macronutrient

Macronutrients,FoodGroups,andEatingPatterns in theManagement of DiabetesA systematic review of the literature, 2010

MADELYN L. WHEELER, MS, RD, FADA, CD1

STEPHANIE A. DUNBAR, MPH, RD2

LINDSAY M. JAACKS, BS3

WAHIDA KARMALLY, DRPH, RD, CDE, CLS, FNLA4

ELIZABETH J. MAYER-DAVIS, MSPH, PHD, RD5

JUDITH WYLIE-ROSETT, EDD, RD6

WILLIAM S. YANCY JR., MD, MHS7

The effectiveness of medical nutritiontherapy (MNT) in the managementof diabetes has been well established(1). Previous reviews have provided com-prehensive recommendations for MNT inthe management of diabetes (2,3). Thegoals of MNT are to 1) attain andmaintainoptimal blood glucose levels, a lipid andlipoprotein prole that reduces the risk ofmacrovascular disease, and blood pres-sure levels that reduce the risk for vasculardisease; 2) prevent and treat the chroniccomplications of diabetes by modifyingnutrient intake and lifestyle; 3) addressindividual nutrition needs, taking into ac-count personal and cultural preferencesand willingness to change; and 4) main-tain the pleasure of eating by only limitingfood choices when indicated by scienticevidence (4).

The literature on nutrition as it relatesto diabetes management is vast. We un-dertook the specic topic of the role ofmacronutrients, eating patterns, and in-dividual foods in response to continuedcontroversy over independent contribu-tions of specic foods and macronu-trients, independent of weight loss, inthe management of diabetes. The positionof the American Diabetes Association(ADA) on MNT is that each person withdiabetes should receive an individual-ized eating plan (4). ADA has received

numerous criticisms because it doesnot recommend one specic mix ofmacronutrients for everyone with diabe-tes. The previous literature review con-ducted by ADA in 2001 supported theidea that there was not one ideal macro-nutrient distribution for all people withdiabetes. This review focuses on literaturethat has been published since that 2001date (5). This systematic review will beone source of information consideredwhen updating the current ADANutritionPosition Statement (4). Other systematicreviews and key research studies thatmay not be included in this review willalso be considered.

When attempting to tease out the roleof macronutrients from other dietary andlifestyle factors, two critical componentsof MNTdenergy balance and a healthfuleating patterndare not addressed. Whileboth are critical components in the man-agement of diabetes as well as the second-ary prevention of complications andpromotion of health, these topics are be-yond the scope of this particular review.The following questions are addressed inthis review:

1. What aspects ofmacronutrient quantityand quality impact glycemic controland cardiovascular disease (CVD) riskin people with diabetes?

2. How do macronutrients combinein whole foods and eating patternsto affect health in people with dia-betes?

3. Is there an optimal macronutrientratio for glycemic management andCVD risk reduction in people withdiabetes?

4. What ndings and needs should directfuture research?

Systematic reviewproceduredA search of the PubMeddatabase was conducted using the searchterms diabetes and one of a number ofwords (low-fat diet, low-carbohydratediet, Mediterranean diet, Mediterraneaneating pattern, vegetarian, vegan, glyce-mic index (GI), dietary carbohydrates, di-etary protein, total fat, dietary fat,saturated fat, omega-3 fatty acid, dietaryber, meats, legumes, nuts, fruit, vegeta-bles, whole grains, milk) to identify arti-cles published between January 2001 andOctober 2010. Certain terms relevant tonutrition therapy in the management ofdiabetes were not included in the searchterms. These terms include trans fattyacids, monounsaturated fatty acids (MU-FAs), polyunsaturated fatty acids (PU-FAs), sucrose, and sugars. The literaturesearch was limited to articles published inEnglish, and multiple publications fromthe same study were limited to the pri-mary study results article.

Studies included in the systematicreview were conducted in people alreadydiagnosed with diabetes; conducted inoutpatient ambulatory care settings;contained a sample size of 10 or moreparticipants in each study group; and oneof the following study designs: clinicaltrials (controlled and randomized con-trolled [RCT]), prospective observationalstudies, cross-sectional observational stud-ies, or case-control studies. Studies wereexcluded if they were published beforeJanuary 2001 or after October 2010; wereconducted in acute care or inpatient set-tings, in women with gestational diabetes,children under 2 years of age, or individ-uals without diabetes or at risk for diabetes;had less than 10 participants in any study

c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c

From 1Nutritional Computing Concepts, Zionsville, Indiana; 2Medical Affairs, American Diabetes Association,Alexandria, Virginia; the 3School of Public Health, Nutritional Epidemiology, The University of North Car-olina, Chapel Hill, Chapel Hill, North Carolina; the 4Irving Institute for Clinical and Translational Research,Columbia University, New York, New York; the 5Department of Nutrition, The University of North Carolina,Chapel Hill, Chapel Hill, North Carolina; the 6Department of Epidemiology and Population Health, AlbertEinstein College of Medicine, Bronx, New York; and the 7Division of General Internal Medicine, Duke Uni-versity School of Medicine, Durham, North Carolina.

Corresponding author: Stephanie A. Dunbar, [email protected]: 10.2337/dc11-2216This article contains Supplementary Data online at http://care.diabetesjournals.org/lookup/suppl/doi:10

.2337/dc11-2216/-/DC1. 2012 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 prot, and thework is not altered. See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.

434 DIABETES CARE, VOLUME 35, FEBRUARY 2012 care.diabetesjournals.org

R e v i e w s / C o n s e n s u s R e p o r t s / A D A S t a t e m e n t sS Y S T E M A T I C R E V I E W

group;were studies lasting only 1 or 2 days;or were not in one of the study designspreviously listed.

In an effort to expand the researchreview, studies were not excluded basedon retention rates; however, this infor-mation is included in Supplementary Ta-ble 1 and only studies with a retentionrate of .80% are included in the keysummary for each topic area. Weightloss is a confounder in some of the studiesand is noted in Supplementary Table 1.

Meta-analyses published during the in-clusionperiod of this systematic reviewwerereviewed for studies meeting this systematicreviews criteria. This information can befound in Supplementary Table 1.

An initial PubMed database searchfound 152 studies after excluding by titleand abstract review. An additional 18studies were found from bibliography re-view. Of these, 72 studies were excluded fornot meeting inclusion criteria. The mostcommon reasons for exclusion were forresults not applicable to the research ques-tion, not published in amajor journal, smallsample size, review articles, and duplicates.

Challenges in evaluatingmacronutrient studies indiabetesmanagementdIsolatingthe effects of dietary macronutrient com-position on glycemic control and CVD riskis difcult due to confounding, especiallyby weight loss and medication changes.Furthermore, altering the level of onemacronutrient affects the proportion ofother macronutrients, making it difcultto isolate the true exposure. Additionalstudy design issues include the difcultyblinding study participants, investigators,and clinicians. Finally, the lack of stan-dardized denitions for terms such as low-fat (or high-fat) diet, low-carbohydrate(or high-carbohydrate) diet, and low-GI (or high-GI) diet makes comparisonsamong study results difcult. These is-sues were addressed by reporting the en-tire macronutrient composition of dietapproaches and potential confounderswhen this information was available.

Question 1: What aspects ofmacronutrient quantity andquality impact glycemiccontrol and CVD risk inpeople with diabetes?

Carbohydrate amountdThere isno consistent denition of low- (orhigh-) carbohydrate diets throughoutthe literature. Based on the studies in this

systematic review, the following deni-tions are used:

c very-low-carbohydrate diet: 2170 g/dayof carbohydrate

c moderately lowcarbohydrate diet: 30to ,40% of kcal as carbohydrate

c moderate-carbohydrate diet: 4065%of kcal as carbohydrate

c high-carbohydrate diet: .65% of kcalas carbohydrate

These denitions are not all-inclusive(e.g., a 100-g/day carbohydrate diet maybe ,30% kcal), but they represent thetypical denitions used by authors, andall published articles t in one of thesecategories.

Many studies use the term conven-tional or traditional macronutrient dis-tribution as a comparison group. Basedon studies in this review, these terms referto an energy contribution from the diet of5565% carbohydrate, #30% fat, and1020% protein. It should be noted thatpeople with diabetes have been shown toconsume an eating pattern that is about45% of calories from carbohydrate (69).The comparison diets referred to as con-ventional or traditional throughout thisreview are higher in carbohydrate thanthose generally consumed by peoplewith diabetes.

Lower (very low and moderately low)carbohydrate

Glycemic control. Eleven clinical trialsexamined the effects of lowering totalcarbohydrate intake on glycemic controlin individuals with diabetes. The carbo-hydrate content goal of the diet was verylow in 7 studies (1016) and moderatelylow in 4 studies (1720).

All studies included adults with type2 diabetes, duration of follow-up rangedfrom 14 days to 1 year, and sample sizesranged from 10 to 55 participants perstudy group. Designs included two feed-ing trials (one crossover clinical trial andone RCT) (10,18) and nine outpatientnutrition counseling interventions (twosingle-arm clinical trials, one crossoverRCT, and six parallel RCTs) (1117,19,20).All studies analyzed participants accordingto treatment assignment, eight studies wererandomized (1113,1519), and for sixstudies, completion of follow-up was 80%or higher (10,12,13,1719).

A1C decreased with a lower-carbohydrate diet in 6 of 10 studies inwhich it was measured (10,1417,20).

Three RCTs found no statistically signif-icant changes in A1C with a very-low-carbohydrate diet (1113) and onefound no difference with a moderatelylowcarbohydrate diet (19). Other gly-cemic parameters such as fasting bloodglucose (FBG), 24-h blood glucose, 24-hinsulin (10), and fasting insulin levels(18) decreased signicantly, and insulinsensitivity increased signicantly (10)on the lower-carbohydrate diet. Glucose-loweringmedicationswere decreased for in-dividuals following the lower-carbohydratediet (1012,14,17) or were more frequentlydecreased than in the comparison diet (16).

CVD risk. Each of the 11 clinical trialsreported at least one serum lipoprotein.The most notable results were that HDLcholesterol increased signicantly morein one very-low-carbohydrate diet group(16) and twomoderately lowcarbohydratediet groups (18,20) compared with thehigher-carbohydrate control diet. Also, tri-glycerides (TGs) decreased more in onemoderately lowcarbohydrate diet group(20) comparedwith the higher-carbohydratecontrol diet. Otherwise, mean changes inserum lipoproteins resulting from a lower-carbohydrate diet were typically benecialbut occurred without a comparison arm orwere not statistically greater than the com-parison arm.

Summary of lower-carbohydrateresearch since 2002In studies reducing total carbohydrateintake, markers of glycemic control andinsulin sensitivity improved, but studieswere small, of short duration, and in somecases were not randomized or had highdropout rates. Serum lipoproteins typi-cally improved with reduction of totalcarbohydrate intake but, with the excep-tion of HDL cholesterol, were not statis-tically greater than with the comparisondiet. The contribution of weight loss tothe results was not clear in some of thesestudies.

Moderate or high carbohydrate

Glycemic control. Seven clinical trialsand two meta-analyses examined theeffects of moderate- or high-carbohydratediets on glycemic control in patients withtype 2 diabetes (2128) or type 1 diabetes(29). Durations of follow-up ranged from5 to 74 weeks, and sample sizes of partic-ipants completing follow-up ranged from10 to 99. All seven studies were RCTsand analyzed participants according totreatment assignment. Four studies had

care.diabetesjournals.org DIABETES CARE, VOLUME 35, FEBRUARY 2012 435

Wheeler and Associates

completion of follow-up of 80% or higher(21,22,25,26). Only one of the studiesblinded participants to diet treatment (22),and none blinded the outcome assessors.

Four studies found no signicantdifferences in glycemic controlwhencom-paring moderate- or high-carbohydratediets with conventional diets (2124).One RCT found A1C improved withhigher carbohydrate (75% of energy in-take) compared with conventional carbo-hydrate (6070% of energy intake) (21)in secondary analyses that used the obser-vation immediately prior to a participantdropping out or having a diabetes medi-cation change as the nal outcome. Theintent-to-treat analyses, however, showedno signicant differences between groups.In another study, A1C decreased signif-icantly more during 5 weeks on a 40%carbohydrate/30% protein versus a 55%carbohydrate/15% protein diet (25). A1Cwas not signicantly different comparedwith a conventional diet in two studiesexamining a moderate-carbohydrate/higher-protein diet (23,24), two studiesexamining a moderate-carbohydrate/higher-fat (MUFA) diet (22,26), and onestudy comparing a 55% carbohydrate/25% fat with a 55% carbohydrate/30%fat diet (29).

Regarding other glycemic parameters,24-h glucose response was lower with a40% carbohydrate/30% protein dietcompared with a 55% carbohydrate/15% protein diet (25); however, plasmaglucose (2224,26), plasma insulin(23,24,26,29), plasma fructosamine(22), and homeostasis model assessment(HOMA) (24) were not signicantly dif-ferent in other diet comparisons.

Two meta-analyses compared lower-carbohydrate diets with conventional car-bohydrate diets (27,28). Of the 19 studiesreviewed in the article byKodama et al. (27),only three (18,26,30) were published dur-ing the date range for this review, and of the13 studies in the meta-analysis by Kirk et al.(28), only four (10,14,22,25) were pub-lished during the review time period. Theseven studies are included in this review.

CVD risk. Six of the seven interventionsreviewed above reported lipoproteins.Two (25,26) reported signicant reduc-tions in TGs on a 40% carbohydrate diet(vs. 5055% carbohydrate), whereas one(21) observed a signicant reduction inLDL cholesterol on a 75% carbohydratediet (vs. 6070% carbohydrate). Threestudies (2224) found no signicant dif-ferences between comparison diets.

Summary of moderate- and high-carbohydrate research since 2002RCTs presenting information onmoderate-and high-carbohydrate diets are diversein terms of fat and protein content aswell as length of study. Only two RCTsfound signicant differences in A1C be-tween groups, with one study ndingsignicantly lower A1C with the higher-carbohydrate diet only in a subgroupanalysis, and the other study ndingsignicantly lower A1C with the lower-carbohydrate diet. In terms of CVD riskfactors, LDL cholesterol improved morewith a high-carbohydrate diet in onestudy, whereas two studies found TGsimproved more with a lower-carbohydratediet.

Carbohydrate typedStudies inthis systematic review addressing thetype of carbohydrate were those of GI/glycemic load or dietary ber.

GIFor studies in this review that providedthe GI that subjects were able to achieve(end-of-study GI numbers), there wasno agreement as to the denition of lowGI (range 3977) or high GI (range5684). The meta-analyses (almost allstudies used were published before2001) found that the average low GIwas 65, and the average high GI was82, but both had wide ranges. This isfurther complicated by the two bases(glucose or white bread) that have beenused to determine GI values for individ-ual foods.

Glycemic control. Five RCTs (19,3134) compared lower-GI diets withhigher-GI diets in individuals with type2 diabetes. Duration of follow-up rangedfrom 4 to 6 weeks and sample sizes weresmall in four studies (1214 in three ofthe studies, 45 in the other), whereas thefth study lasted for 1 year and included156 subjects in the analysis (19). Comple-tion rates were $80% except for 39% inone study (33). Results were mixed withtwo studies nding A1C was signicantlyreduced with the lower-GI versus higher-GI diets (32,33) and the others ndingno differences in glycemic measures(19,31,34).

Three parallel RCTs (16,35,36)compared a lower-GI diet with dietsother than those designated as higher GI(high-ber diet, traditional diet, very-low-carbohydrate diet) in individuals withtype 2 diabetes. Duration of follow-up was

612 months, retention rates were ,80%in two of the studies (16,35), and samplesizes were moderate (range 40155). Com-pared with a higher-ber diet, the lower-GIdiet decreased A1C and FBG signicantly(35). When the lower-GI weight-loss dietwas compared with a conventional weight-loss diet (36), both groups lowered A1Csignicantly with no signicant differencesbetween groups. The lower-GI diet re-duced A1C signicantly less than thevery-low-carbohydrate diet (16).

A study in youth with type 1 diabetes(37) found that individuals advised tofollow a lower-GIdiet had signicant reduc-tions in A1C compared with individualsadvised to follow a carbohydrate-exchangediet, despite the fact that the mean GI forthe two diet groups was not signicantlydifferent. Two studies indicated that educa-tion can change food selection and may(38) ormaynot (39) affect theGIof the diet.

Three meta-analyses (4042) evalu-ated GI. Anderson et al. (41) includedno studies meeting this reviews criteria;Brand-Miller et al. (40) included one (37);and Thomas and Elliott (42) includedthree (32,33,37). These three studiesfrom the meta-analyses are includedabove (32,33,37).

CVD risk. Mixed results were found forthe ve RCTs comparing low-GI withhigh-GI diets for lipoprotein measures.Two studies found a signicant reductionin total cholesterol (31,32) with one of thetwo reporting a signicant reduction inLDL cholesterol and apolipoprotein(apoB) (32) for the lower-GI diet. Theother three studies found no signicantchanges between groups (19,33,34).

Results were mixed in studies compar-ing lowerGIwith other dietary approaches.Signicantly increased HDL cholesterolwas found with lower GI versus higher-cereal ber but no signicant differences inother measured CVD risk markers (35).Total cholesterol was signicantly loweredwith both a lower-GI diet and a traditionaldiet without signicant differences be-tween groups; however, LDL cholesterolwas signicantly higher with the lower-GIdiet versus the traditional diet (36). A very-low-carbohydrate diet reduced TGs sig-nicantly and increased HDL cholesterolsignicantly compared with a lower-GI,reduced-calorie diet, with no signicantdifferences in total cholesterol and LDLcholesterol (16).

A cross-sectional study (43) of menwith type 2 diabetes described a statisti-cally signicant trend toward decreasing


Medical nutrition therapy in managing diabetes

adiponectin with increasing quintiles ofGI.

Summary of GI research since 2002In general, there is little difference inglycemic control and CVD risk factorsbetween low-GI and high-GI or otherdiets. A slight improvement in glycemiamay result from a lower-GI diet; however,confounding by higher ber (16,33,35)must be accounted for in some of thesestudies. Furthermore, standardized de-nitions of low GI need to be developedand low retention rates on lower-GI dietsmust be addressed (16,33,35).

Dietary berThe Institute of Medicine denes dietaryber as consisting of nondigestible (notdigested in the human small intestine)carbohydrates and lignin that are intrinsicand intact in plants (44). Quanticationof the dietary ber in research studies maybe on the basis of dietary recommenda-tions, grams per 1,000 kcals, the amountadded, or its distribution within the studypopulation. Functional bers are beyondthe scope of this systematic review, andthus functional ber and total ber werenot included in this review.

Glycemic control. Seven RCTs exam-ined the effects of moderate amounts ofber supplements (419 g/day) on glyce-mic control in adults with type 2 diabetes(4551). Durations of follow-up rangedfrom 4 to 12 weeks, and sample sizeswere small (1260 participants in the -ber intervention). All studies were random-ized; all studies analyzed participantsaccording to treatment assignment; com-pletion of follow-up was .80% for twoof the studies (46,51); two of the studiesblinded participants to diet treatment(48,51); and two were double-blinded(45,46). In general, these studies supportthe idea that ber supplements may im-prove postprandial glycemia; however, lit-tle improvement in A1C was observed.

Two dietary counseling RCTs exam-ined the effects of dietary ber as part ofan intervention diet. In the rst study,individuals on the low-GI diet showedsmall but signicant improvements inA1C (after controlling for weight loss, ber,or carbohydrate) and FBG at 6 monthscompared with those on the highcereal -ber diet (35). In the second study, individ-uals on a moderate-carbohydrate (51%),high-ber (27 g/1,000 kcals), lower-GI,moderate-fat diet had signicant decreasesin postprandial glucose variability after 4

weeks comparedwith a lower-carbohydrate(44%), lower-ber (8 g/1,000 kcal), higher-GI, higher-MUFA diet (52).

Markers of improved insulin sensitivity(adiponectin) or inammation (C-reactiveprotein [CRP], tumor necrosis factor-R2[TNF-R2]) were assessed in three cross-sectional reports (43,53,54). Higher cerealor fruit ber intakes were associated withhigher levels of adiponectin (43,53,54) andlower levels of CRP (53,54) or TNF-R2(53,54). Another cross-sectional study(55), using a 3-day weighed diet, foundthat individuals with type 2 diabetes andthe metabolic syndrome had signicantlylower intakes of total dietary ber (speci-cally whole grains and fruits) than thosewith diabetes but without the metabolicsyndrome; however, there were no associ-ations between ber intake and A1C orFBG in either group.

The time period of the meta-analysisby Anderson et al. (41) is before any of thearticles in this systematic review werepublished, therefore the meta-analysis re-sults are not included here.

CVD risk. All RCTs described aboveassessed lipoproteins (35,4552). Fourstudies found no signicant differencebetween intervention and control groupsfor thesemeasures (46,4850). One studyfound that psyllium (vs. cellulose) sup-plements (45) signicantly improvedHDL cholesterol; a second study foundthat a higher-ber, lower-fat, and lower-GI diet versus a lower-ber, higher-fat dietproduced signicantly lower total choles-terol, LDL cholesterol, andHDL cholesterol(52). In addition, one cross-sectional studyfound that a diet higher in soluble berfrom whole grains was associated with alower TG level (55). In contrast, Jenkinset al. (35) found that the lower-GI, highcereal ber diet increased HDL cholesterolsignicantly versus the higher-GI diet, andBle-Castillo (51) found that native bananastarch increased TGs, whereas soy milk de-creased TGs.

The Nurses Health Study (NHS)found lower CVD-specic mortality inwomen with diabetes associated withbran intake after adjustments for lifestyleand dietary factors (56).

Summary of ber researchsince 2002The majority of the reviewed evidenceindicates that adding ber supplements inmoderate amounts (419 g) to a daily dietleads to little improvement in glycemiaand CVD risk markers.

Fat amount

Glycemic managementEight clinical trials examined low-fat eat-ing patterns (2123,29,5760). One trialstudied adults with type 1 diabetes (29),whereas the rest studied adults with type2 diabetes; duration of follow-up rangedfrom 3 days to 74 weeks, and sample sizesof participants completing follow-upranged from 10 to 48 participants perstudy group. All eight trials were outpa-tient nutrition counseling interventions:one single-arm (57), two crossover RCTs(22,29), and ve parallel RCTs. Four trialsreduced total fat intake to ,25% of dailyenergy intake (2123,57), and for therest, fat intake was 2530%. All studiesanalyzed participants according to treat-ment assignment, and completion offollow-up was $80% except in threestudies (29,59,60).

A1C decreased with a low-fat dietin one of seven studies in which it wasmeasured (58). In that study (58), in-tensive dietary advice for a lower-fat,moderate-carbohydrate, higher-berdiet in adults with poor glycemic controlsignicantly decreased A1C comparedwith the control group. Insulin sensitiv-ity by euglycemic-hyperinsulinemicclamp improved in the lower-fat dietcompared with the conventional diet inone study (29).

Two weight-loss RCTs by the samegroup compared meal replacements ver-sus conventional diets (59,60) andfound signicant reductions in FBGover short durations with meal replace-ments. One study carried out for 12months showed no persistent differencein FBG between groups, although signif-icantly more subjects in the meal re-placement group had reductions indiabetic medications (60).

In addition to the information fromthe clinical trials, a cross-sectional study(61) found that higher-fat intake corre-lated with signicantly higher A1C.

CVD riskOf the seven studies that measured CVDrisk factors, only one had signicant nd-ings. In a small single-arm study (57)comparing 3 days on a low-fat, ber-richdiet with study participants baselinehigher-fat diet, both total cholesteroland HDL cholesterol decreased signi-cantly.

The cross-sectional study (61) foundthat higher-fat intake correlated withhigher levels of total cholesterol and



LDL cholesterol as well as coronary arterycalcium.

Summary of low-fat research since2002Lowering total fat intake infrequentlyimproved glycemic control or CVD riskfactors in clinical trials involving individ-uals with diabetes. Lowering fat intake inindividuals with diabetes may improvetotal cholesterol and LDL cholesterol butmay also lower HDL cholesterol.

Fat typedFor this review, the type offat refers to the proportion of total energyfrom a specic fatty acid or fatty acidcategory. Categorization may be on thebasis of the number of, the location of, orthe conguration of double bonds. Satu-rated fatty acids (SFAs) may be assessedbased on distribution within the studypopulation or recommended dietary levels.Omega-3 fatty acids are usually evaluatedas milligrams per day or as a distributionwithin the population rather than on thebasis of percent of energy intake.

Saturated fatty acids

Glycemic control. One RCT in individ-uals with type 2 diabetes compared gly-cemic control outcomes for SFAs versusMUFAs with the total fat remaining equal(62) and did not nd a signicant differ-ence between diets for postprandial glu-cose or insulin response.

CVD risk. A 3-week study (62) reportedno improvement in postprandial lipid tol-erance except for a small but signicantreduction in small VLDL TGs when sub-jects consumed a low SFA diet (8% kcalcompared with 17% as SFA).

Summary of SFAs research since 2002The results from the one study relevant tothis topic indicate that the type/amount offatty acid does not affect postprandialglycemic control so long as the amountof total fat is equivalent. An intriguingidea for future research is that loweringSFA or increasing MUFA may increaseglucagon-like peptide-1 activity, therebyreducing postprandial TG.

Omega-3 fatty acids

Glycemic control. Three blinded RCTsin individuals with type 2 diabetes (6365) found that omega-3 fatty acid supple-ments may increase FBG by a small butsignicant amount. However, a fourthblinded RCT (66) observed a signicant

decrease in A1C with supplementationcompared with controls. In the meta-analysis by Hartweg et al. (67), six studiesmet this systematic review criterion(6365,6870) and are included in thissection.

CVD risk. Three RCTs using omega-3fatty acid supplements (4 g/day eicosa-pentaenoic acid [EPA] or docosahexae-noic acid [DHA]) (64), 2.6 g/day EPA plusDHA (65), or 4 g/day sh oil (68) versuscontrols of corn or olive oil observed anincrease in HDL cholesterol, particularlythe HDL-2 and HDL-2b fractions. One ofthese studies (64) also found a decrease inthe HDL-3 fraction with EPA supplemen-tation. Most studies (64,65,6870) ob-served signicant decreases in TGs withEPA, sh oil, or EPA/DHA combination;however, one (64) showed an increase inTGs with supplementation of DHA alone.A one-armed clinical trial (71) alsofound a signicant increase in HDL cho-lesterol and a signicant decrease in TGswith an EPA/DHA combination.

One study (73) focused on whole-food omega-3 intake in a prospective co-hort and found that baseline marineomega-3 fatty acid intake was inverselyassociated with TG.

Summary of omega-3 fatty acidsresearch since 2002Overall it appears that supplementationwith omega-3 fatty acids does not improveglycemic control but may have benecialeffects on CVD risk biomarkers amongindividuals with type 2 diabetes by reduc-ing TGs (in some but not all studies). Otherbenets (e.g., increasing HDL cholesterolor decreasing LDL cholesterol) are notclearly dened.

ProteindThis section reviews studiesexamining the effects of varying the amountof daily protein intake or the source ofprotein intake and further distinguishesthose studies that included individualswith diabetic kidney disease (DKD).

Amount of protein, individualswithout DKDOne metabolic unit-type crossover RCT(25) and two parallel dietary consultationRCTs (23,74) examined the effects ofhigher protein versus usual protein intake(30% vs. 15% of calories as protein withfat remaining constant at 2530%) on gly-cemic control and CVD risk in individualswith type 2 diabetes. Durations of follow-up ranged from 4 to 16weeks, and sample

sizes were small (range 1229 partici-pants in the higher-protein intervention).All studies analyzed participants accord-ing to treatment assignment, completionof follow-up was .80%, and no studieswere blinded.

A 5-week weight-maintenance study(25) observed a signicant reduction inA1C and 24-h glucose response and sig-nicantly lower fasting TGs on thehigher- versus lower-protein eating pat-terns. A study of 8 weeks of weight lossfollowed by 4 weeks of weight mainte-nance (74) found no signicant differencesbetween higher- and lower-protein groupsfor A1C; however, signicant decreasesin serum total cholesterol and LDL cho-lesterol were observed on the higher-versus lower-protein diets. Anotherstudy (23) and a 1-year follow-up ofthe Parker and colleagues study (24) re-ported no signicant differences be-tween groups in glycemic control or CVDrisk factors.

Amount of protein, individualswith DKDFour parallel RCTs examined the effectsof lower versus usual protein intake onglycemic control, CVD risk factors, andrenal function markers in individualswith types 1 and 2 diabetes and micro-albuminuria (75), macroalbuminuria(76,77), or both (78). Durations of follow-up ranged from 1 to 4 years, samplesizes were small (2347 participants inthe intervention groups), and retentionrates were .80% in two studies (76,77).One study blinded physicians to diettreatment (75). Two studies achievedlower protein intakes of 0.860.89 gprotein/kg/day versus usual protein in-takes (1.021.24) (76,77), whereas in theother two studies, the lower-proteingroup had higher actual protein intakesversus the control groups (75,78). Noneof the studies found signicant differencesbetween groups for glycemia, CVD riskfactors, or renal function (glomerularltration rate [GFR], various measuresof proteinuria). At the levels of proteinachieved, no reduction in serum albuminwas noted.

Two meta-analyses addressed proteinrestriction in people with diabetes andmicro- and macroalbuminuria. The meta-analysis by Pan et al. (79) included fourstudies meeting this review s criteria(7578), and the Cochrane analysis byRobertson et al. (80) included three studies(7577). These four studies (7578) are in-cluded above.



Source of protein, individuals withDKDFour RCTs examined the effects of sourceof protein intake on glycemic control, CVDrisk factors, and renal function in indi-viduals with type 2 diabetes and micro-albuminuria (81) or macroalbuminuria(8284). Durations of follow-up rangedfrom 4 weeks to 4 years, and sample sizeswere small (1420 participants in the des-ignated source interventions). Two studieshad completion rates of.80% (81,83).

The nutrition source focus for twoRCTswas soy. HDL cholesterol increased signif-icantly and urinary albumin-to-creatinineratio decreased signicantly with soy pow-der versus casein powder supplementation(82). The 4-year RCT reported that the re-placement of 35% of animal protein withtextured soy protein resulted in signicantimprovements in FBG and total choles-terol, LDL cholesterol, and TGs, but no sig-nicant changes in kidney function versuscontrol (83). In two crossover RCTs fromthe same author group (81,84), the darkchickenmeat group signicantly improvedtotal cholesterol, TGs, and urinary albu-min excretion rate, and the low-protein/vegetables group signicantly improvedtotal cholesterol and GFR versus the redmeat control group.

Summary of amount and source ofprotein research since 2002For individuals without DKD, higher pro-tein eating patterns (30% of calories) may ormay not improve A1C; however, they ap-pear to improve one or more CVD riskmeasures.

For individuals with DKD and eithermicro- or macroalbuminuria, reducing theamount of protein from normal levels doesnot appear to alter glycemic measures, CVDrisk measures, or the course of GFR. For in-dividuals withDKD andmacroalbuminuria,changing the source of protein to be moresoy based may improve CVD risk measuresbut does not appear to alter proteinuria.

Question 2A: How domacronutrients combine infood groups to affectglycemic response and CVDrisk reduction in peoplewith diabetes?

NutsThe high MUFA content of most tree nutsand peanuts and high PUFA content ofwalnuts and pine nuts lends support tothe investigation of potential effects of

nuts on glycemic control and CVD risk inindividuals with diabetes. Since 2002,three RCTs and two reports from theNHS have been published on this topic(30,8589). All studies analyzed partici-pants according to treatment assignment,and two studies blinded participants totreatment. Completion of follow-up wasgreater than 85% for all studies, and twoof the three studies controlled for weightchange.

Glycemic control. Two RCTs (8587)tested the effects of walnuts against gen-eral advice or advice to consume specicPUFA-rich foods. There were no signi-cant differences among groups for glyce-mic control. One double-blinded studycompared 10% of total calories from fatof almonds or olive/canola oil in the con-text of either a high-fat (37%) or low-fat(25%) diet and also did not nd signi-cant differences in glycemic control(30).

CVD risk. Results relating to measures ofCVD risk were mixed. Addition of wal-nuts led to no signicant differences intotal cholesterol and LDL cholesterol;however, improved endothelial functionwas observed (85). In another study (86),the walnut group achieved signicant re-ductions in LDL cholesterol and increasesin HDL cholesterol and the ratio of HDL-to-total cholesterol relative to the othertreatment groups. However, a third study(30) found that HDL cholesterol was sig-nicantly lower in the group receiving al-monds (vs. olive/canola oil). Theseauthors concluded that total dietary fathad a greater effect on serum lipids thandid fat source (30).

Two cross-sectional studies reportedassociations between nut consumptionand lower-risk CVD risk markers. Nuts,as a part of the Mediterranean-style eatingpattern, had an independent effect onadiponectin levels, which were 12%higher in the highest nut intake quintileversus the lowest (88). Consumption ofat least ve servings per week of nuts orpeanut butter was signicantly associatedwith a more favorable lipid prole (lowertotal cholesterol, LDL cholesterol, andapoB-100). There were no signicant as-sociations for inammatory markers(89).

Summary of nuts research since 2002Nut-enriched diets do not alter glycemiain individuals with diabetes. The evidenceis mixed as to whether they have bene-cial effects on serum lipoproteins.

Whole grainsdThe 2010 DietaryGuidelines for Americans (90) deneswhole grains as foods containing the en-tire grain seed (kernel, bran, germ, andendosperm).

Two single-blinded crossover RCTscompared whole grains to ber (47,48) inindividuals with type 2 diabetes. Durationof follow-up was 512 weeks, samplesizes were small (1520 adults), and re-tention rates were 74% or not reported(47). Whole-wheat our products didnot change glycemic measures over 5weeks, while adding ber (arabinoxylan)to whole-wheat our products resulted insignicantly lower postprandial glucose,insulin, and fructosamine (47). In the sec-ond RCT, A1C and FBG were not alteredsignicantly over 12 weeks with Salba (anovel whole grain) or wheat bran (48).Neither study found signicant differen-ces in CVD risk markers.

Two cross-sectional analyses from theNHS found that higher intake of wholegrains was associated with lower levels ofmarkers of inammation (CRP and TNF-R2) (54) and with higher adiponectinconcentrations (88). One of the RCTsalso found CRP was signicantly lowerin the whole grain versus the wheat brangroups (48).

Summary of whole-grains researchsince 2002Whole-grain consumption does not appearto be associated with improved glycemiccontrol in individuals with diabetes. How-ever, diets high inwhole grainsmay reducesystemic inammation.

Legumes

Soybean-based supplementsTwo crossover and four parallel RCTs(50,60,9195) investigated the effects ofsoy-based supplements on individualswith type 2 diabetes. One of the aboveRCTs reported glycemic andCVD informa-tion in separate publications (91,92). Du-rations of follow-up ranged from 6 weeksto 1 year, retention rates were .80% forfour (9195) of the six studies, sample sizeswere small (1538 in the interventiongroup), and four of the studies were double-blinded (9195). Five of the six studiesfound no signicant difference in glycemicmeasures between groups (92,93) (50,94)(60); however, two studies observed im-provements in LDL cholesterol (91,93) ortotal cholesterol (93) versus control. Adiet-counseling, randomized crossovertrial (52) found that legumes as part of



a moderately highcarbohydrate, high-ber, and lower-GI diet improved post-prandial glucose and CVD risk factorscompared with a higher-MUFA diet.

Isolated soy proteins that includedisoavonesThree crossover RCTs compared soy pro-tein for effects on glycemic and CVD riskmarkers in postmenopausal womenwith type 2 diabetes (9698). Durationof follow-up ranged from 4 to 12 weeks,sample sizes were small (1632), and allstudies were double-blinded. Two studiesfound no signicant differences betweengroups in glycemic control measures orlipoproteins (97,98), and one of thesefound no difference in CRP or HOMA-insulin resistance (IR) (97). However,the third (96) showed signicant reduc-tions in A1C, fasting insulin, HOMA-IR,total cholesterol, and LDL cholesterol inthe soy group compared with the controlgroup.

Summary of legumes researchsince 2002While the soy-derived supplements in thestudies were quite different, most studiesdid not indicate a signicant reduction inglycemic measures or CVD risk factorscompared with controls.

Vegetables and fruitdOnesmall short-term RCT addressed vegeta-ble supplements in individuals with type2 diabetes. At four weeks, garlic powdertablets signicantly improved FBG, fruc-tosamine, and TGs (99). Higher-bervegetables as part of a moderately highcarbohydrate, high-ber, and lower-GIdiet improved postprandial glucose andCVD risk factors compared with ahigher-MUFA diet (52). In women withtype 2 diabetes, vegetables and fruit (as acomponent of the Mediterranean-styleeating pattern score) were not associatedwith adiponectin concentrations (88).

Summary of vegetable and fruitresearch since 2002Eating pattern research has not directlyaddressed the role of vegetables and fruitsin people with diabetes. Of the few studiesfound since 2002, results are mixed.

DairydFive RCTs (two crossover andthree parallel feeding trials) examined theeffects of dairy supplements on glycemiccontrol and CVD risk factors (one RCTreported glycemic and CVD informationin separate publications) (91,92). Three

studies included adultswith type2 diabetesand one included youths with type 1 dia-betes (100). Duration of follow-up rangedfrom 6 to 52 weeks, and sample sizesranged from 11 to 59 participants perstudy group. All studies were randomized,analyzed participants according to treat-ment assignment, completion of follow-upwas.80%, and three were double-blinded(9194).

One RCT (100) found that addingcamels milk to the usual diets of youthnewly diagnosed with type 1 diabetes sig-nicantly reduced A1C and mean dose ofinsulin compared with usual diets alone.

Three RCTs comparing soy to dairy(9194) found no signicant differencesbetween groups in glycemic control. How-ever, two of the studies (91,93) did ndLDL cholesterol to be signicantly higherfor the milk protein isolate (91) and casein(93) groups (vs. the soy groups).

An ancillary report of a weight-lossstudy (101) found that there was no re-lationship between dairy calcium and gly-cemic control or CVD risk markers.

Summary of dairy research since 2002None of the components of dairy appearto have an effect on glycemic control orCVD risk reduction.

Meats, poultry, and shdIntwo crossover RCTs from the same researchgroup (81,84), a usual diet with darkchicken meat replacing red meat was com-pared with a low-protein/vegetable diet.Therewere no signicant differences amonggroups for FBG, LDL cholesterol, and HDLcholesterol. Total cholesterol was signi-cantly lower after the chicken and the veg-etable protein diet versus the red meat diet,and TGs were signicantly lower after thechickendiet versus the redmeat diet and thevegetable protein diet. In women with type2 diabetes in theNHS (102), a high intake ofredmeat was signicantly associated with fa-tal coronary heart disease, coronary revascu-larization, and total coronary heart disease. Acase-control study (103) indicated that a highintake of sh protein was associated with adecreased risk ofmicro/macroalbuminuria inyouth with type 1 diabetes.

Summary of meat research since 2002Currently, there is limited evidence toprovide conclusive statements relating tothe intake of meat, poultry, and sh.

Overall summary of Question 2AResearch involving diabetes and foodgroups is sparse and does not indicate

an advantage for specic foods in improv-ing glycemic control. There is a possibilitythat certain CVD risk factors could beimproved with the consumption of nutsor whey.

Question 2B: How domacronutrients combine ineating patterns to affectglycemic response and CVDrisk factors in people withdiabetes?dEating patterns includedbut are not limited todlower carbohydrate,lower fat, lower GI (see the respective sec-tions in Question 1) as well as Mediterra-nean and vegetarian.

Mediterranean-style eating patternAMediterranean-style eating pattern, basedon the reviewed studies, generally includesmore vegetables,whole grains, fruit, legumes,nuts, sh, and MUFA/PUFA; less red meatand SFAs; and some alcohol (wine) com-pared with a traditional diet.

Summary of reviewed studiesFive RCTs (52,104107) compared aMediterranean or modied Mediterranean-style eating pattern to other eating pat-terns over a period of 4 weeks to 4 years.

A4-year study (104) compared aweight-reduction/maintenance Mediterranean-style eating pattern to a lower-fat eating pat-tern. Weight loss was similar, and therewere no signicant differences in glycemiccontrol between groups. Adiponectin in-creased similarly with both eating patterns.

De Natale et al. (52) found that a mod-erately highcarbohydrate, high-ber, andlower-GI Mediterranean-style eating pat-tern signicantly improved postprandialglucose compared with a higher-MUFAMediterranean-style eating pattern.

Three RCTs comparing Greek tradi-tional or fast foods found no signicantdifferences between groups for glycemiccontrol and CVD risk factors (105107).

A cross-sectional study (88) and acase-control study (108) examined theMediterranean-style eating pattern toaddress how adherence was related toselected biomarkers. There were no sig-nicant differences between adherencetertiles for A1C (88,108), total cholesterol(88,108), or LDL cholesterol (88). HDLcholesterol was signicantly higher andTG was signicantly lower in the highesttertile of adherence to the Mediterranean-style eating pattern (88); the highest ter-tile of adherence also was associatedwith a 56% reduction in risk of peripheralarterial disease (108). The NHS (88)



found that adherence to the Mediterranean-style eating pattern was associated withhigher plasma adiponectin concentrationsin women with diabetes, and this was at-tributed mainly to the intake of alcohol,nuts, and whole grains.

An RCT (109) compared 4 oz. of redwine daily to no alcohol. Fasting insulinand HOMA decreased in both groups,with the wine group having a signi-cantly greater decrease. Both groups sig-nicantly reduced total cholesterol andLDL cholesterol with no change in TG.HDL cholesterol was signicantly in-creased in the wine group only, whereasmarkers of inammation (TNF, CRP, andothers) were signicantly increased in thecontrol group.

Summary of Mediterranean-style eatingpattern research since 2002There appears to be no advantage in usingthe Mediterranean-style eating patterncompared with other eating patterns forglycemic control. There are mixed resultsfor CVD risk factors with some studiesindicating that the Mediterranean-styleeating pattern might improve HDL cho-lesterol and TG. Individual componentsof the Mediterranean-style eating pattern(wine, high MUFA/olive oil) do not ap-pear to have independent effects on gly-cemic control, but may be responsible forimprovement in HDL cholesterol.

Vegetarian eatingpatterndOne RCT (21,110) com-paring a low-fat vegan eating patternand a conventional eating pattern foundthat weight and A1C decreased in bothgroups, with no signicant difference be-tween groups in the primary analyses. Inan ancillary analysis that removed partic-ipants who did not complete follow-upor who had medications changed dur-ing follow-up, there was a signicantlygreater decrease in A1C and LDL choles-terol in the vegan group. In a 4-weekcrossover RCT in individuals with earlyDKD, a lacto-vegetarian eating patterndid not show signicant differences inFBG, HDL cholesterol, or LDL cholesterol;however, total cholesterol signicantly de-creased compared with the usual eatingpattern, and GFR signicantly decreasedcompared with both the usual and chickendiets (81,84).

Summary of vegetarian eatingpattern research since 2002Research is limited regarding vegetarianeating patterns. Because of methodological

problems, more research is needed beforeconclusive remarks can be made about theassociations between a vegetarian eatingpattern and glycemic control and CVD riskfactors.

Overall summary of Question 2BStudies examining how eating patternsare related to glycemic control and CVDrisk markers have varied with respect tomacronutrient distribution used to char-acterize low-fat, Mediterranean, low-GI,vegetarian, and lower-carbohydrate eat-ing patterns. While some research sug-gests that these eating patterns improveglycemic and cardiovascular outcomes,variability in research methods and de-nitions have complicated interpretationof ndings. Issues that could affect con-clusions include retention rates, dietaryintervention and assessment methodol-ogy, and data analysis approaches.

Question 3: Is there anoptimal macronutrient ratiofor glycemic managementand cardiovascular riskreduction in people withdiabetes?dVariability in studymethodology, including measurement ofdietary intake, retention rates, and con-founding by weight loss, limits compar-isons as to how macronutrient distributionindependent of weight loss affects out-comes of interest. Although in manyinstances there were not statistically sig-nicant differences between dietary ap-proaches, improvements were often seenfrom baseline to follow-up in both in-tervention groups supporting the ideathat several different macronutrient dis-tributions may lead to improvements inglycemic and/or CVD risk factors (Sup-plementary Table 1).

Question 4: What shouldguide the future directionsof research?dTheevidencepresentedin this review suggests that many differentapproaches to MNT and eating patternsare effective for the target outcomes ofimproved glycemic control and reducedCVD risk among individuals with dia-betes. However, several gaps in the litera-ture remain that warrant mentioninghere.

Most of the studies in the presentreview examined the relationship ofmacronutrients and foods to biochemi-cal markers of glycemic control and CVDrisk. While research has long explored themechanisms underlying the relationship

between nutrition and glycemia, studieshave only just begun examining how nu-trition relates to the endocrine functions offat tissue and other cardiovascular param-eters. For example, future studies shouldaddress:

c The role of adiponectin, which may beresponsive to changes in eating pat-terns and has been associated withbetter diabetes-relatedhealth outcomes inepidemiological studies

c The relationship between ber/whole-grain intake and improved insulin sensi-tivity and markers of inammation (e.g.,CRP and TNF)

c The role of omega-3 fatty acids in relationto adipose tissue inammation, throm-bosis, and lipidmetabolism in the contextof observations that higher intakes areassociated with reduced CVD mortality,particularly sudden cardiac death

c The impact of very-low-carbohydrateand moderately lowcarbohydrate eatingpatterns on long-term complicationssuch as nephropathy

c The impact of postprandial excursionsand hyperglycemia on inammatoryresponse and subsequent CVD risk

In addition to these biochemicalmechanisms underlying nutrition-relatedCVD risk, the interplay between specicnutrients and dietary macronutrient com-position has yet to be thoroughly evalu-ated. The use of technology such ascontinuous glucose monitors to evaluatethe impact of macronutrients in isolation,in the presence of specic nutrients, inthe context of a mixed meal, and in overalleating patterns must be elucidated in orderto fully understand how diet impacts gly-cemic control.

Moving forward, it is essential to con-sider that individuals benet differentlyfrom various nutritional approaches. Stud-ies on gene-diet interactions and the impactof various macronutrient compositionsacross the continuum of dysglycemia/insulin resistance warrant additional inves-tigation. Related to this tailored approachto MNT, it should be noted that individualadherence to nutrition recommendations ishighly variabledand generally suboptimal.Research is needed to develop strategiesthat enhance adherence and to determineif certain nutritional approaches promotegreater adherence than others.

Continued support is needed forlarge, multicenter trials with clinical eventend points. Diabetes care involves mon-itoring risk factors for bothmacrovascular



and microvascular complications andtherefore the sample size needed to detectmultiple biologically and clinically rele-vant effect sizes requires special consid-eration. Furthermore, the duration offollow-up needs to be adequate relativeto the outcomes of interest, and strategiesshould be used to improve retention.When dropout and/or missing data areextensive, special analytic strategies may benecessary to reduce the potential for selec-tion bias. Study design and statistical anal-yses should consider time-varying factors,such as changes inweight andmedications,which may independently impact studyoutcomes, especially in small-scale efcacytrials. Finally, due to the large volumeand variety of research regarding diet anddiabetes-related health outcomes, rigoroussystematic reviews and meta-analyses needto be conducted so that researchers, clini-cians, patients, and funding agencies areaware of the most recent research and thedirection in which it is heading.

AcknowledgmentsdW.K. has reported be-ing a member of the Research Committee forthe American Pistachio Growers. No otherpotential conicts of interest relevant to thisarticle were reported.

M.L.W. researched data, contributed todiscussion, and wrote, reviewed, and editedthe manuscript. S.A.D. contributed to discus-sion and wrote, reviewed, and edited themanuscript. L.M.J. reviewed and edited themanuscript. W.K. researched data, contrib-uted to discussion, and reviewed and editedthe manuscript. E.J.M.-D. reviewed and editedthe manuscript. J.W.-R. researched data,contributed to discussion, and reviewed andedited the manuscript. W.S.Y. researched data,contributed to discussion, and wrote, reviewed,and edited the manuscript.

The authors thankM. Sue Kirkman, MD, forher input into the manuscript and the formerUniversity of North Carolina students forconducting the initial literature search: EmilyFord, MPH, RD; Natalie Peterson, MPH, RD;Cassandra Rico, MPH, RD; Carolyn Wait,MPH, RD; and John Yoon, BS.

References1. Franz MJ, Boucher JL, Green-Pastors J,

Powers MA. Evidence-based nutritionpractice guidelines for diabetes andscope and standards of practice. J AmDiet Assoc 2008;108(Suppl. 1):S52S58

2. American Dietetic Association. Diabetestype 1 and 2 evidence-based nutritionpractice guidelines for adults [articleonline], 2008. Chicago, IL. Availablefrom http://www.adaevidencelibrary.

com/topic.cfm?=3252. Accessed 10 No-vember 2011

3. Franz MJ, Powers MA, Leontos C, et al.The evidence for medical nutritiontherapy for type 1 and type 2 diabetes inadults. J Am Diet Assoc 2010;110:18521889

4. Bantle JP, Wylie-Rosett J, Albright AL,et al.; American Diabetes Association.Nutrition recommendations and inter-ventions for diabetes: a position statementof the American Diabetes Association.Diabetes Care 2008;31(Suppl. 1):S61S78

5. Franz MJ, Bantle JP, Beebe CA, et al.Evidence-based nutrition principlesand recommendations for the treatmentand prevention of diabetes and relatedcomplications. Diabetes Care 2002;25:148198

6. Delahanty LM, Nathan DM, Lachin JM,et al.; Diabetes Control andComplicationsTrial/Epidemiology of Diabetes. Associa-tion of diet with glycated hemoglobinduring intensive treatment of type 1 di-abetes in the Diabetes Control and Com-plications Trial. Am J Clin Nutr 2009;89:518524

7. Eeley EA, Stratton IM, Hadden DR,Turner RC, Holman RR; UK ProspectiveDiabetes Study Group. UKPDS 18: esti-mated dietary intake in type 2 diabeticpatients randomly allocated to diet, sul-phonylurea or insulin therapy. DiabetMed 1996;13:656662

8. Vitolins MZ, Anderson AM, DelahantyL, et al.; Look AHEAD Research Group.Action for Health in Diabetes (LookAHEAD) trial: baseline evaluation ofselected nutrients and food group in-take. J Am Diet Assoc 2009;109:13671375

9. Oza-Frank R, Cheng YJ, Narayan KM,Gregg EW. Trends in nutrient intakeamong adults with diabetes in the UnitedStates: 1988-2004. J Am Diet Assoc2009;109:11731178

10. Boden G, Sargrad K, Homko C, MozzoliM, Stein TP. Effect of a low-carbohydratediet on appetite, blood glucose levels,and insulin resistance in obese patientswith type 2 diabetes. Ann Intern Med2005;142:403411

11. Daly ME, Paisey R, Paisey R, et al. Short-term effects of severe dietary carbohydrate-restriction advice in type 2 diabetesdarandomized controlled trial. Diabet Med2006;23:1520

12. Davis NJ, Tomuta N, Schechter C, et al.Comparative study of the effects of a1-year dietary intervention of a low-carbohydrate diet versus a low-fat dieton weight and glycemic control in type2 diabetes. Diabetes Care 2009;32:11471152

13. Dyson PA, Beatty S, Matthews DR.A low-carbohydrate diet is more effectivein reducing body weight than healthy

eating in both diabetic and non-diabeticsubjects. Diabet Med 2007;24:14301435

14. YancyWS Jr, FoyM, Chalecki AM, VernonMC, Westman EC. A low-carbohydrate,ketogenic diet to treat type 2 diabetes. NutrMetab (Lond) 2005;2:34

15. Stern L, Iqbal N, Seshadri P, et al. Theeffects of low-carbohydrate versus con-ventional weight loss diets in severelyobese adults: one-year follow-up of arandomized trial. Ann Intern Med 2004;140:778785

16. Westman EC, Yancy WS Jr, MavropoulosJC, Marquart M, McDufe JR. The ef-fect of a low-carbohydrate, ketogenicdiet versus a low-glycemic index diet onglycemic control in type 2 diabetes mel-litus. Nutr Metab (Lond) 2008;5:36

17. Haimoto H, Sasakabe T, Wakai K,Umegaki H. Effects of a low-carbohydratediet on glycemic control in outpatientswith severe type 2 diabetes. Nutr Metab(Lond) 2009;6:21

18. Miyashita Y, Koide N, Ohtsuka M, et al.Benecial effect of low carbohydrate inlow calorie diets on visceral fat reductionin type 2 diabetic patients with obesity.Diabetes Res Clin Pract 2004;65:235241

19. Wolever TM, Gibbs AL, Mehling C, et al.The Canadian Trial of Carbohydrates inDiabetes (CCD), a 1-y controlled trialof low-glycemic-index dietary carbohy-drate in type 2 diabetes: no effect onglycated hemoglobin but reduction inC-reactive protein. Am J Clin Nutr 2008;87:114125

20. Jnsson T, Granfeldt Y, Ahrn B, et al.Benecial effects of a Paleolithic diet oncardiovascular risk factors in type 2 di-abetes: a randomized cross-over pilotstudy. Cardiovasc Diabetol 2009;8:35

21. Barnard ND, Cohen J, Jenkins DJ, et al. Alow-fat vegan diet and a conventionaldiabetes diet in the treatment of type 2diabetes: a randomized, controlled, 74-wk clinical trial. Am J Clin Nutr 2009;89:1588S1596S

22. Gerhard GT, Ahmann A, Meeuws K,McMurry MP, Duell PB, Connor WE.Effects of a low-fat diet compared withthose of a high-monounsaturated fat dieton body weight, plasma lipids and lipo-proteins, and glycemic control in type 2diabetes. Am J Clin Nutr 2004;80:668673

23. Wycherley TP, Noakes M, Clifton PM,Cleanthous X, Keogh JB, BrinkworthGD. A high-protein diet with resistanceexercise training improves weight lossand body composition in overweightand obese patients with type 2 diabetes.Diabetes Care 2010;33:969976

24. Brinkworth GD, Noakes M, Parker B,Foster P, Clifton PM. Long-term effectsof advice to consume a high-protein,low-fat diet, rather than a conventional



weight-loss diet, in obese adults withtype 2 diabetes: one-year follow-up of arandomised trial. Diabetologia 2004;47:16771686

25. Gannon MC, Nuttall FQ, Saeed A,Jordan K, Hoover H. An increase in di-etary protein improves the blood glucoseresponse in persons with type 2 diabetes.Am J Clin Nutr 2003;78:734741

26. Rodrguez-Villar C, Prez-Heras A,Mercad I, Casals E, Ros E. Comparisonof a high-carbohydrate and a high-monounsaturated fat, olive oil-rich dieton the susceptibility of LDL to oxidativemodication in subjects with type 2 di-abetes mellitus. Diabet Med 2004;21:142149

27. Kodama S, Saito K, Tanaka S, et al. In-uence of fat and carbohydrate pro-portions on the metabolic prole inpatients with type 2 diabetes: a meta-analysis. Diabetes Care 2009;32:959965

28. Kirk JK, Graves DE, Craven TE, LipkinEW, Austin M, Margolis KL. Restricted-carbohydrate diets in patients with type2 diabetes: a meta-analysis. J Am DietAssoc 2008;108:91100

29. Rosenfalck AM, Almdal T, Viggers L,Madsbad S, Hilsted J. A low-fat diet im-proves peripheral insulin sensitivity inpatients with type 1 diabetes. DiabetMed 2006;23:384392

30. Lovejoy JC, Most MM, Lefevre M,Greenway FL, Rood JC. Effect of dietsenriched in almonds on insulin actionand serum lipids in adults with normalglucose tolerance or type 2 diabetes. AmJ Clin Nutr 2002;76:10001006

31. Kabir M, Oppert JM, Vidal H, et al. Four-week low-glycemic index breakfast witha modest amount of soluble bers in type2 diabetic men. Metabolism 2002;51:819826

32. Rizkalla SW, Taghrid L, Laromiguiere M,et al. Improved plasma glucose con-trol, whole-body glucose utilization, andlipid prole on a low-glycemic index dietin type 2 diabetic men: a randomizedcontrolled trial. Diabetes Care 2004;27:18661872

33. Jimenez-Cruz A, Bacardi-Gascon M,Turnbull WH, Rosales-Garay P, Severino-Lugo I. A exible, low-glycemic indexMexican-style diet in overweight andobese subjects with type 2 diabetes im-proves metabolic parameters during a6-week treatment period. Diabetes Care2003;26:19671970

34. Heilbronn LK,NoakesM,Clifton PM. Theeffect of high- and low-glycemic indexenergy restricted diets on plasma lipid andglucose proles in type 2 diabetic subjectswith varying glycemic control. J Am CollNutr 2002;21:120127

35. Jenkins DJ, Kendall CW, McKeown-Eyssen G, et al. Effect of a low-glycemicindex or a high-cereal ber diet on type 2

diabetes: a randomized trial. JAMA2008;300:27422753

36. Ma Y, Olendzki BC, Merriam PA, et al.A randomized clinical trial comparinglow-glycemic index versus ADA di-etary education among individualswith type 2 diabetes. Nutrition 2008;24:4556

37. Gilbertson HR, Brand-Miller JC,Thorburn AW, Evans S, Chondros P,Werther GA. The effect of exible lowglycemic index dietary advice versusmeasured carbohydrate exchange dietson glycemic control in children with type1 diabetes. Diabetes Care 2001;24:11371143

38. Burani J, Longo PJ. Low-glycemic indexcarbohydrates: an effective behavioralchange for glycemic control and weightmanagement in patients with type 1 and2 diabetes. Diabetes Educ 2006;32:7888

39. Cheong SH, McCargar LJ, Paty BW,Tudor-Locke C, Bell RC. The First StepFirst Bite Program: guidance to increasephysical activity and daily intake of low-glycemic index foods. J Am Diet Assoc2009;109:14111416

40. Brand-Miller J, Hayne S, Petocz P,Colagiuri S. Low-glycemic index dietsin the management of diabetes: a meta-analysis of randomized controlled trials.Diabetes Care 2003;26:22612267

41. Anderson JW, Randles KM, Kendall CW,Jenkins DJ. Carbohydrate and ber rec-ommendations for individuals with di-abetes: a quantitative assessment andmeta-analysis of the evidence. J Am CollNutr 2004;23:517

42. Thomas D, Elliott EJ. Low glycaemicindex, or low glycaemic load, diets fordiabetes mellitus. Cochrane DatabaseSyst Rev 2009;1):CD006296

43. Qi L, Rimm E, Liu S, Rifai N, Hu FB.Dietary glycemic index, glycemic load,cereal ber, and plasma adiponectinconcentration in diabetic men. DiabetesCare 2005;28:10221028

44. Institute of Medicine. Dietary referenceintakes for energy, carbohydrate, ber, fat,fatty acids, cholesterol, protein, and aminoacids (macronutrients). Washington, DC,The National Academies Presses, 2005,p. 340341

45. Ziai SA, Larijani B, Akhoondzadeh S,et al. Psyllium decreased serum glucoseand glycosylated hemoglobin signi-cantly in diabetic outpatients. J Ethno-pharmacol 2005;102:202207

46. Magnoni D, Rouws CH, Lansink M, vanLaere KM, Campos AC. Long-term useof a diabetes-specic oral nutritionalsupplement results in a low-postprandialglucose response in diabetes patients.Diabetes Res Clin Pract 2008;80:7582

47. Lu ZX, Walker KZ, Muir JG, ODea K.Arabinoxylan bre improves metabolic

control in people with type II diabetes.Eur J Clin Nutr 2004;58:621628

48. Vuksan V, Whitham D, Sievenpiper JL,et al. Supplementation of conventionaltherapywith the novel grain Salba (Salviahispanica L.) improves major and emerg-ing cardiovascular risk factors in type 2diabetes: results of a randomized con-trolled trial. Diabetes Care 2007;30:28042810

49. Jenkins DJ, Kendall CW, Augustin LS,et al. Effect of wheat bran on glycemiccontrol and risk factors for cardiovascu-lar disease in type 2 diabetes. DiabetesCare 2002;25:15221528

50. Cho SH, Kim TH, Lee NH, Son HS, ChoIJ, Ha TY. Effects of Cassia tora bersupplement on serum lipids in Koreandiabetic patients. J Med Food 2005;8:311318

51. Ble-Castillo JL, Aparicio-Trpala MA,Francisco-Luria MU, et al. Effects of na-tive banana starch supplementation onbody weight and insulin sensitivity inobese type 2 diabetics. Int J Environ ResPublic Health 2010;7:19531962

52. De Natale C, Annuzzi G, Bozzetto L,et al. Effects of a plant-based high-carbohydrate/high-ber diet versus high-monounsaturated fat/low-carbohydratediet on postprandial lipids in type 2 di-abetic patients. Diabetes Care 2009;32:21682173

53. Qi L, Meigs JB, Liu S, Manson JE,Mantzoros C, Hu FB. Dietary bers andglycemic load, obesity, and plasma adi-ponectin levels in women with type 2diabetes. Diabetes Care 2006;29:15011505

54. Qi L, van Dam RM, Liu S, Franz M,Mantzoros C, Hu FB. Whole-grain, bran,and cereal ber intakes and markers ofsystemic inammation in diabetic women.Diabetes Care 2006;29:207211

55. Steemburgo T, DallAlba V, Almeida JC,Zelmanovitz T, Gross JL, de Azevedo MJ.Intake of soluble bers has a protectiverole for the presence of metabolic syn-drome in patients with type 2 diabetes.Eur J Clin Nutr 2009;63:127133

56. He M, van Dam RM, Rimm E, Hu FB, QiL. Whole-grain, cereal ber, bran, andgerm intake and the risks of all-cause andcardiovascular disease-specic mortalityamong women with type 2 diabetesmellitus. Circulation 2010;121:21622168

57. Mostad IL, Qvigstad E, Bjerve KS, GrillVE. Effects of a 3-day low-fat diet on met-abolic control, insulin sensitivity, lipids andadipocyte hormones in Norwegian sub-jects with hypertriacylglycerolaemia andtype 2 diabetes. Scand J Clin Lab Invest2004;64:565574

58. Coppell KJ, Kataoka M, Williams SM,Chisholm AW, Vorgers SM, Mann JI. Nu-tritional intervention inpatientswith type 2diabetes who are hyperglycaemic despite



optimised drug treatmentdLifestyle Overand Above Drugs in Diabetes (LOADD)study: randomised controlled trial. BMJ2010;341:c3337

59. Yip I, Go VL, DeShields S, et al. Liquidmeal replacements and glycemic controlin obese type 2 diabetes patients. ObesRes 2001;9(Suppl. 4):341S347S

60. Li Z, Hong K, Saltsman P, et al. Long-term efcacy of soy-based meal replace-ments vs an individualized diet plan inobese type II DM patients: relative effectson weight loss, metabolic parameters,and C-reactive protein. Eur J Clin Nutr2005;59:411418

61. Snell-Bergeon JK, Chartier-Logan C,Maahs DM, et al. Adults with type 1 di-abetes eat a high-fat atherogenic diet thatis associated with coronary artery cal-cium. Diabetologia 2009;52:801809

62. Rivellese AA, Giacco R, Annuzzi G, et al.Effects of monounsaturated vs. saturatedfat on postprandial lipemia and adiposetissue lipases in type 2 diabetes. ClinNutr 2008;27:133141

63. Mostad IL, Bjerve KS, Bjorgaas MR,Lydersen S, Grill V. Effects of n-3 fattyacids in subjects with type 2 diabetes:reduction of insulin sensitivity and time-dependent alteration from carbohydrateto fat oxidation. Am J Clin Nutr 2006;84:540550

64. Woodman RJ,Mori TA, Burke V, PuddeyIB,Watts GF, Beilin LJ. Effects of puriedeicosapentaenoic and docosahexaenoicacids on glycemic control, blood pres-sure, and serum lipids in type 2 diabeticpatients with treated hypertension. Am JClin Nutr 2002;76:10071015

65. Pedersen H, Petersen M, Major-PedersenA, et al. Inuence of sh oil supplemen-tation on in vivo and in vitro oxidationresistance of low-density lipoprotein intype 2 diabetes. Eur J Clin Nutr 2003;57:713720

66. Pooya Sh, Jalali MD, Jazayery AD,Saedisomeolia A, Eshraghian MR, ToorangF. The efcacy of omega-3 fatty acid sup-plementation on plasma homocysteine andmalondialdehyde levels of type 2 diabeticpatients. Nutr Metab Cardiovasc Dis 2010;20:326331

67. Hartweg J, Farmer AJ, Holman RR, NeilA. Potential impact of omega-3 treat-ment on cardiovascular disease in type2 diabetes. Curr Opin Lipidol 2009;20:3038

68. Petersen M, Pedersen H, Major-PedersenA, Jensen T, Marckmann P. Effect of shoil versus corn oil supplementation onLDL and HDL subclasses in type 2 di-abetic patients. Diabetes Care 2002;25:17041708

69. Kabir M, Skurnik G, Naour N, et al.Treatment for 2 mo with n 3 poly-unsaturated fatty acids reduces adiposityand some atherogenic factors but doesnot improve insulin sensitivity in women

with type 2 diabetes: a randomizedcontrolled study. Am J Clin Nutr 2007;86:16701679

70. Shidfar F, Keshavarz A, Hosseyni S, AmeriA, Yarahmadi S. Effects of omega-3 fattyacid supplements on serum lipids, apoli-poproteins and malondialdehyde in type2 diabetes patients. East Mediterr HealthJ 2008;14:305313

71. Kesavulu MM, Kameswararao B, ApparaoCh, Kumar EG, Harinarayan CV. Effect ofomega-3 fatty acids on lipid peroxidationand antioxidant enzyme status in type 2diabetic patients. Diabetes Metab 2002;28:2026

72. This reference was withdrawn.73. Belalcazar LM, Reboussin DM, Haffner

SM, et al.; LookAHEAD(Action forHealthin Diabetes) Obesity, Inammation, andThrombosis Research Group. Marineomega-3 fatty acid intake: associationswith cardiometabolic risk and responseto weight loss intervention in the LookAHEAD (Action for Health in Diabetes)study. Diabetes Care 2010;33:197199

74. Parker B, Noakes M, Luscombe N,Clifton P. Effect of a high-protein, high-monounsaturated fat weight loss diet onglycemic control and lipid levels in type2 diabetes. Diabetes Care 2002;25:425430

75. Pijls LT, de Vries H, van Eijk JT, DonkerAJ. Protein restriction, glomerular ltra-tion rate and albuminuria in patientswith type 2 diabetes mellitus: a random-ized trial. Eur J Clin Nutr 2002;56:12001207

76. Meloni C, Tatangelo P, Cipriani S, et al.Adequate protein dietary restriction indiabetic and nondiabetic patients withchronic renal failure. J Ren Nutr 2004;14:208213

77. Hansen HP, Tauber-Lassen E, Jensen BR,Parving HH. Effect of dietary proteinrestriction on prognosis in patients withdiabetic nephropathy. Kidney Int 2002;62:220228

78. Dussol B, Iovanna C, Raccah D, et al. Arandomized trial of low-protein diet intype 1 and in type 2 diabetes mellituspatients with incipient and overt ne-phropathy. J Ren Nutr 2005;15:398406

79. Pan Y, Guo LL, Jin HM. Low-proteindiet for diabetic nephropathy: a meta-analysis of randomized controlled trials.Am J Clin Nutr 2008;88:660666

80. Robertson L, Waugh N, Robertson A.Protein restriction for diabetic renal dis-ease. Cochrane Database Syst Rev 2007;4):CD002181

81. Gross JL, Zelmanovitz T, Moulin CC,et al. Effect of a chicken-based diet onrenal function and lipid prole in pa-tients with type 2 diabetes: a randomizedcrossover trial. Diabetes Care 2002;25:645651

82. Teixeira SR, Tappenden KA, Carson L,et al. Isolated soy protein consumptionreduces urinary albumin excretion andimproves the serum lipid prole in menwith type 2 diabetes mellitus and ne-phropathy. J Nutr 2004;134:18741880

83. Azadbakht L, Atabak S, Esmaillzadeh A.Soy protein intake, cardiorenal indices,and C-reactive protein in type 2 diabeteswith nephropathy: a longitudinal ran-domized clinical trial. Diabetes Care2008;31:648654

84. de Mello VD, Zelmanovitz T, PerassoloMS, Azevedo MJ, Gross JL. Withdrawalof red meat from the usual diet reducesalbuminuria and improves serum fattyacid prole in type 2 diabetes patientswith macroalbuminuria. Am J Clin Nutr2006;83:10321038

85. Ma Y, Njike VY, Millet J, et al. Effects ofwalnut consumption on endothelial func-tion in type 2 diabetic subjects: a random-ized controlled crossover trial. DiabetesCare 2010;33:227232

86. Tapsell LC, Gillen LJ, Patch CS, et al.Including walnuts in a low-fat/modied-fat diet improves HDL cholesterol-to-total cholesterol ratios in patients withtype 2 diabetes. Diabetes Care 2004;27:27772783

87. Gillen LJ, Tapsell LC, Patch CS, Owen A,Batterham M. Structured dietary adviceincorporating walnuts achieves optimalfat and energy balance in patients withtype 2 diabetes mellitus. J Am Diet Assoc2005;105:10871096

88. Mantzoros CS, Williams CJ, Manson JE,Meigs JB, Hu FB. Adherence to the Med-iterranean dietary pattern is positivelyassociated with plasma adiponectin con-centrations in diabetic women. Am J ClinNutr 2006;84:328335

89. Li TY, Brennan AM, Wedick NM,Mantzoros C, Rifai N, Hu FB. Regularconsumption of nuts is associated with alower risk of cardiovascular disease inwomen with type 2 diabetes. J Nutr2009;139:13331338

90. U.S. Department of Health and HumanServices. Dietary Guidelines for Ameri-cans, 2010 (Internet). Available fromhttp://health.gov/dietaryguidelines/2010.asp. Accessed 30 June 2011

91. Pipe EA, Gobert CP, Capes SE, DarlingtonGA, Lampe JW, Duncan AM. Soy proteinreduces serum LDL cholesterol and theLDL cholesterol:HDL cholesterol and apo-lipoprotein B:apolipoprotein A-I ratios inadults with type 2 diabetes. J Nutr 2009;139:17001706

92. Gobert CP, Pipe EA, Capes SE,Darlington GA, Lampe JW, Duncan AM.Soya protein does not affect glycaemiccontrol in adults with type 2 diabetes. BrJ Nutr 2010;103:412421

93. Hermansen K, Sndergaard M, Hie L,Carstensen M, Brock B. Benecial effectsof a soy-based dietary supplement on lipid



levels and cardiovascular risk markers intype 2 diabetic subjects. Diabetes Care2001;24:228233

94. Kim JI, Kim JC, KangMJ, Lee MS, Kim JJ,Cha IJ. Effects of pinitol isolated fromsoybeans on glycaemic control and car-diovascular risk factors in Korean patientswith type II diabetes mellitus: a random-ized controlled study. Eur J Clin Nutr2005;59:456458

95. Fujita H, Yamagami T, Ohshima K.Long-term ingestion of a fermentedsoybean-derived Touchi-extract withalpha-glucosidase inhibitory activity issafe and effective in humans with bor-derline and mild type-2 diabetes. J Nutr2001;131:21052108

96. Jayagopal V, Albertazzi P, Kilpatrick ES,et al. Benecial effects of soy phytoes-trogen intake in postmenopausal womenwith type 2 diabetes. Diabetes Care 2002;25:17091714

97. Gonzlez S, Jayagopal V, Kilpatrick ES,Chapman T, Atkin SL. Effects of iso-avone dietary supplementation on car-diovascular risk factors in type 2 diabetes.Diabetes Care 2007;30:18711873

98. Howes JB, Tran D, Brillante D, HowesLG. Effects of dietary supplementationwith isoavones from red clover onambulatory blood pressure and endo-thelial function in postmenopausaltype 2 diabetes. Diabetes Obes Metab2003;5:325332

99. Sobenin IA, Nedosugova LV, FilatovaLV, Balabolkin MI, Gorchakova TV,

Orekhov AN. Metabolic effects of time-released garlic powder tablets in type 2diabetes mellitus: the results of double-blinded placebo-controlled study. ActaDiabetol 2008;45:16

100. Mohamad RH, Zekry ZK, Al-MehdarHA, et al. Camel milk as an adjuvanttherapy for the treatment of type 1 di-abetes: verication of a traditionalethnomedical practice. J Med Food 2009;12:461465

101. Shahar DR, Abel R, Elhayany A, Vardi H,Fraser D. Does dairy calcium intake en-hance weight loss among overweightdiabetic patients? Diabetes Care 2007;30:485489

102. Qi L, van Dam RM, Rexrode K, Hu FB.Heme iron from diet as a risk factor forcoronary heart disease in women withtype 2 diabetes. Diabetes Care 2007;30:101106

103. Mllsten AV, Dahlquist GG, Stattin EL,Rudberg S. Higher intakes of sh proteinare related to a lower risk of micro-albuminuria in young Swedish type 1diabetic patients. Diabetes Care 2001;24:805810

104. Esposito K, Maiorino MI, Ciotola M,et al. Effects of a Mediterranean-stylediet on the need for antihyperglycemicdrug therapy in patients with newly di-agnosed type 2 diabetes: a randomizedtrial. Ann Intern Med 2009;151:306314

105. Karantonis HC, Fragopoulou E,Antonopoulou S, Rementzis J, Phenekos

C, Demopoulos CA. Effect of fast-foodMediterranean-type diet on type 2 dia-betics and healthy human subjects plate-let aggregation. Diabetes Res Clin Pract2006;72:3341

106. Antonopoulou S, Fragopoulou E,Karantonis HC, et al. Effect of traditionalGreek Mediterranean meals on plateletaggregation in normal subjects and inpatients with type 2 diabetes mellitus.J Med Food 2006;9:356362

107. Aronis P, Antonopoulou S, KarantonisHC, Phenekos C, Tsoukatos DC. Effectof fast-food Mediterranean-type diet onhuman plasma oxidation. J Med Food2007;10:511520

108. Ciccarone E, Di Castelnuovo A, SalcuniM, et al.; Gendiabe Investigators. A high-score Mediterranean dietary pattern is as-sociated with a reduced risk of peripheralarterial disease in Italian patients with type2 diabetes. J Thromb Haemost 2003;1:17441752

109. Marfella R, Cacciapuoti F, Siniscalchi M,et al. Effect of moderate red wine intakeon cardiac prognosis after recent acutemyocardial infarction of subjects withtype 2 diabetes mellitus. Diabet Med2006;23:974981

110. Turner-McGrievy GM, Barnard ND,Cohen J, Jenkins DJ, Gloede L, GreenAA. Changes in nutrient intake and dietaryquality among participants with type 2diabetes following a low-fat vegan diet or aconventional diabetes diet for 22 weeks.J Am Diet Assoc 2008;108:16361645



macronutrient

Documents