbumin (µg/mg) 0.19 ± 7.70 .85 !0.42 ± 1.84 .11 .32Total cholesterol (mg/dL) !9.75 ± 24.4 .004 !12.76 ± 33.31 .01 .95Triglycerides (mg/dL) !23.50 ± 69.2 .01 !31.76 ± 55.36 <.0001 .48HDL (mg/dL) !0.75 ± 19.67 .79 0.107 ± 12.36 .95 .93Insulin resistance (HOMA) !0.29 ± 0.93 .02 !0.37 ± 0.57 <.0001 .12Blood glucose (mg/dL) !6.14 ± 30.83 .14 !6.96 ± 9.37 <.0001 .07

-7

-6

-5

-4

-3

-2

-1

0

1

2

3

BASELINE FINAL

WEI

GHT

LO

SS (k

ilogr

ams)

LOW FRUCTOSE GROUPMODERATE NATURAL FRUCTOSE GROUP

Fig. 2 – Weight changes for each diet intervention group.

6 M E T A B O L I S M C L I N I C A L A N D E X P E R I M E N T A L X X ( 2 0 1 1 ) X X X – X X X


compared with baseline. Because subjects were assigned todifferent caloric intakes, ANCOVA was then performed,adjusting for baseline and caloric intake. Data from ANCOVAare reported as adjusted least square means ± SE. Paired ttests were used to compare within-group changes frombaseline to 6 weeks. The data were analyzed first forcompleters of the study because those who dropped outtended to drop out after the baseline visit and second asintention to treat. The study was powered to detect an effectsize of 0.5 kg (difference in 1 kg between groups as an effectof the fructose content on each diet with a 2-kg standard

deviation). The number of patients required to have a powerof 80% was 64 patients in each arm.

3. Results

3.1. Participants

A total of 107 (82%) of the 131 subjects recruited into the studycompleted the trial, with baseline characteristics shown inTable 1. Nine patients from the low-fructose arm and 13

Table 3 –Within- and between-group changes in the low-fructose group and themoderate-fructose groupwith natural fruitsupplements

Δ = final ! baseline Low fructose Moderate natural fructose Comparison betweenintervention groups

Δ Mean ± SD P value Δ Mean ± SD P value P value

Weight (kg) !2.94 ± 2.18 <.0001 !4.07 ± 2.39 <.0001 .002Systolic BP (mm/Hg) !9.46 ± 7.77 <.0001 !7.85 ± 8.73 <.0001 .09Diastolic BP (mm/Hg) !5.17 ± 4.69 <.0001 !6.04 ± 5.40 <.0001 .57Fat (%) !2.09 ± 6.32 .02 !2.89 ± 6.33 .002 .10Waist to hip ratio !0.03 ± 0.02 <.0001 !0.18 ± 1.04 .21 .41BMI (kg/m2) !1.18 ± 0.82 <.0001 !1.57 ± 1.08 <.0001 .02Uric acid (mg/dL) !0.24 ± 0.60 .004 !0.22 ± 0.56 .01 .90sICAM (ng/dL) !0.28 ± 0.78 .01 !0.42 ± 0.67 <.0001 .19Urine microalbumin (µg/mg) 0.19 ± 7.70 .85 !0.42 ± 1.84 .11 .32Total cholesterol (mg/dL) !9.75 ± 24.4 .004 !12.76 ± 33.31 .01 .95Triglycerides (mg/dL) !23.50 ± 69.2 .01 !31.76 ± 55.36 <.0001 .48HDL (mg/dL) !0.75 ± 19.67 .79 0.107 ± 12.36 .95 .93Insulin resistance (HOMA) !0.29 ± 0.93 .02 !0.37 ± 0.57 <.0001 .12Blood glucose (mg/dL) !6.14 ± 30.83 .14 !6.96 ± 9.37 <.0001 .07

-7

-6

-5

-4

-3

-2

-1

0

1

2

3

BASELINE FINAL

WEI

GHT

LO

SS (k

ilogr

ams)

LOW FRUCTOSE GROUPMODERATE NATURAL FRUCTOSE GROUP

Fig. 2 – Weight changes for each diet intervention group.

6 M E T A B O L I S M C L I N I C A L A N D E X P E R I M E N T A L X X ( 2 0 1 1 ) X X X – X X X


ARTICLE










Introduction






FRUTA: 160-1435 g/d

ORIGINAL ARTICLE

Metabolic and physiologic improvements fromconsuming a paleolithic, hunter-gatherer type diet

LA Frassetto, M Schloetter, M Mietus-Synder, RC Morris Jr and A Sebastian

Department of Medicine, University of California San Francisco School of Medicine, San Francisco, CA, USA

Background: The contemporary American diet figures centrally in the pathogenesis of numerous chronic diseases—‘diseases ofcivilization’. We investigated in humans whether a diet similar to that consumed by our preagricultural hunter-gathererancestors (that is, a paleolithic type diet) confers health benefits.Methods: We performed an outpatient, metabolically controlled study, in nine nonobese sedentary healthy volunteers,ensuring no weight loss by daily weight. We compared the findings when the participants consumed their usual diet with thosewhen they consumed a paleolithic type diet. The participants consumed their usual diet for 3 days, three ramp-up diets ofincreasing potassium and fiber for 7 days, then a paleolithic type diet comprising lean meat, fruits, vegetables and nuts, andexcluding nonpaleolithic type foods, such as cereal grains, dairy or legumes, for 10 days. Outcomes included arterial bloodpressure (BP); 24-h urine sodium and potassium excretion; plasma glucose and insulin areas under the curve (AUC) during a 2horal glucose tolerance test (OGTT); insulin sensitivity; plasma lipid concentrations; and brachial artery reactivity in response toischemia.Results: Compared with the baseline (usual) diet, we observed (a) significant reductions in BP associated with improved arterialdistensibility (!3.1±2.9, P"0.01 and #0.19±0.23, P" 0.05);(b) significant reduction in plasma insulin vs time AUC, duringthe OGTT (P"0.006); and (c) large significant reductions in total cholesterol, low-density lipoproteins (LDL) and triglycerides(!0.8±0.6 (P"0.007), !0.7±0.5 (P" 0.003) and !0.3±0.3 (P"0.01) mmol/l respectively). In all these measured variables,either eight or all nine participants had identical directional responses when switched to paleolithic type diet, that is, nearconsistently improved status of circulatory, carbohydrate and lipid metabolism/physiology.Conclusions: Even short-term consumption of a paleolithic type diet improves BP and glucose tolerance, decreases insulinsecretion, increases insulin sensitivity and improves lipid profiles without weight loss in healthy sedentary humans.European Journal of Clinical Nutrition advance online publication, 11 February 2009; doi:10.1038/ejcn.2009.4

Keywords: paleolithic diet; blood pressure; glucose tolerance; insulin sensitivity; lipids

Introduction

In 1985, anthropologists Eaton and Konner (1985) intro-duced the general medical community in ‘paleolithicnutrition. A consideration of its nature and current implica-tions’. ‘Paleolithic’ refers to the period of history of the genusHomo, beginning more than 2 million years ago andcontinuing until about 10000 years ago (10 kya, when theneolithic period began) when humans began to cultivate

plants (predominantly cereal grains) and domesticateanimals (Brandt, 2007). During that interval (more than2-million year), culminating in the emergence of today’s soleHomo species, Homo sapiens, about 200 kya (McDougall et al.,2005), our ancestors, including Homo sapiens, lived ashunter-gatherers, eating wild animal-source foods (leanmeats, internal organs, bone marrow, but no dairy) anduncultivated plant-source foods (mostly fruits, nongrain,vegetables, nuts, but no legumes). As the 10000 or so yearssince the beginning of the agriculture and animal domes-tication began,that is less than 1% of Homo evolutionarytime—leaves little time for evolutionary forces to redesignthe core metabolic and physiological processes in a majorway in response to the major dietary changes introduced bythe agricultural revolution and food animal domestication,

Received 25 July 2008; revised 20 November 2008; accepted 30 December2008

Correspondence: Dr LA Frassetto, San Francisco School of Medicine, Campusbox 0126 505 Parnassus Avenue, room 1202M San Francisco, CA 94143,USAE-mail: [email protected]

European Journal of Clinical Nutrition (2009), 1–9& 2009 Macmillan Publishers Limited All rights reserved 0954-3007/09 $32.00

www.nature.com/ejcn

Table 2 Menu: 1 week of increasing fiber and potassium diets (‘ramp’ diets) and Paleo diet (per 3000 kcal)

Diet Ramp 1 Ramp 2 Ramp 3 Paleo diet

Potassium content 125mmol 180mmol 259mmol 339mmol

BreakfastHoney Honey Honey HoneyFresh pineapple Orange juice Orange juice Carrot JuiceScrambled eggs Fresh pineapple Fresh pineapple Fresh pineapple

Pork tenderloin Pork tenderloin Pork tenderloin

AM snackLettuce, cucumber, celeryand tomatoes with oil andvinegar dressing

Celery, cucumber, redpeppers and tomatoes withoil and vinegar dressing

Low salt tomato soup Almonds

Carrot juice Carrot juice

LunchStir-fried fresh zucchini Carrot juice Carrot juice Carrot juiceTuna salad (tuna, radish,shallots, mayo) on lettuce

Tuna salad/mayo on lettuce Tuna salad/mayo on lettuce Tuna Salad (mayo, radishes,shallots) on lettuce

Applesauce Honey Low-salt tomato soup withchopped tomatoes

Honey

Day SnackLettuce, carrot and pepperswith oil and vinegar dressing

Turkey/mayo in lettuce wrap Turkey/mayo with lettucewrap

Turkey, guacamole and tomatolettuce roll-ups

Canned pears Carrots and tomatoes with oiland vinegar dressing

Carrots and tomatoes with oilnd vinegar dressing

Honey Tomato juice

DinnerChicken breast stir-fry withbroccoli and garlic

Chicken breast stir-fry withfresh spinach, garlic andbroccoli

Chicken breast stir-fry withfresh spinach, broccoli andgarlic

Chicken breast stir-fry with freshspinach, garlic and broccoli

Mandarin oranges Mandarin oranges Mandarin oranges Roasted parsnips andmushrooms with thyme

Honey Tomato juice Low salt tomato soup

PM SnackTurkey and tomatoes withmayo in lettuce wrap

Cantaloupe Cantaloupe Cantaloupe

Carrot juice Carrot juice Carrot juiceHoney

Table 3 Effect of the paleolithic diet on metabolic variables

Category Variable Usual diet Paleolithic diet % Change P-value

Lipids Total cholesterol (mmol/l) 4.7±0.9 4.0±0.7 !16 0.007HDL (mmol/l) 1.3±0.2 1.3±0.3 "4 NSLDL (mmol/l) 3.0±0.7 2.3±0.6 !22 0.003VLDL (mmol/l) 0.4±0.2 0.3±0.1 !35 0.01Triglycerides (mmol/l) 0.9±0.4 0.6±0.1 !35 0.01

Fasting Fasting insulin (pmol/l) 69±63 21±7 !68 0.07Insulin and Glucose Fasting glucose (mmol/l) 18±3 17±2 !5 NS

OGTT Insulin AUC (pmol#h/l) 533±222 361±194 !39 0.006HOMAa 3.2±3.2 1.0±0.4 !72 0.07

Abbreviation: HOMA$homeostatic assessment.aEquation for HOMA$ (fasting insulin# fasting glucose)/22.4.

Health benefits of a Paleo dietLA Frassetto et al

5

European Journal of Clinical Nutrition

ORIGINAL ARTICLE






Introduction






www.nature.com/ejcn

Table 2 Menu: 1 week of increasing fiber and potassium diets (‘ramp’ diets) and Paleo diet (per 3000 kcal)

Diet Ramp 1 Ramp 2 Ramp 3 Paleo diet

Potassium content 125mmol 180mmol 259mmol 339mmol

BreakfastHoney Honey Honey HoneyFresh pineapple Orange juice Orange juice Carrot JuiceScrambled eggs Fresh pineapple Fresh pineapple Fresh pineapple

Pork tenderloin Pork tenderloin Pork tenderloin

AM snackLettuce, cucumber, celeryand tomatoes with oil andvinegar dressing

Celery, cucumber, redpeppers and tomatoes withoil and vinegar dressing

Low salt tomato soup Almonds

Carrot juice Carrot juice

LunchStir-fried fresh zucchini Carrot juice Carrot juice Carrot juiceTuna salad (tuna, radish,shallots, mayo) on lettuce

Tuna salad/mayo on lettuce Tuna salad/mayo on lettuce Tuna Salad (mayo, radishes,shallots) on lettuce

Applesauce Honey Low-salt tomato soup withchopped tomatoes

Honey

Day SnackLettuce, carrot and pepperswith oil and vinegar dressing

Turkey/mayo in lettuce wrap Turkey/mayo with lettucewrap

Turkey, guacamole and tomatolettuce roll-ups

Canned pears Carrots and tomatoes with oiland vinegar dressing

Carrots and tomatoes with oilnd vinegar dressing

Honey Tomato juice

DinnerChicken breast stir-fry withbroccoli and garlic

Chicken breast stir-fry withfresh spinach, garlic andbroccoli

Chicken breast stir-fry withfresh spinach, broccoli andgarlic

Chicken breast stir-fry with freshspinach, garlic and broccoli

Mandarin oranges Mandarin oranges Mandarin oranges Roasted parsnips andmushrooms with thyme

Honey Tomato juice Low salt tomato soup

PM SnackTurkey and tomatoes withmayo in lettuce wrap

Cantaloupe Cantaloupe Cantaloupe

Carrot juice Carrot juice Carrot juiceHoney

Table 3 Effect of the paleolithic diet on metabolic variables

Category Variable Usual diet Paleolithic diet % Change P-value

Lipids Total cholesterol (mmol/l) 4.7±0.9 4.0±0.7 !16 0.007HDL (mmol/l) 1.3±0.2 1.3±0.3 "4 NSLDL (mmol/l) 3.0±0.7 2.3±0.6 !22 0.003VLDL (mmol/l) 0.4±0.2 0.3±0.1 !35 0.01Triglycerides (mmol/l) 0.9±0.4 0.6±0.1 !35 0.01

Fasting Fasting insulin (pmol/l) 69±63 21±7 !68 0.07Insulin and Glucose Fasting glucose (mmol/l) 18±3 17±2 !5 NS

OGTT Insulin AUC (pmol#h/l) 533±222 361±194 !39 0.006HOMAa 3.2±3.2 1.0±0.4 !72 0.07

Abbreviation: HOMA$homeostatic assessment.aEquation for HOMA$ (fasting insulin# fasting glucose)/22.4.


5


ORIGINAL ARTICLE






Introduction






www.nature.com/ejcn

and in the brachial artery cross-sectional area, all arecorrelated (r!0.71, P!0.047; r!0.72, P!0.04; r!0.74,P!0.04, respectively).In all the consistency data reported, the Wilcoxon

signed-rank test gave significant P-values, ranging fromo0.004 to o0.04.

Discussion

We found, in a small group of sedentary, slightly overweight(body mass index: 27. 8±2.4 kg/m2), but not obese adulthumans, that switching from their usual diet to a paleolithic-type diet, which contained no cereal grains, dairy andlegumes, resulted, after only a short period of time andwithout any weight loss or increase in activity levels, in(a) significant reductions in BP associated with improvedarterial distensibility; (b) large significant reductions andincreases, respectively, in 24-h urine sodium and potassiumexcretion; (c) reduced urinary calcium excretion; (d) asignificant reduction in fasting plasma insulin concentra-tion; (e) a significant reduction in the plasma insulin vs timeAUC, during an OGTT; (f) the highest reductions in plasma

insulin AUC occurring in participants with the highestincrease in urinary potassium (index of dietary potassium);(g) a significantly lower integrated value of plasma insulinconcentration over the period of the OGTT (that is, theplasma insulin AUC) relative to the corresponding integratedvalue of plasma glucose concentration (that is, the plasmaglucose AUC)—an index of improved insulin sensitivity;(h) improved insulin sensitivity by homeostatic modelassessment; (i) improved insulin sensitivity proportional tothe degree of baseline insulin resistance; (j) large significantreductions in total cholesterol, LDL and triglycerides; and (k)no significant change in HDL. In addition to the quantitativestatistical significance, in all these measured variables, eithereight or all of the nine participants had identical directionalresponses to the switch from the usual diet to the paleolithic-type diet. The direction of change indicated very consis-tently improved metabolic and physiological status withrespect to circulatory, carbohydrate and lipid metabolism/physiology. We did not consider the BP changes surprising,in as much as switching to the paleolithic diet greatlyreduced sodium excretion and increased potassium excre-tion, with a fivefold increase in the urine K/Na ratio—reflecting, presumably, corresponding changes in the BP

! In

sulin

AU

C, p

mol

*hr/

L

! Daily Urine K excretion, mmol/day

-560

-420

-280

-140

0

140

280

! A

rter

ial c

ross

-sec

tion

al a

rea,

cm

2

! Daily Urine K excretion, mmol/day

-0.04

-0.02

0.00

0.02

0.04

0.06

r=0.90, p=0.005r=0.74, p=0.04

0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 160 180

Figure 1 (a and b) Physiological changes correlating with increases in average daily urine potassium excretion (an index of dietary potassiumintake. (a) Increasing arterial dilation and (b) decreasing insulin secretion during 2-h oral glucose tolerance test.

Table 4 Resting blood pressure measurements and brachial artery reactivity data

Factor Days –2 to 0 (usual diet) Days 15 to 17 (Paleo diet) P-value

Resting BPSystolic BP (mmHg) 116±10 "2.6±5.1 NSDiastolic BP (mmHg) 71±6 "3.4±2.7 0.006MAP (mmHg) 86±7 "3.1±2.9 0.01

Brachial artery diameter at baseline (BAD; mm) 3.97±0.88 3.98±0.85 0.14Peak brachial artery diameter during hyperemia (pkFMD; mm) 4.25±0.83 4.35±0.73 0.05Absolute difference (pkFMD-BAD; mm) 0.288±0.089 0.371±0.158 0.06

Abbreviations: BAD, brachial artery diameter; BP, blood pressure.


6


Lindeberg S, Lundh B. J Intern Med. 1993; 233: 269-275!

1990: Não existia Electricidade, Telefones, Veículos a Motor

Alguns produtos ocidentais eram recebidos da Nova Guiné, mas a influência do estilo de vida ocidental era minímo

DIETA TÍPICA EM KITAVA!Nov-Dec 1990"

"

"Alimentos !Total !Prot ! Gor !HC !Ener!(médias) "(g) "(g) "(g) "(g) "(kJ) "" "

"Tubérculos "1200 "25 "2 "300 "5600"Fruta "400 "3 "<1 "50 "920"Coco "110 "4 "43 "7 "1865"Peixe "85"Porco "<1 "- "- "- "-"Outras Carnes "<5 "- "- "- "-"Outros "200 "5 "<1 "14 "360"Processados "<1 "0 "<1 "<1 "20""Total "2 000 "54 "50 "370 "9200" "

Lindeberg S et al. AJCN 1997;66:845

0

1

2

3

4

5

6

7

8

25-39 40-59 60-74 25-39 40-59 60-74

Insu

lina

jeju

m

Insulina jejum!(Ul/ml)!!

Kitava Suécia

Homens Mulheres

Idade

Lindeberg S, Eliasson M, Lindahl B, Ahren B. Metabolism 1999; 48:1216-9 !

-3

-2

-1

0

1

2

3

4

Kitava Sweden

Males

-3

-2

-1

0

1

2

3

4

Kitava Sweden

Females

Fig. log HOMA in the age group 50-86 years in Kitava and Sweden. HOMA = fP-Glucose in mmol/L x fP-Insulin in µU/mL / 22.5.

ÍNDICE HOMA!

Carrera-Bastos P, Fontes-Villalba M, O'Keefe JH, Lindeberg S, Cordain L. Res Rep Clin Cardiol 2011;2:15-35.

42

44

46

48

50

52

54

Homens Mulheres

Índice Cintura (cm)/altura (m)

Kitava Suécia

Lindeberg, S, Soderberg, S, Ahren, B, Olsson, T. J Intern Med, 2001; 249: 553-8!

0.5

0

-0.5

-1

-1.5

-2

1.5

0.5

0

-0.5

-1

-1.5

-2

20 30 40 50 60 70 80

Age (y)

FIGURE 1. Systolic and diastolic blood pressure in Kitavan and Swedish males and females in relation to means and SDs in Swedish males aged 20-29y old (systolic blood pressure: 129 ± 10 mm, diastolic blood pressure: 84 ± 7 mm). Values are expressed as ratios of Swedish SDs (z scores). For example,systolic blood pressure in Kitavan males aged 20-69 y is l.5 SD below the mean of Swedish 20-29-y-old males, whereas systolic blood pressure ofKitavan males aged 70-79 y is at the same level as that of 20-29-y-old Swedish males.

848 LINDEBERG ET AL

G)00C’)N

:,Cl)Cl)

0. 000

.00

0Cl)

C’)

a)00Cl)N

a)

U)U)a)0.

000

.000U)CO0

Total cholesterol, triacylglycerol, LDL cholesterol, and apoB increased in males between 20 and 50 y of age, whereasHDL cholesterol and apo A-I decreased. Among the fewerfemales there was a slightly positive (nonsignificant) age trendonly for apo B. A slight linear age-related increase in lipopro-tein(a) was present in males.

Plasma fibrinogen increased with age in Kitavan males andfemales (r = 0.28 and 0.33, P 0.0018 and 0.03, respec-tively), as did FVIIc (r = 0. 1 8 and 0.28, P = 0.048 and 0.07,respectively), FVHIc (r = 0.39 and 0.39, P 0.0001 and0.0083, respectively), and vWF:Ag (r = 0.40 and 0.46, P =

0.0001 and 0.0024, respectively). PAI-l tended to decreasewith age in females.

DISCUSSION

We believe our findings regarding age relations of cardio-vascular risk factors to be valid for the following reasons. Ageestimates were considered to be accurate to within 3 y for most

subjects, which appears exceptional for surveys in traditionalpopulations who are mostly unaware of individual ages. TheKitavan study population was a random sample aged 20-86 ymixed with a substantial number of self-selected subjects in theage group of 20-49 y. This could infer some health-relatedselection bias although none of the measured variables differedbetween randomly assigned and self-selected subjects. Forexample, eligible subjects with low body weight or low bloodpressure as a result of concomitant disease could be less willingto participate. Such subjects would probably tend to be older,which would lead to underestimation of the noted age-relateddecrease in BMI. If nonparticipants aged < 50 y were lesshealthy and also had higher blood pressures than did partici-pants, the conclusion that blood pressure does not increase withage may be erroneous. However, randomly assigned and non-participating subjects did not by appearance differ in bodycomposition, agility, or amount of physical activity except inthe age group of 80-96 y, for which participants tended to beyounger and less often disabled than nonparticipants. This

at Lund University Libraries on D

ecember 23, 2005

ww

w.ajcn.org

Dow

nloaded from

Lindeberg S et al. AJCN 1997;66:845

ü  ENFARTE E AVC RAROS!

ü  INEXISTÊNCIA DE SOBREPESO, HIPERTENSÃO E SUBNUTRIÇÃO!

ü BAIXOS VALORES DE INSULINA E LEPTINA!

ü MUITAS PESSOAS ALCANÇAM + 75 ANOS!

Lindeberg S et al. J Intern Med 1993;233:269-75; Lindeberg S et al. J Intern Med 1994;236:331-40; Lindeberg S et al. Metabolism 1999;48:1216-19

Simopoulos AP (ed): Evolutionary Aspects of Nutrition and Health.Diet, Exercise, Genetics and Chronic Disease.World Rev Nutr Diet. Basel, Karger, 1999, vol 84, pp 19–73

............................Cereal Grains:Humanity’s Double-Edged Sword

Loren Cordain

Department of Exercise and Sport Science, Colorado State University, Fort Collins,Colo., USA

‘Here is bread, which strengthens man’s heart, and therefore called the staV of life’(Mathew Henry: 1662–1714, Commentary on Psalm 104)

yet,‘Man cannot live on bread alone’ (Bible, Matthew 4:4)

Contents

20 Introduction22 Archaeological Perspective24 Dietary Imbalances of Cereal Grains26 Vitamins A, C and Beta-Carotene27 B Vitamins29 Minerals34 Essential Fatty Acids36 Amino Acids41 Antinutrients in Cereal Grains43 Alkylresorcinols43 Alpha-Amylase Inhibitors44 Protease Inhibitors45 Lectins47 Autoimmune Diseases and Cereal Grain Consumption48 Autoimmunity49 Molecular Mimicry49 Genetic and Anthropological Factors51 Autoimmune Diseases Associated with Cereal Grain Consumption56 Psychological and Neurological Illnesses Associated with Cereal Grain Consumption58 Conclusions60 Acknowledgments60 References

Simopoulos AP (ed): Evolutionary Aspects of Nutrition and Health.Diet, Exercise, Genetics and Chronic Disease.World Rev Nutr Diet. Basel, Karger, 1999, vol 84, pp 19–73

............................Cereal Grains:Humanity’s Double-Edged Sword

Loren Cordain

Department of Exercise and Sport Science, Colorado State University, Fort Collins,Colo., USA

‘Here is bread, which strengthens man’s heart, and therefore called the staV of life’(Mathew Henry: 1662–1714, Commentary on Psalm 104)

yet,‘Man cannot live on bread alone’ (Bible, Matthew 4:4)

Contents

20 Introduction22 Archaeological Perspective24 Dietary Imbalances of Cereal Grains26 Vitamins A, C and Beta-Carotene27 B Vitamins29 Minerals34 Essential Fatty Acids36 Amino Acids41 Antinutrients in Cereal Grains43 Alkylresorcinols43 Alpha-Amylase Inhibitors44 Protease Inhibitors45 Lectins47 Autoimmune Diseases and Cereal Grain Consumption48 Autoimmunity49 Molecular Mimicry49 Genetic and Anthropological Factors51 Autoimmune Diseases Associated with Cereal Grain Consumption56 Psychological and Neurological Illnesses Associated with Cereal Grain Consumption58 Conclusions60 Acknowledgments60 References

ü Glicoproteínas com capacidade para se ligarem de forma reversiva a mono e oligossacarídeos específicos!

ü  São resistentes à cozedura e às enzimas proteoliticas e são reconhecidas como o maior antinutriente dos alimentos.!

Leguminosas

Cereais

LECTINAS

Cordain L, et al. Br J Nutr. 2000 Mar;83(3):207-17 Cordain L. World Rev Nutr Diet 1999; 84:19-73

Adaptado de Cordain L, 2009 (com permissão)

WGA

ü  Gérmen de Trigo: 300 – 350 mg/kg WGA (1)

ü  Farinha de trigo integral: 30-50 mg/kg WGA (2)

ü  Farinha de trigo refinado: 4.4 mg/kg WGA (2)

1. Vincenzi S, et al. J Agric Food Chem. 2002 Oct 23;50(22):6266-70. 2.  Matucci A et al. Food Control 2004;15: 391-95

Enterócitos

Villi

Integridade Hiperpermeabilidade

WGA/PHA

Junções Estreitas

1.  Sjolander A et al. The effect of concanavalin A and wheat germ agglutinin on the ultrastructure and permeability of rat intestine. Int Arch Allergy Appl Immunol 1984; 75, 230–236. 2.  Greer F & Pusztai A. Toxicity of kidney bean (Phaseolus vulgaris) in rats: changes intestinal permeability. Digestion 1985; 32, 42–46.

3.  Pellegrina CD et al. Plant lectins as carriers for oral drugs: Is wheat germ agglutinin a suitable candidate? Toxicol Appl Pharmacol 2005;207:170-78

LUMEN

Adaptado de Cordain L, 2009 (com permissão)

Proc. Nat. Acad. Sci. USAVol. 70, No. 2, pp. 485-489, February 1973

Insulin-Like Activity of Concanavalin A and Wheat Germ Agglutinin-DirectInteractions with Insulin Receptors

(glucose transport/lipolysis/adenylate cyclase/affinity chromatography/lymphocyte transformation/growth factors)

PEDRO CUATRECASAS AND GUY P. E. TELL

Department of Medicine, and Department of Pharmacology and Experimental Therapeutics, The Johns Hopkins UniversitySchool of Medicine, Baltimore, Maryland 21205

Communicated by Saul Roseman, December 7, 1972

ABSTRACT Concanavalin A and wheat germ agglu-tinin are as effective as insulin in enhancing the rate ofglucose transport and in inhibiting epinephrine-stimu-lated lipolysis in isolated adipocytes. These lectins, also likeinsulin, inhibit basal as well as epinephrine-stimulatedadenylate cyclase activity of membranes obtained fromhomogenates of fat cells. Low concentrations of wheatgerm agglutinin enhance the specific binding of insulin toreceptors of fat cells and liver membranes. Higher con-centrations of this plant lectin, as well as of concanavalinA, competitively displace the binding of insulin to recep-tors in these tissues. These effects are equally apparentin insulin-binding proteins solubilized from membranes,indicating that the plant lectins interact directly withinsulin receptors. All of the effects observed with the plantlectins are reversed by simple sugars that bind specificallyto these plant proteins. Agarose derivatives of the plantlectins effectively adsorb solubilized insulin-binding pro-teins, and these can be eluted with buffers containingspecific simple sugars. The possible implications of thesefindings to certain biological properties (mitogenicity) ofthese lectins and to the mechanism of action of othergrowth-promoting substances are considered.

Concanavalin A and wheat germ agglutinin are plant proteinsthat can bind to specific carbohydrate determinants on thesurface of mammalian cells, and they can agglutinate variousnormal and neoplastic animal cells. In addition concanavalin Acan, by unknown mechanisms, stimulate mitosis and blasto-genic transformation of lymphocytes (1), inhibit phagocytosisby polymorphonuclear leukocytes (2), and prevent lympho-cyte cap and patch formation induced by anti-immunoglobulin(3). In this report, we demonstrate that very low concen-

trations of these plant lectins have profound insulin-likeeffects on metabolic processes of isolated fat cells, and thatthey can interact directly with the cell surface receptors forinsulin in these cells. These observations may be of importancein understanding the basis of the biological properties (e.g.,lymphocyte transformation) of these lectins, and they may

also provide a better understanding of the mechanism of actionof insulin.

METHODSIsolated fat cells were prepared from 80-140 g Sprague-Dawley rats (4). The procedures for the preparation of liver(5, 6) and fat cell membranes (7), and for measurement of thespecific binding of [1251 linsulin to cell (8) and to membrane (7)receptors have been described. Insulin receptor proteins were

solubilized from liver membranes with Triton X-100 (6).Glycerol in the incubation medium was determined by themethod of Ryley (9). Adenylate cyclase activity was deter-

485

mined by a modification (10) of the technique of Pohl et al.(11). Fat cell membranes (10) were freshly prepared for eachexperiment by homogenization (Polytron), centrifugation, andsuspension of the pellet in 50 mMI Tris - HCl (pH 7.6); aden-ylate cyclase assays were begun within 10 min. ConcanavalinA (three-times crystallized) was from Miles. Wheat germagglutinin, a gift from Dr. V. Marchesi, was purified byaffinity chromatography and was homogeneous on Nadodecylsulfate disc gel electrophoresis (12).Agarose derivatives of the lectins were prepared by the

CNBr procedure (13), or by reaction with activated N-hydroxysuccinimide esters of diaminodipropylaminosuccinyl-agarose (14). 40 ml of Sepharose 4B was activated with 6 g ofCNBr and reacted with 60 ml of ice-cold 0.1 MI sodium phos-phate buffer (pH 7.4) containing 500 mg of concanavalin Aand 0.1 MI a-methyl-D-mannopyranoside. After 16 hr at 40, 2 gof glycine was added and the incubation was continued for 8hr at 240. This adsorbent contained 5.5 mg of protein per mlof gel. Wheat germ agglutinin (1.4 mg//ml) was similarlycoupled to activated agarose in the presence of 0.1 M N-acetyl-D-glucosamine; 1.1 mg of protein was coupled per ml of gel.'251-labeled plant lectins were used during the couplingprocedures. By phase-contrast microscopy, the lectin-beadsbecame heavily coated with erythrocytes when these weremixed at 240 for 1-2 hr. The cells could be rapidly desorbedwith 50 mM N-acetyl-D-glucosamine.

RESULTSEffects on glucose transport

Concanavalin A and wheat germ agglutinin are very effectivein enhancing the rate of [14C]glucose oxidation in isolated fatcells (Fig. 1). The maximal effects are similar to those thatcan be achieved with insulin. The concentration required forthe half-maximal effect is about 20 nM for concanavalin A(molecular weight: 100,000) and about 4 nMI for wheat germagglutinin (molecular weight: 25,000). These effects resultfrom interactions of the lectins with high-affinity binding sitesthat represent only a fraction of the total number of lectin-binding sites on the fat cells, since direct binding studies withiodinated lectins demonstrate that saturation requires con-centrations greater than 0.1 mg/ml (manuscript in prepara-tion). Fat cells do not agglutinate with either lectin undervarious conditions studied.The increased rates of glucose oxidation induced by con-

canavalin A and by wheat germ agglutinin can be completelyand selectively abolished by addition of the specific sugars

BioMed Central


BMC Endocrine Disorders

Open AccessResearch articleAgrarian diet and diseases of affluence – Do evolutionary novel dietary lectins cause leptin resistance?Tommy Jönsson*1, Stefan Olsson2, Bo Ahrén1, Thorkild C Bøg-Hansen3, Anita Dole3 and Staffan Lindeberg1

Address: 1Department of Clinical Sciences, Lund University, Lund, Sweden, 2Department of Ecology, The Royal Veterinary and Agricultural University, Copenhagen, Denmark and 3Institute of Molecular Pathology, University of Copenhagen, Copenhagen, Denmark

Email: Tommy Jönsson* - [email protected]; Stefan Olsson - [email protected]; Bo Ahrén - [email protected]; Thorkild C Bøg-Hansen - [email protected]; Anita Dole - [email protected]; Staffan Lindeberg - [email protected]* Corresponding author

AbstractBackground: The global pattern of varying prevalence of diseases of affluence, such as obesity,cardiovascular disease and diabetes, suggests that some environmental factor specific to agrariansocieties could initiate these diseases.

Presentation of the hypothesis: We propose that a cereal-based diet could be such anenvironmental factor. Through previous studies in archaeology and molecular evolution weconclude that humans and the human leptin system are not specifically adapted to a cereal-baseddiet, and that leptin resistance associated with diseases of affluence could be a sign of insufficientadaptation to such a diet. We further propose lectins as a cereal constituent with sufficientproperties to cause leptin resistance, either through effects on metabolism central to the properfunctions of the leptin system, and/or directly through binding to human leptin or human leptinreceptor, thereby affecting the function.

Testing the hypothesis: Dietary interventions should compare effects of agrarian and non-agrarian diets on incidence of diseases of affluence, related risk factors and leptin resistance. A non-significant (p = 0.10) increase of cardiovascular mortality was noted in patients advised to eat morewhole-grain cereals. Our lab conducted a study on 24 domestic pigs in which a cereal-free hunter-gatherer diet promoted significantly higher insulin sensitivity, lower diastolic blood pressure andlower C-reactive protein as compared to a cereal-based swine feed. Testing should also evaluatethe effects of grass lectins on the leptin system in vivo by diet interventions, and in vitro in variousleptin and leptin receptor models. Our group currently conducts such studies.

Implications of the hypothesis: If an agrarian diet initiates diseases of affluence it should bepossible to identify the responsible constituents and modify or remove them so as to make anagrarian diet healthier.


BMC Endocrine Disorders 2005, 5:10 doi:10.1186/1472-6823-5-10

Received: 24 June 2005Accepted: 10 December 2005

This article is available from: http://www.biomedcentral.com/1472-6823/5/10

© 2005 Jönsson et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

GLIADINA LIGA-SE A RECEPTOR CXR3

Fasano A. Scientific American. August 2009

RESPOSTA INFLAMATÓRIA

Podolsky. N Engl J Med. 2002 Aug 8;347(6):417-29

TLR4 EM MONÓCITOS DE DIABÉTICOS TIPO II!

Dasu MR. Increased Toll-Like Receptor (TLR) Activation and TLR Ligands in Recently Diagnosed Type 2 Diabetic Subjects. Diabetes Care 33:861–868, 2010

NUTRITION MYTHS

OBESITY 1

nature publishing group ARTICLESINTERVENTION AND PREVENTION

INTRODUCTIONManipulation of physiological pathways in order to reduce obesity and symptoms of the metabolic syndrome is a major focus of research worldwide. Recent data show that adipose tissue, the energy storage site of the body, is also an endocrine organ that synthesizes and secretes a variety of adipocytokines. !is includes hormones that regulate hunger and satiety as well as those associated with the development of insulin resistance, the metabolic syndrome and in"ammation (1).

Leptin “the satiety hormone” has been described as the “information provider” of adipose tissue status to receptors in the brain. In short term, it regulates hunger, satiety, and food intake (1–3). Previous studies have described a typical diurnal pattern of leptin secretion that falls during the day from 0800 to 1600 hours, reaching a nadir at 1300 hours and increases from 1600 with a zenith at 0100 hours (4,5). Ironically, this

crucial hormone responsible for satiety is at its highest levels when individuals are sleeping.

Adiponectin is considered to be “the link between obesity, insulin resistance, and the metabolic syndrome” (6). Adiponectin plays a role in energy regulation as well as in lipid and carbohy-drate metabolism, reducing serum glucose and lipids, improving insulin sensitivity and having an anti-in"ammatory e#ect (7). Adiponectin’s diurnal secretion pattern has been described in obese individuals (particularly with abdominal obesity), as low throughout the day. In normal weight subjects or overweight subjects following weight loss, a general increase in adiponectin concentrations is detected as well as a rise in the diurnal pattern during the daytime, with zeniths at 1100 and 0100 hours and a decline at night, reaching a nadir at 0400 hours (5,8).

Innovative dietary regimens that will be able to modify these hormonal secretion patterns may be bene$cial to people

Greater Weight Loss and Hormonal Changes After 6 Months Diet With Carbohydrates Eaten Mostly at DinnerSigal Sofer1,2, Abraham Eliraz1, Sara Kaplan2, Hillary Voet1, Gershon Fink3, Tzadok Kima4 and Zecharia Madar1

This study was designed to investigate the effect of a low-calorie diet with carbohydrates eaten mostly at dinner on anthropometric, hunger/satiety, biochemical, and inflammatory parameters. Hormonal secretions were also evaluated. Seventy-eight police officers (BMI >30) were randomly assigned to experimental (carbohydrates eaten mostly at dinner) or control weight loss diets for 6 months. On day 0, 7, 90, and 180 blood samples and hunger scores were collected every 4 h from 0800 to 2000 hours. Anthropometric measurements were collected throughout the study. Greater weight loss, abdominal circumference, and body fat mass reductions were observed in the experimental diet in comparison to controls. Hunger scores were lower and greater improvements in fasting glucose, average daily insulin concentrations, and homeostasis model assessment for insulin resistance (HOMAIR), T-cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, C-reactive protein (CRP), tumor necrosis factor- (TNF- ), and interleukin-6 (IL-6) levels were observed in comparison to controls. The experimental diet modified daily leptin and adiponectin concentrations compared to those observed at baseline and to a control diet. A simple dietary manipulation of carbohydrate distribution appears to have additional benefits when compared to a conventional weight loss diet in individuals suffering from obesity. It might also be beneficial for individuals suffering from insulin resistance and the metabolic syndrome. Further research is required to confirm and clarify the mechanisms by which this relatively simple diet approach enhances satiety, leads to better anthropometric outcomes, and achieves improved metabolic response, compared to a more conventional dietary approach.

Obesity (2011) doi:10.1038/oby.2011.48

1The Robert H. Smith Faculty of Agriculture, Food and Environment, Institute of Biochemistry and Food Science, The Hebrew University of Jerusalem, Rehovot, Israel; 2Meuhedet Medical Services, Diet and Nutrition Department, Israel; 3Kaplan Medical Center, Rehovot, Israel; 4Israeli Police Force, Tel Aviv District, Israel. Correspondence: Zecharia Madar ([email protected])

Received 28 June 2010; accepted 29 January 2011; advance online publication 7 April 2011. doi:10.1038/oby.2011.48Obesity 19, 2006-2014 (October 2011)

OBESITY 1









Obesity (2011) doi:10.1038/oby.2011.48


Received 28 June 2010; accepted 29 January 2011; advance online publication 7 April 2011. doi:10.1038/oby.2011.48

OBESITY 1









Obesity (2011) doi:10.1038/oby.2011.48


Received 28 June 2010; accepted 29 January 2011; advance online publication 7 April 2011. doi:10.1038/oby.2011.48

OBESITY 5

ARTICLESINTERVENTION AND PREVENTION

Table 3 Changes in anthropometric parameters after 6 months on diet

Units Experimental group (n = 30) Control group (n = 33) Comparison of groups

Weight loss

(kg) 11.6 ± 0.84* 9.06 ± 0.84* P = 0.024

(%) 11.7 ± 0.66* 9.96 ± 0.79* P = 0.053

BMI reduction

Original (g/m2) 3.99 ± 0.24* 3.16 ± 0.27*

Adjusted for baseline differences (g/m2) 3.85 ± 0.25* 3.28 ± 0.24* P = 0.115

(%) 11.7 ± 0.66* 9.68 ± 0.79* P = 0.053

Abdominal circumference decrease

Original (cm) 11.7 ± 0.89* 9.39 ± 0.98*

Adjusted for baseline differences (cm) 11.1 ± 0.92* 10.0 ± 0.88* P = 0.408

(%) 10.5 ± 0.70* 8.80 ± 0.90* P = 0.159

Body fat percent reduction

Absolute (%) 6.98 ± 0.95* 5.13 ± 0.59* P = 0.710

Relative (%) 18.1 ± 2.45* 14.1 ± 1.71* P = 0.122

Mean ± s.e. Analysis by two-factor ANOVA.*Significant difference from day 0 (P < 0.0001).

150%

100%

50%

99.9%94.1%

103.7%

113.7%

Day 90 Day 180

Experimental diet

Control diet

*

Cha

nge

in H

-SS

c

a

150%

100%

50%

Experimental diet

Control diet

Experimental diet

Control diet

Cha

nge

in H

-SS

c

150%

100%

50%

Cha

nge

in H

-SS

c

Day 90 Day 180

Morning Noon Afternoon Evening Morning Noon Afternoon Evening

80.7% 77.6%

93.4%

125.1% 128.0%127.7%

**

**

#

*

b

Figure 2 Hunger and satiety scales. (a) Least square mean ± s.e. hunger-satiety scores (H-SSc) on day 90 and day 180 as a percentage of baseline (average daily satiety on day 0 and 7) in the experimental (n = 18) and the control (n = 21) groups. Comparison of groups by repeated measures ANOVA. *P < 0.05 for difference from baseline. (b) Mean ± s.e. percent of H-SSc at day 90 and at day 180 compared to scores at parallel hours on the first week of the diet (average day 0 and day 7). *P < 0.05 as compared to the same hour in the first week. #P = 0.030 comparing control and experimental groups by contrast t-test following repeated measures ANOVA at day 180.

Obesity 19, 2006-2014 (October 2011)

OBESITY 1









Obesity (2011) doi:10.1038/oby.2011.48


Received 28 June 2010; accepted 29 January 2011; advance online publication 7 April 2011. doi:10.1038/oby.2011.48Obesity 19, 2006-2014 (October 2011)

OBESITY 7

ARTICLESINTERVENTION AND PREVENTION

Serum inflammatory parameters levelMeasurements of in!ammatory markers are shown in Table 4. A trend of a greater CRP reduction was observed in the experi-mental group (27.8 vs. 5.8%). Signi"cant di#erences were not achieved a$er adjusting for baseline di#erences. On day 180, subjects on the experimental diet had signi"cantly lower TNF-% concentration, with a 9.2% decrease from baseline measure-ments. In contrast, the control diet led to a 16.1% increase in TNF-% (P = 0.034 for di#erence between groups). Both diets low-ered IL-6 concentrations at day 90 and 180 compared to baseline. At day 180, the experimental diet led to a signi"cant reduction of 37.8% (P < 0.01) whereas a smaller insigni"cant reduction of 23.7% was found in the control diet. On day 90 a similar trend was observed (15.8% vs. 10% reduction from baseline, respectively).

Serum hormonal levelsBoth diets decreased average 12-h leptin concentrations on day 90 and day 180 compared to baseline (P < 0.05) (Figure 3a).

A trend to smaller reductions from baseline was observed in the experimental group (29.3 and 20.6% decrease in the exper-imental group, respectively, 31.4 and 26.2% decrease in the control group, respectively).

&e experimental diet led to a signi"cant increase (43.5%, P < 0.05) in average 12-h adiponectin concentrations, whereas the control diet led to a smaller and insigni"cant (13.9%) increase a$er 180 days (Figure 3b). &e same trend was observed on day 90 (15.3 vs. 1.9%, respectively).

DISCUSSION&is randomized clinical trial, performed in a sample of police o'cers with BMI >30, examined the e#ects of a low-calorie diet based on carbohydrates eaten mostly at dinner, in compar-ison to an identical low-calorie diet providing carbohydrates throughout the day.

Greater weight loss, abdominal circumference, and body fat mass reductions were observed in the experimental diet in comparison to controls (Table 3). &e experimental diet group’s H-SScs were higher in comparison to baseline (Figure 2). A$er 180 days, a drop in averaged 12-h leptin concentrations was observed in both diet groups. A trend to smaller reduc-tion in averaged 12-h leptin concentrations from baseline was observed in the experimental group (Figure 3a). &e decrease observed in overall daily leptin concentrations for both groups has been documented in previous studies (2,14–16) and may be explained by reduced body fat mass (Table 3). Reduced lev-els of leptin during weight loss programs is commonly associ-ated with a decline in satiety levels (14,15) as was observed in our control group. In the experimental group, this expected satiety reduction did not occur. On the contrary, at the end of the study, the experimental diet group had higher H-SSc in comparison to baseline.

It is proposed that the smaller reduction in averaged 12-h leptin concentration, induced by the experimental diet, may be an important factor in the higher levels of satiety reported during the day. Previous studies with di#erent diets reported that during weight loss, leptin concentrations decreased, sati-ety levels were reduced, food intake renewed and a slow regain of body weight occurred (14,15,17). &us, dietary manipula-tions that will maintain higher daytime leptin concentrations during daylight hours in weight loss process may be bene"cial. Our experimental diet might manipulate daily leptin secre-tion, leading to higher relative concentrations throughout the day. We propose that this modi"cation of hormone secretion helped participants experience greater satiety during waking hours, enhance diet maintenance over time and have better anthropometric outcomes.

Although, no speci"c nutritional guidance regarding glucose balance, lipids pro"les or in!ammation status was given to par-ticipants, improvements in these parameters were observed. It is of great interest that for nearly all of these parameters, signif-icantly greater improvements were observed in the experimen-tal diet group (Table 4). Signi"cantly higher improvements of glucose balance and insulin resistance (HOMAIR), lipid pro"le (total cholesterol, LDL-cholesterol, HDL-cholesterol) and the

35

30

25

20

15

10

5

0

70.7% 68.6%79.4% 73.8%

Experimental diet

Control diet

Lept

in (

ng/m

l)

Day 0 Day 90 Day 180

* ** *

a

Day 0 Day 90 Day 180

Experimental diet

Control diet

Adi

pone

ctin

(ng

/ml)

*80

70

60

50

40

30

20

10

0

115.3%

101.9%

143.5%

113.9%

b

Figure 3 Mean ± s.e. for absolute values, least squares mean for percentage of baseline (shown in boxes on bars) in the experimental (n = 18) and the control (n = 21) groups. Average daily (a) leptin and (b) adiponectin at days 0, 90, and 180. Comparison of groups for percentages of baseline by repeated measures ANOVA. *P < 0.05 for difference from baseline by t-test using standard errors from ANOVA. Percentage of baseline values were calculated for each subject and averaged.

High-glycaemic index and -glycaemic load meals increase the availabilityof tryptophan in healthy volunteers

Christopher P. Herrera1*†, Keir Smith2, Fiona Atkinson2, Patricia Ruell1, Chin Moi Chow1,Helen O’Connor1 and Jennie Brand-Miller2,3

1Discipline of Exercise and Sport Science, Faculty of Health Sciences, University of Sydney, Sydney, NSW, Australia2School of Molecular Bioscience, Faculty of Science, University of Sydney, Sydney, NSW, Australia3Boden Institute of Obesity, Nutrition, and Exercise, University of Sydney, Sydney, NSW, Australia

(Received 12 July 2010 – Revised 29 October 2010 – Accepted 12 November 2010)

AbstractThe purpose of the present study was to determine the influence of the glycaemic index (GI) and glycaemic load (GL) on the ratioof tryptophan (TRP) relative to other large neutral amino acids (LNAA). Ten healthy men (age 22·9 (SD 3·4) years; BMI 23·5(SD 1·6) kg/m2) underwent standard GI testing, and later consumed each of a mixed-macronutrient (1915 kJ; 66·5% carbohydrate(CHO), 17% protein and 16·5% fat) high-GI (MHGI), an isoenergetic, mixed-macronutrient low-GI (MLGI) and a CHO-only (3212 kJ;90% CHO, 8% protein, 2% fat) high-GI (CHGI) meal on separate days. The GI, GL and insulin index values (e.g. area under thecurve) were largest after the CHGI meal (117, 200, 158), followed by the MHGI (79, 59, 82) and MLGI (51, 38, 56) meals, respectively(all values were significantly different, P,0·05). After the MHGI and MLGI meals but not after the CHGI meal, TRP was elevated at 120and 180 min (P,0·05). After the CHGI, LNAA was lower compared with the MLGI (P,0·05); also the rate of decline in LNAA washigher after CHGI compared with MHGI and MLGI (both comparisons P,0·05). The percentage increase from baseline in TRP:LNAAafter CHGI (23%) was only marginally higher than after the MHGI meal (17%; P!0·38), but it was threefold and nearly significantly greaterthan MLGI (8%; P!0·05). The present study demonstrates that the postprandial rise in TRP:LNAA was increased by additional CHOingestion and higher GI. Therefore, the meal GL appears to be an important factor influencing the postprandial TRP:LNAA concentration.

Key words: Glycaemic index: Glycaemic load: Tryptophan: Amino acids: Carbohydrate

The availability of dietary tryptophan (TRP) to the braindepends on its concentration relative to other large neutralamino acids (LNAA), which compete for a common trans-port mechanism across the blood–brain barrier(1).Increased consumption of carbohydrate (CHO) elicits amarked demand in insulin secretion, which enhances per-ipheral, skeletal muscle uptake of LNAA. However, TRP islargely albumin bound and therefore protected from thisabsorption(2). Previous studies have shown that the post-prandial concentration of TRP:LNAA increases between20 and 50% compared with baseline after predominatelyCHO-rich meals(3–5). In one study, consumption of asucrose-based food elicited a larger postprandial increasein TRP:LNAA compared with a raw starch-based food(3).Although the glycaemic index (GI) of the foods was notmeasured, these authors hypothesised that a high-GICHO-based meal compared with a low-GI CHO-based

meal would elicit a greater insulin release, and thereforea larger postprandial rise in plasma TRP:LNAA levels(3,6,7).

The proportion of CHO relative to either protein or fat maysimilarly influence the TRP:LNAA response after mixed-macronutrientmeals. Berry et al.(8) suggested that ameal con-taining a CHO:protein ratio of approximately 5:1 wouldneither raise nor lower postprandial LNAA, given that theLNAA-lowering effect of the insulin demand would be offsetby the contribution of LNAA provided by the proteinsource. Since TRP is the least abundant amino acid foundin protein, a protein-rich, low-CHO meal lowers TRP:LNAAdue to a greater contribution of LNAA relative to TRP inthe meal(3). Conversely, the addition of fat to a meal retardsgastric emptying(9), and lowers the peak glycaemic and insu-lin response(10). Thus, increased fat content in a meal mayattenuate the postprandial TRP:LNAA response after mixed-macronutrient meals when compared with a CHO-only meal.

† Present address: Research and Education Centre, ASPETAR – Qatar Orthopaedic and Sports Medicine Hospital, PO Box 29 222, Doha, Qatar.

*Corresponding author: Dr C. P. Herrera, fax "974 4413 2020, email [email protected]

Abbreviations: CHGI, carbohydrate-only high glycaemic index; CHO, carbohydrate; GI, glycaemic index; GL, glycaemic load; LNAA, large neutral amino

acids; MHGI, mixed-macronutrient high glycaemic index; MLGI, mixed-macronutrient low glycaemic index; TRP, tryptophan.

British Journal of Nutrition (2011), page 1 of 7 doi:10.1017/S0007114510005192q The Authors 2011

British

JournalofNutrition

Recently, the GI of a CHO-rich meal has been shown tosignificantly influence sleep initiation in healthy sleepers.Afaghi et al.(11) reported a 50% reduction in the timerequired to fall asleep after a CHO-rich, high-GI meal com-pared with an isoenergetic low-GI meal when consumed4 h before bedtime. This improvement in sleep initiationafter a high-GI CHO meal was probably due to anincreased plasma TRP:LNAA concentration and increasedserotonin; however, biochemical data were not collectedto support this claim.

Therefore, the purpose of the present study was toinvestigate the influence of the GI and glycaemicload (GL) on the TRP:LNAA response after a CHO-onlyhigh-GI (CHGI) meal compared with two isoenergetic,mixed-macronutrient high- and low-GI (MHGI and MLGI)meals. We hypothesise that the glycaemic and insulinresponse to the CHGI meal will be larger than the MHGIand MLGI meals, and the corresponding increase inTRP:LNAA will be proportional to the GL of the mealsdue to differences in insulin demand.

Methods and materials

Participants

Ten healthy mixed-ethnicity men (age 22·9 (SD 3·4) years)of normal weight (BMI 23·5 (SD 1·6) kg/m2) were recruitedfrom a university student population. Exclusion criteriaincluded a self-reported current or past history of medical,psychiatric or sleep disorder, current use of prescribedmedication, recreational drug use, allergy related to thestudy meals or habitual use of a restrictive diet. The presentstudy was conducted according to the guidelines laiddown in the Declaration of Helsinki, and all proceduresinvolving human subjects were approved by the HumanResearch Ethics Committee of the University of Sydney.Written informed consent was obtained from all subjectsbefore participation.

Meals

The energy and macronutrient composition of the meals aresummarised in Table 1. The CHGI meal, 3212 kJ, was repli-cated from a previous study(11) and consisted of a large

portion of rice (Jasmine GI approximately 109; RivianaFoods, Sydney, NSW, Australia) served with a tomato-based vegetable puree. The mixed-macronutrient meals(MHGI and MLGI) were isoenergetic, approximately1915 kJ, and consisted of rice (MHGI: Jasmine GI approxi-mately 109; MLGI: Doongara GI approximately 46; RivianaFoods) served with a sachet of chicken with sun-driedtomato sauce (965 kJ; 7·9 g fat, 14·1 g protein and 24·7 gCHO;Lean Cuisinee; Nestle Australia Limited, Rhodes,NSW, Australia). Meals were prepared in the University ofSydney Human Nutrition Unit kitchens. Uncooked rice(raw weight of CHGI, 200 g; MHGI, 64·7 g; MLGI, 64·5 g)was prepared using an electric rice cooker before thetesting day, with a rice:water ratio of 1:1·5. Cooked ricewas frozen (2208C) in individual portions and reheatedin a microwave before serving. Frozen Lean Cuisineesachets were heated in the microwave according to themanufacturer’s recommendations and poured over therice immediately before serving. All meals were presentedto the participants with 250ml of cool water and consumedwithin 15min. Participants were required to consume theentire meal, which was assessed visually by a researcher(C. P. H. and K. S.) and by weighing the plate before andafter eating.

Procedure

Participants presented to the testing facility in the morningby at least 10.00 hours, having fasted overnight for a periodof at least 8 h. Participants were required to avoid vigorousexercise and abstain from alcohol for at least 24 h beforetesting, and were instructed to avoid over- or under-eating. Smokers were instructed to abstain on the morningbefore testing. On the night before testing, participantswere instructed to consume a high-CHO, low-fat eveningmeal devoid of legumes in order to avoid extremehunger and variation in basal blood glucose concentration.Self-reported compliance with these instructions was eval-uated by a researcher (C. P. H. and K. S.) each morningbefore testing.

Participants initially completed three independentreference glucose tests separated by at least 48 hapart. Each reference test used a standard glucose drink

Table 1. Energy and macronutrient composition of the carbohydrate (CHO) high-glycaemic index (CHGI), mixed-macro-nutrient high-glycaemic index (MHGI) and mixed-macronutrient low-glycaemic index (MLGI) meals

Fat Protein CHO

Meal Energy (kJ) g Energy (%) g Energy (%) g Energy (%) GI* GL†

CHGI‡ 3212 0·4 1·6 16·8 8 171·4 90·4 117 200MHGI 1916 7·9 16·2 18·2 17·2 75 66·6 79 59MLGI 1913 7·9 16·1 18·6 17·5 75 66·4 51 38

GI, glycaemic index; GL, glycaemic load.* GI was determined using the average glucose response (n 9) and the incremental area under the curve method(12)).† GL was calculated by multiplying each meal GI by the available CHO (g).‡ GI for the CHGI meal was approximate due to a larger CHO content of this test meal compared with the reference glucose drink providing

75 g CHO(12).

C. P. Herrera et al.2

British

JournalofNutrition

High-glycaemic index and -glycaemic load meals increase the availabilityof tryptophan in healthy volunteers

Christopher P. Herrera1*†, Keir Smith2, Fiona Atkinson2, Patricia Ruell1, Chin Moi Chow1,Helen O’Connor1 and Jennie Brand-Miller2,3

1Discipline of Exercise and Sport Science, Faculty of Health Sciences, University of Sydney, Sydney, NSW, Australia2School of Molecular Bioscience, Faculty of Science, University of Sydney, Sydney, NSW, Australia3Boden Institute of Obesity, Nutrition, and Exercise, University of Sydney, Sydney, NSW, Australia

(Received 12 July 2010 – Revised 29 October 2010 – Accepted 12 November 2010)

AbstractThe purpose of the present study was to determine the influence of the glycaemic index (GI) and glycaemic load (GL) on the ratioof tryptophan (TRP) relative to other large neutral amino acids (LNAA). Ten healthy men (age 22·9 (SD 3·4) years; BMI 23·5(SD 1·6) kg/m2) underwent standard GI testing, and later consumed each of a mixed-macronutrient (1915 kJ; 66·5% carbohydrate(CHO), 17% protein and 16·5% fat) high-GI (MHGI), an isoenergetic, mixed-macronutrient low-GI (MLGI) and a CHO-only (3212 kJ;90% CHO, 8% protein, 2% fat) high-GI (CHGI) meal on separate days. The GI, GL and insulin index values (e.g. area under thecurve) were largest after the CHGI meal (117, 200, 158), followed by the MHGI (79, 59, 82) and MLGI (51, 38, 56) meals, respectively(all values were significantly different, P,0·05). After the MHGI and MLGI meals but not after the CHGI meal, TRP was elevated at 120and 180 min (P,0·05). After the CHGI, LNAA was lower compared with the MLGI (P,0·05); also the rate of decline in LNAA washigher after CHGI compared with MHGI and MLGI (both comparisons P,0·05). The percentage increase from baseline in TRP:LNAAafter CHGI (23%) was only marginally higher than after the MHGI meal (17%; P!0·38), but it was threefold and nearly significantly greaterthan MLGI (8%; P!0·05). The present study demonstrates that the postprandial rise in TRP:LNAA was increased by additional CHOingestion and higher GI. Therefore, the meal GL appears to be an important factor influencing the postprandial TRP:LNAA concentration.

Key words: Glycaemic index: Glycaemic load: Tryptophan: Amino acids: Carbohydrate

The availability of dietary tryptophan (TRP) to the braindepends on its concentration relative to other large neutralamino acids (LNAA), which compete for a common trans-port mechanism across the blood–brain barrier(1).Increased consumption of carbohydrate (CHO) elicits amarked demand in insulin secretion, which enhances per-ipheral, skeletal muscle uptake of LNAA. However, TRP islargely albumin bound and therefore protected from thisabsorption(2). Previous studies have shown that the post-prandial concentration of TRP:LNAA increases between20 and 50% compared with baseline after predominatelyCHO-rich meals(3–5). In one study, consumption of asucrose-based food elicited a larger postprandial increasein TRP:LNAA compared with a raw starch-based food(3).Although the glycaemic index (GI) of the foods was notmeasured, these authors hypothesised that a high-GICHO-based meal compared with a low-GI CHO-based

meal would elicit a greater insulin release, and thereforea larger postprandial rise in plasma TRP:LNAA levels(3,6,7).

The proportion of CHO relative to either protein or fat maysimilarly influence the TRP:LNAA response after mixed-macronutrientmeals. Berry et al.(8) suggested that ameal con-taining a CHO:protein ratio of approximately 5:1 wouldneither raise nor lower postprandial LNAA, given that theLNAA-lowering effect of the insulin demand would be offsetby the contribution of LNAA provided by the proteinsource. Since TRP is the least abundant amino acid foundin protein, a protein-rich, low-CHO meal lowers TRP:LNAAdue to a greater contribution of LNAA relative to TRP inthe meal(3). Conversely, the addition of fat to a meal retardsgastric emptying(9), and lowers the peak glycaemic and insu-lin response(10). Thus, increased fat content in a meal mayattenuate the postprandial TRP:LNAA response after mixed-macronutrient meals when compared with a CHO-only meal.

† Present address: Research and Education Centre, ASPETAR – Qatar Orthopaedic and Sports Medicine Hospital, PO Box 29 222, Doha, Qatar.

*Corresponding author: Dr C. P. Herrera, fax "974 4413 2020, email [email protected]

Abbreviations: CHGI, carbohydrate-only high glycaemic index; CHO, carbohydrate; GI, glycaemic index; GL, glycaemic load; LNAA, large neutral amino

acids; MHGI, mixed-macronutrient high glycaemic index; MLGI, mixed-macronutrient low glycaemic index; TRP, tryptophan.

British Journal of Nutrition (2011), page 1 of 7 doi:10.1017/S0007114510005192q The Authors 2011

British

JournalofNutrition

CHGI meal. In addition, there was an initial increase in TRPafter the MHGI and MLGI meals, whereas after the CHGImeal, changes in TRP were unremarkable. Conversely,the rate of decline in LNAA was markedly higher afterthe CHGI meal compared with the MHGI and MLGImeals. As a result, there was a substantial postprandialrise in TRP:LNAA after the CHGI meal, which was slightlyattenuated after the MHGI meal, and relatively small afterthe MLGI meal.

The decrease in the glycaemic and insulin response ofthe MHGI and MLGI meals confirms our initial hypothesisand demonstrates substantial clinical improvement com-pared to the CHGI meal. The GL of the MHGI (GL 59)and MLGI (GL 38) was significantly lower than the CHGImeal (GL 200). Epidemiological studies have linked high-GI and -GL diets to increased risk of CVD and for thedevelopment of type 2 diabetes(19,20). A GL value of.120 over a single day is regarded as high(21). Hence,this study confirms that the GL of the CHGI meal wasover 1·5 times greater than what is clinically recommendedfor an entire day(11). Conversely, given regular consump-tion of the MHGI or MLGI as an evening meal, the cumu-lative daily GL could be kept below high levels. Inaddition, the MHGI and MLGI meals were similar in taste,had higher palatability ratings and were approximately50% less in energy relative to the CHGI meal (Table 1).Moreover, the mixed-macronutrient profile of the MHGIand MLGI meals is consistent with Western dietary guide-lines(22,23). Thus, it is likely that these meals would beeasily accepted as part of a regular diet in future studies.

The biochemical results of the study suggest that asimple manipulation of the amount and type of CHO in asingle meal can lead to a substantial change in postprandialamino acid concentration. Given that the macronutrient

composition of the MHGI and MLGI meals was identicaland each contained the same protein source (e.g. chickenbreast sachet), it is likely that these meals contributedan identical amount of dietary TRP, and that this amountwas greater than the CHGI meal, which comprisedmainly rice and vegetables(24). Indeed, there was a signifi-cant increase in TRP concentration at 120 and 180 min afterthe MHGI and MLGI meals but not after the CHGI meal(Table 2). Conversely, there was an overall decline frombaseline in LNAA at the end of the study period after allmeals (240 min; Table 2), and a higher rate of change inLNAA after the CHGI meal compared with the MHGI andMLGI meals. These findings together suggest that the post-prandial decline in LNAA was largely dependent on thetype (high-GI v. low-GI rice) and amount of CHO (75 gin the mixed-macronutrient meals v. 171 g in CHGI meal)in the meals (Table 1). Furthermore, the decline in LNAApresented after the MHGI and MLGI meals indicates thata CHO:protein ratio of approximately 4:1 results in asignificant uptake of LNAA into skeletal muscle. Thus,the present study suggests that the postprandial changein LNAA is largely dependent on the meal GL and not onthe ratio of CHO:protein, as suggested earlier to beresponsible for the change in LNAA(8).

The present data further indicate that the increase inpostprandial TRP:LNAA was inversely proportional to themagnitude and rate of decline in LNAA. In fact, theTRP:LNAA response after the CHGI meal is in line withthose from previous studies(3–5). Importantly, the post-prandial concentration of TRP:LNAA dictates the synthesisof central nervous system serotonin(25), a key neuro-transmitter involved in the regulation of sleep(26). In aseparate study, the CHGI meal was previously shown toimprove sleep initiation when provided 4 h before bedtimecompared with an isoenergetic low-GI version(11). Further-more, Blum et al. reported that the consumption of asimilar-sized CHO breakfast meal resulted in a 461%increase from baseline in the postprandial insulin responseand a 3·5-fold increase in platelet poor plasma serotoninin healthy subjects(27). Thus, the peak percentage changefrom baseline in insulin after the CHGI meal (650%) indi-cates that postprandial serotonin was probably increased,which may therefore explain the improvement in sleepinitiation after the CHGI meal(11). In addition, given thepeak insulin response to the MHGI meal (450%), it remainsa possibility that such a mixed-macronutrient meal maysimilarly improve sleep initiation(4,28). However, giventhe reduced TRP:LNAA response, we may speculate thatthis improvement would be present only when sleeppropensity is high, such as in the post-lunch period orin the late evening(29–31).

There are several limitations with the present study.It was difficult to estimate a priori the sample size requiredto identify meal difference in amino acid concentrationdue to a limited number of published data usingmixed-macronutrient meals. Future studies should inflate

2401801200

Time (min)

5

10

15

20

25

0

Cha

nge

from

bas

elin

e (%

)

Fig. 2. Effects of the carbohydrate-only high-glycaemic index (CHGI), mixed-macronutrient high-glycaemic index (MHGI) and mixed macronutrient lowglycaemic index (MLGI) meals on postprandial tryptophan (TRP)/large neu-tral amino acid (LNAA) concentrations in healthy participants. There was asignificant postprandial rise after each meal with a peak percentage increasebetween 180 and 240min after meal consumption. The corresponding peakpercentage rise was approximately 23, 17 and 8% after the CHGI, MHGIand MLGI meals, respectively. The incremental area under the curve dataindicate that TRP:LNAA levels after the CHGI meal were greater than afterthe MLGI meal (two-tailed t test, P!0·054; one-tailed t test, P!0·03).V, CHGI; B, MHGI; X, MLGI.

Meal composition and tryptophan 5

British

JournalofNutrition

Tryp/LNAA

High-glycemic-index carbohydrate meals shorten sleep onset1–3

Ahmad Afaghi, Helen O’Connor, and Chin Moi Chow

ABSTRACTBackground: Dietary carbohydrate intake has been shown to in-crease the plasma concentration of tryptophan, a precursor of sero-tonin and sleep-inducing agent.Objective: To investigate the role of carbohydrate in sleep induc-tion, we explored the effect of glycemic index (GI) and meal time onsleep in healthy volunteers.Design: We compared the effect of high- and low-GI carbohydrate–based meals ingested 4 h before bedtime on sleep quality. We alsoevaluated the effect of the timing of high-GI meals (4 h comparedwith 1 h) on sleep quality. Twelve healthy men (aged 18-35 y) wereadministered standard, isocaloric (3212 kJ; 8% of energy as protein,1.6% of energy as fat, and 90.4% of energy as carbohydrate) mealsof either Mahatma (low GI ! 50) or Jasmine (high GI ! 109) rice 4 hbefore their usual bedtime. On another occasion, the high-GI mealwas given 1 h before bedtime. The participants underwent a famil-iarization night followed by 3 test nights in random order 1 wk apart.Results: A significant (P ! 0.009) reduction in the mean ("SD)sleep onset latency (SOL) was observed with a high-GI (9.0 " 6.2min) compared with a low-GI (17.5 " 6.2 min) meal consumed 4 hbefore bedtime. The high-GI meal given 4 h before bedtime showeda significantly shortened SOL compared with the same meal given1 h before bedtime (9.0 " 6.2 min compared with 14.6 " 9.9 min;P ! 0.01). No effects on other sleep variables were observed.Conclusions: We showed that a carbohydrate-based high-GI mealresulted in a significant shortening of SOL in healthy sleepers com-pared with a low-GI meal and was most effective when consumed 4 hbefore bedtime. The relevance of these findings to persons with sleepdisturbance should be determined in future trials. Am J ClinNutr 2007;85:426–30.

KEY WORDS Carbohydrates, glycemic index, sleep quality,sleep timing

INTRODUCTION

Common sleep difficulties include sleep initiation and main-tenance. According to The Gallup Organization, 49% of adults inthe United States do not sleep well for #5 nights/mo, 10–40%have intermittent insomnia, and 10 –15% have long-termsleep difficulties (1). In Australia, a survey reported an in-somnia prevalence of 17% in men and 25% in women in anurban community (2).

The current treatment options for insomnia are pharmacother-apy and cognitive behavioral therapy. Treatments are consideredeffective if they shorten sleep onset latency (SOL) or increasetotal sleep time by 30 min (3). Cognitive behavioral therapy is

considered the best practice. Other popular remedies used to treatsleep difficulties include prescribed sedatives and tranquilizers,herbal extracts and complimentary medicines, massage and re-laxation techniques, regular physical activity, and avoidance ofstimulants such as caffeine before sleeping.

Both the timing (4, 5) and macronutrient content (6–9) ofmeals are known to influence sleep. A meal consumed close tobedtime is associated with sleep disturbance (4). A number ofmacronutrients influence sleep through tryptophan (Trp), whichserves as a precursor for brain serotonin, a sleep-inducing agent(10, 11). A factor that promotes the entry of Trp into the brain isits plasma concentration relative to that of the other large neutralamino acids (LNAAs: tyrosine, phenylalanine, leucine, isoleu-cine, valine, and methionine) (12). It is now known that high-glycemic-index (GI) carbohydrates have the ability to increasethe ratio of circulating Trp to LNAAs (Trp:LNAA) via a directaction of insulin, which promotes a selective muscle uptake ofLNAAs (13). Thus, a high-GI meal would be expected to pro-mote sleep via an increase in brain Trp and serotonin as theplasma Trp:LNAA increases (12). It would also be expected thata meal containing a high protein content, which contributes lessTrp to the circulating blood compared with the other LNAAs (12)and thus a lower plasma Trp:LNAA, would reduce serotonin.Serotonin function may be measured indirectly through changesin melatonin concentrations, because serotonin is an intermedi-ary product in the production of melatonin, a pineal hormone(14). Urinary 6-hydroxymelatonin sulfate, a stable end productof melatonin, is often used as a surrogate measure of melatonin,given their linear relation (14).

Therefore, the aim of the present study was to investigate therole of carbohydrate in inducing sleep, and specifically the effectof GI on sleep patterns in healthy sleepers. We hypothesized thatcarbohydrate-based high- compared with low-GI meals ingested4 h before bedtime would improve sleep quality because of agreater insulin response and that the timing of the high-GI meal(4 h compared with 1 h) before bedtime would also influencesleep quality.

1 From the School of Exercise and Sport Science, Faculty of Health Sci-ences, The University of Sydney, Sydney, Australia.

2 Supported by Sydney University’s PhD student research budget. The ricewas provided by Riviana Food Pty Ltd, Victoria, Australia.

3 Reprints not available. Address correspondence to CM Chow, DeltaSleep Research Unit, School of Exercise and Sport Science, Faculty of HealthSciences, The University of Sydney, PO Box 170, Lidcombe NSW 1825Australia. E-mail: [email protected].

Received June 8, 2006.Accepted for publication October 10, 2006.

426 Am J Clin Nutr 2007;85:426–30. Printed in USA. © 2007 American Society for Nutrition

by guest on March 26, 2011


bedtime (2.3 ! 0.6) than they did after the low-GI meal ingested4 h before bedtime (2.1 ! 0.3, P " 0.1). However, they weresignificantly sleepier after the high-GI meal ingested 4 h beforebedtime than after that ingested 1 h before bedtime (1.9 ! 0.5,P " 0.04).

Hunger and fullness scale

The VAS rating of hunger or fullness confirmed that the largerice serving given was adequate so that subjects were not hungryafter the meal. Respective ratings immediately after the meal andat bedtime were 3.21 and 2.17 for the low-GI meal ingested 4 hbefore bedtime; 3.17 and 2.13 for high-GI meal ingested 4 hbefore bedtime; and 3.17 and 2.25 for the high-GI meal ingested

1 h before bedtime. On a 4-point scale, a rating of 3 indicates thatthe subject feels “full” after the meal. These ratings were notsignificantly different between the high- and low-GI meals in-gested 4 h before bedtime (P " 0.67 and P " 0.72 after the mealand at bedtime, respectively), between the high-GI meals in-gested 4 h and 1 h before bedtime (P " 1.0 and P " 0.28 after themeal and at bedtime, respectively), and between the low-GI mealingested 4 h before bedtime and the high-GI meal ingested 1 hbefore bedtime (P " 0.72 and P " 0.17 after the meal and atbedtime, respectively).

6-SM analysis

The evening collection concentration of 6-SM, a metabolite ofmelatonin, showed no significant differences between thehigh-GI meal ingested 4 h before bedtime (661.8 ! 228.1 ng/h;CV " 34%), the high-GI meal ingested 1 h before bedtime(556.4 ! 209.8 ng/h; CV " 37%; P " 0.3), and the low-GI mealingested 4 h before bedtime (602.4 ! 208.8 ng/h; CV " 34%;P " 0.5). Higher concentrations of 6-SM, as expected, were seenfor the night collection (1783.6 ! 618.8 ng/h; CV " 34%; and1718.5 ! 484.0 ng/h; CV " 28%; P " 0.6, and 1571.6 ! 482.6ng/h, CV% " 30%; P " 0.2, respectively).

Blood glucose

The blood glucose response to both the high- and low-GImeals is shown in Figure 2. Blood glucose rose to a peak at #30and 45 min after meal ingestion, followed by a steady decrease.The repeated-measures ANOVA confirmed that the blood glu-cose profiles over time differed between the 2 meal types (P "0.001 for the group by time interaction). The AUC was signifi-cantly greater for the high-GI (336.2 ! 61.9) than for the low-GI(237.1 ! 69.3) meal (Student’s t test, P " 0.009). The bloodglucose response to ingestion of a high-GI meal 1 h before bed-time was similar to that observed after ingestion of a high-GI

TABLE 1Effect of the glycemic index (GI) and timing of meals on sleep1

Sleep variable

High-GI meal

Low-GI meal, 4 h P2 P31 h 4 h

SOL (min) 14.6 ! 9.94 9.0 ! 6.2 17.5 ! 6.2 0.01 0.009ROL (min) 84.1 ! 39.8 97.0 ! 33.2 82.6 ! 35.7 0.32 0.16SE (%) 92.0 ! 2.2 92.4 ! 2.7 90.7 ! 2.7 0.63 0.06Arousal index (no./h)

REM 15.4 ! 10.4 15.4 ! 7.8 14.5 ! 7.1 0.99 0.70NREM 12.1 ! 5.3 10.7 ! 4.5 10.6 ! 5.8 0.15 0.93Total 12.6 ! 5.1 11.5 ! 4.2 11.4 ! 5.3 0.32 0.95

Sleep stage 1 (%) 5.7 ! 2.0 6.3 ! 2.8 17.5 ! 6.2 0.37 0.08Sleep stage 2 (%) 56.2 ! 5.4 54.5 ! 4.8 82.6 ! 35.7 0.18 0.43Sleep stage 3 (%) 5.3 ! 2.0 4.7 ! 2 90.7 ! 2.7 0.37 0.17Sleep stage 4 (%) 14.7 ! 5.3 14.9 ! 7.2 17.5 ! 6.2 0.82 0.64NREM sleep (%) 81.9 ! 4.3 80.6 ! 4.5 82.6 ! 35.7 0.15 0.77REM sleep (%) 18.0 ! 4.3 19.4 ! 4.5 90.7 ! 2.7 0.14 0.75Total sleep time (min) 478 ! 68.7 472.0 ! 66.4 464.1 ! 70.1 0.78 0.74Total wake time (min) 26.0 ! 9.0 27.6 ! 7.55 29.3 ! 12.7 0.66 0.59

1 n " 12. SOL, sleep onset latency; ROL, rapid eye movement latency; REM, rapid eye movement; SE, sleep efficiency (the ratio of total sleep time inbed); NREM, non–rapid eye movement. Sleep stages and proportion of NREM and REM sleep are presented as a percentage of total sleep time.

2 Comparison of high-GI meal given 4 h and 1 h before bedtime.3 Comparison between high-GI meal ingested 4 h before bedtime with low-GI meal ingested 4 h before bedtime.4 x! ! SD (all such values).

FIGURE 1. Comparison of sleep onset latency (SOL) between the high-glycemic-index (GI) meal ingested 1 h or 4 h before bedtime and the low-GImeal ingested 4 h before bedtime. PC, subject who showed an inappropriatetrend in the SOL for the 3 meals. The mean (!SD) SOLs for the low-GI mealgiven 4 h before usual bedtime and the high-GI meals given 4 h and 1 h beforethe usual bedtime were 17.5 ! 6.2, 9.0 ! 6.2, and 14.6 ! 9.9 min, respec-tively (n " 12). The SOL was significantly shortened after the high-GI mealcompared with the low-GI meal when given 4 h before usual bedtime (P "0.009). When considering the timing of the high-GI meals, the SOL of themeal ingested 4 h before bedtime was significantly shortened compared to themeal ingested 1 h before bedtime, P " 0.01 (repeated-measures ANOVA).

428 AFAGHI ET AL

by guest on March 26, 2011

ww

w.ajcn.org

Dow

nloaded from

Am. J. Clin. Nutr. 2007 Feb.;85(2):426–430.

NUTRITION MYTHS

"""

!!

“A ASSOCIAÇÃO ENTRE DIETA E

ACNE É UM MITO!”!""""""

""

Dose Extra Batata Doce

Controlo

1.  Smolinski KN, Yan AC. Curr Opin Pediatr 2004; 16: 385-391

2.  Thiboutot DM, Strauss JS: Diseases of the sebaceous glands, in Freedberg IM, Eisen AZ, Wolff K, et al, (eds.): Fitzpatrick’s Dermatology in General Medicine, vol 1. (ed 6). New York: McGraw-Hill, 2003 p 683

3.  Cunliffe WJ, Simpson NB: Disorders of sebaceous glands. In: Champion RH, Wilkinson DS, Ebling FJG, et al, (eds): Rook/Wilkinson/Ebling Textbook of Dermatology (ed 6). Oxford: Blackwell Science, Ltd, 1998 p 1951

??

ALUNOS DE MEDICINA RESPONDEM IMEDIATAMENTE QUE ACNE E DIETA É UM MITO

!

ü 28-61 % Crianças (10-12 a)!

ü 79-95 % Adolescentes (16-18 a) !

White GM. J Am Acad Dermatol 1998;39:S34-S37

Goulden V, Stables, GI Cunliffe WJ. J Am Acad Dermatol 1999;41:577-80

EUA!

OKINAWA

!1946: !!!!

!Médicos americanos administraram extensos questionários a médicos locais que mostraram:!

!

Steiner PE. Arch Pathol 1946;42:359-380.

AUSÊNCIA DE ACNE VULGARIS NESTA POPULAÇÃO RURAL

ALASCA

Schaefer O. When the Eskimo comes to town. Nutr Today 1971;6:8-16

ACNE NOS ESQUIMÓS

“Another condition has become prevalent, one obvious even to the layman: acne vulgaris. The condition used to be unknown among Eskimos, but one can see it readily amongst teenagers on the streets of Inuvik, Frobisher Bay, and Cambridge Bay. It is far less Prevalent in the smaller centers” “Old Northmen, such as missionaries, traders, trappers, men of the Royal Canadian Mounted Police and others who have known and watched the Eskimos closely for many years, frequently remark to their physician friends on the change in the complexions of the young people. Many Eskimos themselves blame their pimples on the pop, chocolate, and candies the youngsters consume as if addicted

Otto Schaefer, M.D. – 1971 Esquimó, 1913


DIETA ORIGINAL: !

ü Carne de Animais selvagens (caça e peixe)!

ü Vegetação e bagas (Verão)!

ü Modificações na dieta à medida que foram ocidentalizados:!

Açúcar, Doces, Cereais, Lacticínios"

Controlo


HABITANTES DE KITAVA!

ü  Área de 25 Km2!

ü  Nº Habitantes: 2250!

ü  Estilo de Vida: Horticultura e Pesca!

ü  Durante 7 semanas, em 1990, o Dr. Lindeberg visitou 494 casas!ü  AVALIOU 1200 pessoas com 10 anos ou mais!ü  Incluindo 300 pessoas entre 15-25 anos !

AVALIAÇÃO MÉDICA!

Cordain L, et al. Arch Dermatol 2002;138:1584-90

Consumo muito reduzido de:!

ü  Álcool, café e chá!

ü  Lacticínios!

ü  Cereais!

ü  Açúcar!

ü  Sal refinado!

ü  Óleos vegetais e Margarina!

ü  Alimentos processados!



!NÃO ENCONTROU UM ÚNICO CASO DE ACNE NESTA POPULAÇÃO.!

ACNE É RARA OU INEXISTENTE EM POPULAÇÕES PRIMITIVAS!

"


Controlo

Cordain L, Lindeberg S, Hurtado M, et al: Acne vulgaris: a disease of Western civilization. Arch Dermatol 2002;138:1584-90

ESTUDO DE INTERVENÇÃO DIETÉTICA RANDOMIZADO E CONTROLADO

Dieta LGL: •  25% P •  45% HC de Bx IG •  30% L

Smith RN et al. Am J Clin Nutr. 2007 Jul;86(1):107-15




Costa, A et al. An. Bras. Dermatol. 2010; 85(3): 346-353

ASSOCIAÇÃO POSITIVA COM CARGA GLICÉMICA ELEVADA

E COM LACTICÍNIOS


Controlo

LEITE E ACNE

ESTUDOS EPIDEMIOLÓGICOS!Adebamowo CA, Spiegelman D, Danby FW, et al High school dietary dairy intake and teenage acne. J Am Acad Dermatol; 52(2):207-14, 2005.

Adebamowo CA, Spiegelman D, Berkey CS, et al. Milk consumption and acne in adolescent girls. Dermatol Online J; 12(4):1, 2006. Adebamowo CA, Spiegelman D, Berkey CS, et al. Milk consumption and acne in teenaged boys. J Am Acad Dermatol. 2008 May;58(5):787-93

NUTRITION MYTHS

DDR: 0,8 G/KG/DIA

National Academy of Sciences. Dietary Reference Intakes – The essential guide to nutritional requirements. The National Academy Press, 2006.

ATLETAS: 1,2 – 2 G/KG/DIA

Greenwood M, Kalman D, Antonio J. Nutritional Supplements in Sports and Exercise. Humana Press, Totowa, NJ, 2009

PROTEÍNA tem > Efeito na Saciedade que

GORDURA e HC

Porrini M et al. Physiol Behav 1997;62:563-70

Batterham RL, et al. Cell Metabolism 2006 Sept; 4:223-233.

N Engl J Med 2010;363:2102-13

N Engl J Med 2010;363:2102-13

PROTEÍNA E DCV

NURSES HEALTH STUDY:

ü  80,082 Mulheres

seguidas desde 1976

ü  Proteína (animal e vegetal) associada com menor risco de DAC

Hu FB et al. Am J Clin Nutr 1999;70:221-7

DIETA HIPERPROTEICA MELHORA DISLIPIDEMIA

Substituição de CHO (11% ingestão calórica total diária) por Proteína:

1. Diminuiu LDL (9%), 2. Reduziu Col total /HDL (15%) 3. Diminuiu Triglicéridos (23%) 4. Aumentou HDL (12%)

Wolfe BM et al. Can J Cardiol 1995 11:127G-31G

DIETA HIPERPROTEICA MELHORA DISLIPIDEMIA, GLICEMIA E SENSIBILIDADE À INSULINA EM DT2,

OBESIDADE E/OU SÍNDROME METABÓLICA

1.  O’Dea K. Diabetes 1984; 33, 596-603.

2.  O’Dea K, et al. J. Am. Diet. Assoc. 1989; 89, 1076-1086.

3.  Wolfe BM & Piche LA. Clin. Invest. Med. 1999; 22, 140-148.

4.  Layman DK, et al. J Nutr. 2003 Feb;133(2):411-7.

5.  Farnsworth E, et al. Am J Clin Nutr. 2003 Jul;78(1):31-9.

6.  Aude YW, et al. Arch Intern Med. 2004 Oct 25;164(19):2141-6.

7.  McAuley KA, et al. Diabetologia. 2005 Jan;48(1):8-16.

8.  Luscombe-Marsh ND, et al. Am J Clin Nutr. 2005 Apr;81(4):762-72

9.  Noakes M, Keogh JB, Foster PR, Clifton PM. Am J Clin Nutr. 2005 Jun;81(6):1298-306.

10.  Appel LJ, et al. JAMA. 2005 Nov 16;294(19):2455-64

Hodgson JM, et al. Am J Clin Nutr. 2006 Apr;83(4):780-7

Substituição

moderada de CHO por Proteína poderá reduzir PA em Hipertensos

NUTRITION MYTHS

PROTEÍNA ANIMAL AUMENTA RISCO DE OSTEOPOROSE

!

PRAL DE ALGUNS ALIMENTOS

Peixe 14,6 Carne 12,4 Aves 7,8 Ovo 7,3

Marisco 7,3 Queijo 3,3 Leite 1,3

Cereais 1,1

Oleaginosas -1,1

Fruta -5,2

Tubérculos -5,4

Cogumelos -11,2

Raízes (Cenoura, nabo) -17,1

Tomate -17,5

Hortaliças -23,4

ALCALINIZANTES ACIDIFICANTES

PRAL(mEq/100 kcal)

Leguminosas -0,4

PAEL (mEq/100 kcal)

PRAL (mEq/100 kcal)

NEUTROS

Frassetto L.A. et al. J Nephrol. 2006 Mar-Apr;19 Suppl 9:S33-40.

Dietary protein and calcium interact to influence calcium retention: acontrolled feeding study1–4

Janet R Hunt, LuAnn K Johnson, and ZK Fariba Roughead

ABSTRACTBackground: The effect of meat protein on calcium retention atdifferent calcium intakes is unresolved.Objective: The objective was to test the effect of dietary protein oncalcium retention at low and high intakes of calcium.Design: In a randomized controlled feeding study with a 2 3 2factorial crossover design, healthy postmenopausal women (n !27) consumed either ’675 or ’1510 mg Ca/d, with both low andhigh protein (providing 10% and 20% energy) for 7 wk each, sep-arated by a 3-wk washout period. After 3 wk, the entire diet wasextrinsically labeled with 47Ca, and isotope retention was monitoredby whole-body scintillation counting. Clinical markers of calciumand bone metabolism were measured.Results: High compared with low dietary protein significantly in-creased calcium retention from the low-calcium (29.5% comparedwith 26.0% absorbed) but not the high-calcium diet (18% absorbed).For the low-calcium diet, this effect nearly balanced a protein-related0.5-mmol/d greater urinary calcium excretion. Protein-related calciu-retic effects were independent of dietary calcium. Testing at 1, 2, 3, 5,and 7wk showed no long-term adaptation in urinary acidity or urinarycalcium excretion. High comparedwith low dietary protein decreasedurinary deoxypyridinoline and increased serum insulin-like growthfactor I without affecting parathyroid hormone, osteocalcin, bone-specific alkaline phosphatase, or tartrate-resistant acid phosphatase.Conclusions: In healthy postmenopausal women, a moderate in-crease in dietary protein, from 10% to 20% of energy, slightlyimproved calcium absorption from a low-calcium diet, nearly com-pensating for a slight increase in urinary calcium excretion. Underpractical dietary conditions, increased dietary protein from animalsources was not detrimental to calcium balance or short-term indi-cators of bone health. Am J Clin Nutr 2009;89:1357–65.

INTRODUCTION

Whether dietary protein has an anabolic or catabolic effect onbone remains a controversial issue. Although dietary protein frompurified protein sources generally increases urinary calciumexcretion (1–4), this effect is reduced if not eliminated whenmeat is the protein source (5–7). Furthermore, the calciureticeffect of protein may be counterbalanced by an increase in di-etary calcium absorption (8, 9). Protein is a primary bone con-stituent, and the anabolic effect of dietary protein may bebeneficial for bone mass, especially if dietary calcium is ade-quate (10). However, concerns persist about the effect of proteinon acid-base balance, its associated effect on calcium excretion,

and possible negative consequences for bone health (11, 12),especially if calcium intakes are low.

In this controlled feeding study, the objectives were to de-termine the nature of interaction between dietary protein andcalcium on 1) calcium retention (by using 47Ca radiotracer andwhole-body scintillation counting) and 2) blood and urinarybiomarkers of calcium and bone metabolism. Multiple urinarymeasurements during a 7-wk dietary treatment period also en-abled examination of possible adaptations in urinary acidifica-tion and calciurea.


General protocol and treatment assignment

This controlled feeding trial was conducted as a randomizedcrossover design (2 3 2 factorial); each subject consumed oneamount of dietary calcium but alternately consumed 2 amountsof dietary protein. After statistically blocking to attain an evendistribution of body mass index (BMI; in kg/m2), subjects wererandomly assigned to either high-calcium (HC) or low-calcium(LC) treatment groups. Subjects were further randomly assignedto a sequence for consuming both low-protein (LP) and high-protein (HP) diets. Thus, every subject consumed 2 experimentaldiets for 7 wk each. The 2 diet periods were separated by a 3-wkwashout period when diets were not controlled.

Subjects

Healthy postmenopausal women were recruited through publicadvertising and direct mailings. The women were selected afteran interview and blood analysis if they met the following criteria:

1 From the US Department of Agriculture, Agricultural Research Service,Grand Forks Human Nutrition Research Center, Grand Forks, ND (JRH andZKFR), and the University of North Dakota, Grand Forks, ND (LKJ).

2 Mention of a trademark or proprietary product does not constitute a guar-antee or warranty of the product by the US Department of Agriculture anddoes not imply its approval to the exclusion of other products that may alsobe suitable.

3 Supported by the USDA Agricultural Research Service, the NationalCattlemen’s Beef Association, and the North Dakota Beef Commission.

4 Reprints not available. Address correspondence to JR Hunt, US Depart-ment of Agriculture, ARS, GFHNRC, 2420 2nd Avenue N, STOP 9034,Grand Forks, ND 58202-9034. E-mail: [email protected].

Received November 4, 2008. Accepted for publication January 26, 2009.First published online March 11, 2009; doi: 10.3945/ajcn.2008.27238.


at Lund University Libraries on February 16, 2010

ww

w.ajcn.org

Dow

nloaded from

ü  Proteína aumenta Excreção de Ca!

ü  Proteína aumenta Absorção de Ca!

!!!!!!!


Nutrient Physiology, Metabolism, and Nutrient-Nutrient Interactions

A Diet High in Meat Protein and Potential RenalAcid Load Increases Fractional CalciumAbsorption and Urinary Calcium Excretionwithout Affecting Markers of Bone Resorptionor Formation in Postmenopausal Women1–4

Jay J. Cao,* LuAnn K. Johnson, and Janet R. Hunt

USDA, Agricultural Research Service, Grand Forks Human Nutrition Research Center, Grand Forks, ND 58202

Abstract

Our objective in this study was to determine the effects of a high-protein and high-potential renal acid load (PRAL) diet on

calcium (Ca) absorption and retention and markers of bone metabolism. In a randomized crossover design, 16

postmenopausal women consumed 2 diets: 1 with low protein and low PRAL (LPLP; total protein: 61 g/d; PRAL: 248

mEq/d) and 1with high protein and high PRAL (HPHP; total protein: 118 g/d; PRAL: 33mEq/d) for 7 wk each separated by a

1-wk break. Ca absorption was measured by whole body scintillation counting of radio-labeled 47Ca. Compared with the

LPLP diet, the HPHP diet increased participants’ serum IGF-I concentrations (P , 0.0001), decreased serum intact PTH

concentrations (P, 0.001), and increased fractional 47Ca absorption (mean6 pooled SD: 22.3 vs. 26.56 5.4%; P, 0.05)

and urinary Ca excretion (156 vs. 2036 63 mg/d; P = 0.005). The net difference between the amount of Ca absorbed and

excreted in urine did not differ between 2 diet periods (55 vs. 28 6 51 mg/d). The dietary treatments did not affect other

markers of bonemetabolism. In summary, a diet high in protein and PRAL increases the fractional absorption of dietary Ca,

which partially compensates for increased urinary Ca, in postmenopausal women. The increased IGF-I and decreased PTH

concentrations in serum, with no change in biomarkers of bone resorption or formation, indicate a high-protein diet has no

adverse effects on bone health. J. Nutr. doi: 10.3945/jn.110.129361.

Introduction

Although being essential to bone health, protein intake, espe-cially from animal sources, in high amounts has also beenconsidered a risk factor for osteoporosis or bone fractures (1–4)due to the increase in urinary calcium (Ca) excretion resultingfrom the metabolic acidity of protein metabolism (5–8). How-ever, contrary to the assumption that high dietary protein

impairs bone, many epidemiological observations link a high-protein intake with bone anabolism, including an associationwith increased bone mineral density or decreased fracture risk(9–14), with few reports showing negative associations (15,16).

The results of well-controlled human trials with Ca isotopesshow that a high-protein intake increases intestinal Ca absorp-tion (17–20). Whether such an increase in intestinal Ca absorp-tion can offset hypercalciuria in a diet with a large difference inacid load remains unclear. Furthermore, the alleged detrimentaleffect of high-protein intake on bone may also be dependentupon other dietary factors, such as Ca intake (20).

Our previous finding that a diet with high compared with lowmeat protein did not adversely affect the retention of 47Ca orinduce calciurea (17) was questioned because of a relativelysmall difference (;32 mEq/d) in potential renal acid load(PRAL)5 between the 2 diets (21). PRAL, as a measure of the

1 Supported by the USDA Agricultural Research Service (ARS) program “MineralIntakes for Optimal Bone Development and Health” Current ResearchInformation System no. 5450-51000-039-00D as part of the authors’ officialduties. USDA, ARS, Northern Plains Area, is an equal opportunity/affirmativeaction employer and all agency services are available without discrimination.Mention of a trademark or proprietary product does not constitute a guarantee orwarranty of the product by the USDA and does not imply its approval to theexclusion of other products that may also be suitable. This work was also partiallyfunded by the Beef Check-off through the National Cattlemen’s Beef Association(NCBA). The NCBA did not participant in the design, implementation, interpre-tation, or analysis of the research.2 Author disclosures: J. J. Cao, L. K. Johnson, and J. R. Hunt, no conflicts ofinterest.3 This trial was registered at clinicaltrials.gov as NCT00620763.4 Supplemental Figure 1 and Table 1 are available with the online posting of thispaper at jn.nutrition.org* To whom correspondence should be addressed. E-mail: [email protected].

5 Abbreviations used: CTX, carboxyterminal cross-linking telopeptide; DPD,deoxypyridinoline; HPHP: high protein and high potential renal acid load; LPLP,low protein and low potential renal acid load; NTX, aminoterminal cross-linkingtelopeptide of bone collagen; OC, osteocalcin; OPG, osteoprotegerin; PRAL,potential renal acid load; PTH, parathyroid hormone; sRANKL, soluble receptoractivator nuclear factor-kB ligand; TRAP, tartrate-resistant acid phosphatase.

ã 2011 American Society for Nutrition.Manuscript received July 17, 2010. Initial review completed August 23, 2010. Revision accepted November 30, 2010. 1 of 7doi: 10.3945/jn.110.129361.

The Journal of Nutrition. First published ahead of print January 19, 2011 as doi: 10.3945/jn.110.129361.


by Janet Hunt on February 3, 2011

jn.nutrition.orgD

ownloaded from

NaCl determina ~ 50% da PRAL

Frassetto LA, Morris RC Jr, Sebastian A. Am J Physiol Renal Physiol. 2007 Aug;293(2):F521-5

Jajoo R, et al. J Am Coll Nutr. 2006 Jun;25(3):224-30.

ü  40 H e M > 50 anos

ü  Duração: 60 dias

ü  2 Grupos:

GRUPO 1: Elevada ingestão de Fruta e Vegetais

GRUPO 2: Fruta e Legumes substituidas por Cereais

GRUPO 2:

ü  PTH

ü  Excreção UCa

ü  Excreção urinária de N-Telopéptido

Jajoo R, et al. J Am Coll Nutr. 2006 Jun;25(3):224-30.

NUTRITION MYTHS

Aumento da ingestão de proteína em pacientes com insuficiência renal acelera a progressão

Brenner BM, Meyer TW, Hostetter TH. N Engl J Med. 1982 Sep 9;307(11):652-9

INSUFICIÊNCIA RENAL E PROTEÍNA

ü  Dieta em que Proteína foi 25 % energia não levou a alterações renais adversas ao fim de 6 meses

ü  Sem alterações na albumina urinária (pré/pós dieta)

ü  Taxa de Filtração Glomerular por volume de rim não mudou

ü Conclusão: Rim adapta-se a dieta hiperproteica em pessoas sem patologia renal

Skov AR et al. Int J Obes Metab Disord 1999;23:1170-77.

CONCLUSÃO: !"“Apesar da restrição de Proteína poder ser apropriada no tratamento de patologias renais, não encontrámos evidencia que una ingestão elevada de proteína afecte de forma negativa a função renal de indivíduos saudáveis.”!

""""

NUTRITION MYTHS

LEITE, CA E FRACTURAS

Bischoff-Ferrari HA et al. J Bone Mineral Res 2011; online Oct 2010, DOI 10.1002/jbmr.279!

Pooled analysis for categories of milk intake and hip fracture risk in women from prospective cohort studies (6 studies, 195 102 women, 3574 fractures). !

RR/+1glass"

LEITE E FRACTURAS"

Sem associação em adultos

."""."

""

""""

"

Bischoff-Ferrari HA, Dawson-Hughes B, Baron JA, et al. Am J Clin Nutr. 2007 Dec;86(6):1780-90

248

ü  7 estudos prospectivos, 170 991 mulheres, 2954 fracturas da anca.

ü  5 estudos prospectivos, 68 606 Homens, 214 fracturas da anca.

ü  5 intervenções (5666 Mulheres + 1074 Homens), 814 fracturas não-vertebrais.

ü  4 intervenções c/ resultados separados para fracturas da anca (6504 pessoas, 139 fracturas da anca).

249

Bischoff-Ferrari HA, Dawson-Hughes B, Baron JA, et al. Am J Clin Nutr. 2007 Dec;86(6):1780-900

Total n: 252,841

ü Estudos Prospectivos NÃO MOSTRAM ASSOCIAÇÃO ENTRE INGESTÃO DE CA E FRACTURAS DA ANCA em Homens e Mulheres

ü RCTs observam um LIGEIRO AUMENTO DO RISCO DE FRACTURAS DA ANCA com a SUPLEMENTAÇÃO DE CA

. .

Bolland MJ, et al. BMJ. 2008 Feb 2;336(7638):262-6

NUTRITION MYTHS

Obrigado

www.nutriscience.pt

[email protected]

!

Health & Medicine

Mitos da nutrição