Effect of Lowering the GlycemicLoad With Canola Oil on GlycemicControl and Cardiovascular RiskFactors: A Randomized ControlledTrialDiabetes Care 2014;37:1806–1814 | DOI: 10.2337/dc13-2990

OBJECTIVE

Despite their independent cardiovascular disease (CVD) advantages, effects ofa-linolenic acid (ALA), monounsaturated fatty acid (MUFA), and low-glycemic-load (GL) diets have not been assessed in combination. We therefore determinedthe combined effect of ALA, MUFA, and low GL on glycemic control and CVD riskfactors in type 2 diabetes.

RESEARCH DESIGN AND METHODS

The study was a parallel design, randomized trial wherein each 3-month treat-ment was conducted in a Canadian academic center between March 2011 andSeptember 2012 and involved 141 participants with type 2 diabetes (HbA1c 6.5%–8.5% [48–69mmol/mol]) treated with oral antihyperglycemic agents. Participantswere provided with dietary advice on either a low-GL diet with ALA and MUFAgiven as a canola oil–enriched bread supplement (31 g canola oil per 2,000 kcal)(test) or a whole-grain diet with a whole-wheat bread supplement (control). Theprimary outcome was HbA1c change. Secondary outcomes included calculatedFramingham CVD risk score and reactive hyperemia index (RHI) ratio.

RESULTS

Seventy-nine percent of the test group and 90% of the control group completedthe trial. The test diet reduction in HbA1c units of20.47% (25.15mmol/mol) (95%CI20.54% to20.40% [25.92 to24.38 mmol/mol]) was greater than that for thecontrol diet (20.31% [23.44 mmol/mol] [95% CI 20.38% to 20.25% (24.17 to22.71 mmol/mol)], P = 0.002), with the greatest benefit observed in those withhigher systolic blood pressure (SBP). Greater reductions were seen in CVD riskscore for the test diet, whereas the RHI ratio increased for the control diet.

CONCLUSIONS

A canola oil–enriched low-GL diet improved glycemic control in type 2 diabetes,particularly in participants with raised SBP, whereas whole grains improved vas-cular reactivity.

1Department of Nutritional Sciences, Faculty ofMedicine, University of Toronto, Toronto, ON,Canada2Department of Medicine, Faculty of Medicine,University of Toronto, Toronto, ON, Canada3Clinical Nutrition and Risk Factor ModificationCenter, St. Michael’s Hospital, Toronto, ON, Can-ada4Division of Endocrinology and Metabolism, St.Michael’s Hospital, Toronto, ON, Canada5Li Ka Shing Knowledge Institute of St. Michael’sHospital, Toronto, ON, Canada6College of Pharmacy andNutrition, University ofSaskatchewan, Saskatoon, SK, Canada7Medical School, Faculty of Medicine, Universityof Ottawa, Ottawa, ON, Canada8School of Medicine, Faculty of Health Sciences,Queen’s University, Kingston, ON, Canada9Department of Clinical Epidemiology and Bio-statistics, Faculty of Health Sciences, McMasterUniversity, Hamilton, ON, Canada10Department of Pathology and Molecular Med-icine, Faculty of Health Sciences, McMaster Uni-versity, Hamilton, ON, Canada

Corresponding author: David J.A. Jenkins,nutritionproject@smh.ca.

Received 20 December 2013 and accepted 15April 2014.

Clinical trial reg. no. NCT01348568, clinicaltrials.gov.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-2990/-/DC1.

A slide set summarizing this article is availableonline.

© 2014 by the American Diabetes Association.See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.

David J.A. Jenkins,1,2,3,4,5 CyrilW.C. Kendall,1,3,6

Vladimir Vuksan,1,3,5 Dorothea Faulkner,1,3

Livia S.A. Augustin,3 Sandra Mitchell,1,3

Christopher Ireland,1,3 Korbua Srichaikul,3,7

Arash Mirrahimi,3,8 Laura Chiavaroli,1,3

Sonia Blanco Mejia,1,3 Stephanie Nishi,1,3

Sandhya Sahye-Pudaruth,1,3

Darshna Patel,1,3

Balachandran Bashyam,1,3

Edward Vidgen,3 Russell J. de Souza,3,9

John L. Sievenpiper,3,5,10 Judy Coveney,3

Robert G. Josse,1,2,3,4,5 and

Lawrence A. Leiter 1,2,3,4,5

1806 Diabetes Care Volume 37, July 2014

DIABETES

CARESYMPOSIUM

New pharmacological treatments for di-abetes are required to be tested for car-diovascular safety before licensing (1)due to concerns over possible increasedcardiovascular disease (CVD) risk insome studies (2). Dietary strategies,although less effective in improvingglycemic control, may have the advan-tage of actually reducing CVD risk (3,4).Low-glycemic-load (GL) diets have

been associated in cohort studieswith a reduction in both diabetes inci-dence and CVD events (3–5), especiallyin overweight individuals (3), and havebeen recommended by many diabetesassociations (6–8). Monounsaturatedfatty acids (MUFAs) and short-chain-length n-3 fatty acids (a-linolenic acid[ALA]) reduced CVD risk in randomizedcontrolled trials (9,10). Furthermore,high ALA and MUFA intake may alsolower the GL of the diet. An increasedproportion of vegetable oil calories inthe meal would be expected to reducepostprandial glycemia both by decreas-ing the carbohydrate content of themeal and by delaying gastric emptying,whereas the increase in vegetable oilover the longer term would predict a re-duction in serum lipids. This combineddietary approach may therefore benefitboth glycemia and CVD risk in diabetes.Despite these possible advantages, theeffects of ALA and MUFA as part of alow-GL diet have not been tested intype 2 diabetes.To determine the possible advantages

of this combination, we tested the effectof a commonly used oil, canola oil, con-taining both ALA (9.1%) and MUFA (63%)when used as part of a low–glycemic in-dex (GI) diet. This dietary interventionwas compared with a high-whole-grain–cereal diet. Such whole-grain diets haveinvariably been associatedwith a reducedrisk of diabetes (11,12) and CVD in cohortstudies (12–14), despite generally havingno effect on conventional CVD risk fac-tors (15).

RESEARCH DESIGN AND METHODS

ParticipantsParticipants were recruited from news-paper, public transportation, and hospi-tal clinic advertisements. One hundredand forty-one participants were eligibleand randomized (Fig. 1). Recruitmenttook place from 28 March 2011 to 17September 2012, with the last studyvisit on 4 December 2012. Eligible

participants had at least a 6-month his-tory of type 2 diabetes based on clinicalcriteria, were taking a stable dose of oralantihyperglycemic agents for at leastthe previous 2 months, and had HbA1cvalues between 6.5% (48 mmol/mol)and 8.5% (69 mmol/mol) both at theinitial screening and at the visit 1 weekbefore randomization (Fig. 1). No partic-ipants had clinically significant cardio-vascular, renal (creatinine.150 mmol/L),or liver (alanine aminotransferase levelmore than three times the upper limit ofnormal) disease or a history of cancer.None were smokers, and alcohol intakewas two or fewer drinks a day for menand one or fewer drinks a day forwomen. Participation rate and reasonsfor exclusion are given in Fig. 1.

ProtocolThe study followed a randomized, paralleldesign with two treatment arms of 3months duration as follows: 1) a low-GLdietwith a canola oil–enriched bread pro-vided as a supplement (test) or 2) a highwheat-fiber diet emphasizing whole-wheat foods (control). After stratificationby sex and HbA1c .7.1% or #7.1% (54mmol/mol) but without a predeterminedblock size, participants were randomizedin a blinded fashion by a statistician whowas geographically separate from thestudy center. The dietitians and partici-pants could not be blinded, but equalemphasis was placed on the potentialimportance of both diets for health.The analytical technicians, statistician,and study investigators were blinded totreatment up to and including the anal-ysis of the primary outcome.

Participants attended the Risk FactorModification Centre of St. Michael’sHospital, a teaching hospital of the Uni-versity of Toronto, for screening andweeks 21, 0, 2, 4, 8, 10, and 12 of thestudy. They were weighed at each visit;waist circumference was measuredwhile standing at the level of the umbi-licus, and fasting blood samples weretaken at all visits except week 2. Seatedblood pressure was measured in tripli-cate with an automatic sphygmoma-nometer (Omron HEM 907 XL; OmronHealthcare Inc., Burlington, ON, Canada)and the mean taken. Seven-day food re-cords covering the week before eachvisit were discussed with the dietitian.No specific exercise advice was given,but participants were asked to keep

exercise constant. Baseline exercise rou-tine was recorded and any subsequentchange noted. The study conformed tothe same general principles as otherstudies of this duration run from thecenter (16).

The study was approved by the re-search ethics board of St. Michael’sHospital and the University of Toronto,and written consent was obtained fromall participants. The study was regis-tered with ClinicalTrials.gov (identifier:NCT01348568).

Dietary InterventionsThe test diet included 4.5 slices of ca-nola oil–enriched whole-wheat bread(500 kcal/day) provided as a supple-ment. The supplement delivered 31 gcanola oil or 14% of total dietary caloriesof a 2,000-kcal diet (Supplementary Ta-ble 1). The control diet included 7.5slices of whole-wheat bread without ca-nola oil per day (500 kcal) (Supplemen-tary Table 1). Dietary advice on the testdiet emphasized low-GI foods, includinglegumes, barley, pasta, parboiled rice,and temperate-climate fruit, as outlinedin previous studies (17). For the con-trol diet, participants were instructedto avoid white-flour products and re-place them with whole-wheat breakfastcereals, study breads, brown rice, andso forth.

Dietary AssessmentParticipants provided 7-day food re-cords covering the previous 7 days be-fore clinic visits. These records werediscussed with the dietitians for clarifi-cation for future formal dietary analysesand to indicate where further dietaryadvice was required. The different na-ture of the diets precluded blinding;however, the advantages of both dietswere emphasized with reference totheir benefits as recorded in the litera-ture (11–14). Adherence to the diet wasassessed from the 7-day food records;106 participants provided complete di-etary records for the 3-month study.Participants ranked their level of satietyon a scale of 24 (starved/feeling weak)to +4 (painfully full) and palatability ofstudy breads and diets at each visit on ascale of 1–10 (1 = strongly dislike, 10 =like very much).

Biochemical and Dietary AnalysesHbA1c, blood glucose, and serum lipidswere measured in the hospital routine

care.diabetesjournals.org Jenkins and Associates 1807

analytical laboratory by techniques aspreviously described (17). The reactivehyperemia index, as a marker of flow-mediated vasodilatation, was measuredwith the EndoPAT system (Itamar Med-ical Ltd., Franklin, MA), which assessesthe capillary blood refill to finger tipsafter a 5-min occlusion of the forearmwith a cuff inflated to 50mmHg above theparticipant’s resting systolic blood pres-sure and expressed as a ratio of the bloodflow in theopposite arm (18). Diet recordswere analyzed using a computer program(ESHA Food Processor SQL version 10.9;ESHA, Salem, OR) based on U.S. Depart-ment of Agriculture data (19) and interna-tional GI tables (20) using the bread scale(where bread = 100; for the glucose scale,bread scale values were multiplied by0.71) (21) (Supplementary Table 1).

Statistical Analyses

Results are expressed as mean 6 SEMor 95% CI. Both the absolute and the

relative CVD risk score were calculatedusing the Framingham risk equation fortotal 10-year cardiovascular events (22),in which only systolic blood pressureand total and HDL cholesterol (HDL-C)changed during the study. All patientswho met the inclusion criteria were in-cluded in the analysis (n = 141). Week0 HbA1c was taken as baseline, andweeks 8, 10, and 12 were selected asend of study to allow for stabilizationof HbA1c as the main outcome. Treat-ment differences in physical and bio-chemical measures were assessed fromall available data. The analysis of treat-ment effect within a repeated-measuresstudy design used the mixed (random-effects) linear model, with change frombaseline over time as the response vari-able and diet (low-GL canola vs. wheatbran) and time (weeks 8, 10, and 12) asthe main effects. Neither baseline norother covariates were used in the pri-mary analysis, which was performed

with SAS 9.2 software (23). Within-treatment changes for all variables wereestimated by the least squares meanstechnique within the mixed model.

Changes in medication use were as-sessed either by two-tailed Fisher exacttest in the case of 23 2 tables or by theMantel-Haenszel test for larger contin-gency tables. For the values used in Fig.2 and associated Supplementary Table2, multiple imputation using five setsof randomly imputed values for missingdata was generated by PROC MI and ana-lyzed by PROC MIXED, and the five sets ofresults were pooled by PROC MIANALYZEin SAS 9.2 (23).

We also assessed the interactions be-tween the effect of diet on HbA1c andthe baselinemeasures of components ofthe metabolic syndrome (waist circum-ference, systolic and diastolic bloodpressure, HDL-C, fasting triglyceridelevel, and blood glucose level) togetherwith the additional components of the

Figure 1—Flow of participants through the study.

1808 Canola Oil and Glycemic Control in Diabetes Diabetes Care Volume 37, July 2014

Framingham risk score (age, sex, totalcholesterol:HDL-C) (SupplementaryTable 2). HbA1c response datawere strat-ified according to whether the partici-pants’ baseline measures were above(or equal to) or below the cut point formetabolic syndrome components (24) orthe median for CVD risk factor compo-nents of the Framingham risk score (22).The HbA1c data for upper and lower sys-tolic blood pressure cut points are alsopresented in graphic form (Fig. 2). Thetreatment differences between thesesubgroups and the significance of theinteraction with baseline measures(Supplementary Table 2) were calculatedfor both raw and multiply imputeddata. To determine whether any baselinemeasures affected the HbA1c response by.10% and might thus be considereda modifier of the effect size, we under-took a bivariate regression analysis ofHbA1c change (repeated measures) in-volving baseline measures of predictorssuggested by the metabolic syndrome di-agnostic criteria (6) and Framingham CVDrisk factors (3) by using one predictor ata time.Initially, we planned to recruit 120

participants. However, because of alarger-than-expected dropout at thestart and to capture smaller effect sizesseen in our more-recent studies, partic-ipant recruitment numbers were in-creased to 140 (16,25). On the basis ofdata from a 12-week study in type 2 di-abetes (16) from an ANCOVA model, wewould require 116 completers todetect a treatment difference in HbA1c

change of 0.15% with an SD of 0.48%[assuming a = 0.05, 1 2 b = 0.8, usingr = 0.8 to account for the high degreeof correlation between successivemeasures (26)].

RESULTS

Fifty-five of 70 participants (79%) com-pleted the test diet (i.e., provided atleast one blood sample in the finalmonth), compared with 64 of 71 (90%)on the control diet (Fig. 1). Of the 119participants with data in the last month(completers), 3 on the test diet and 7 onthe control diet weremissing one or twoof the three final values. The attritionrates were not significantly different be-tween treatments (Fig. 1). No baselinedifferences were seen (Table 1) exceptfor a higher baseline dietary GI in thetest group compared with the controlgroup (3 GI units [95% CI 1.1–4.9], P =0.003) (Supplementary Table 3). Thetest bread was rated more palatablethan the control bread, as was the over-all test diet compared with the controldiet (Supplementary Table 3).

By design, the test diet resulted insignificantly greater increases in MUFAand ALA intake and corresponding lowercarbohydrate intake, and hence GL, rel-ative to the control diet (SupplementaryTable 3). The relative GI and GL reduc-tions for the test diet comparedwith thecontrol diet were 219 GI units (95% CI220 to 217, P , 0.0001) and 252 GLunits (95% CI259 to245, P, 0.0001),respectively, and compliance with thetest bread was 89% (95% CI 86%–93%)

versus the control bread 77% (95% CI74%–80%) (P , 0.0001).

Glycemic Control and Body WeightOral antihyperglycemic medication dos-ages increased in one and were reducedin five participants on the test diet. Theydecreased in four participants on thecontrol diet, with no significant treat-ment differences.

The mean HbA1c change was20.47%(25.15mmol/mol) absolute HbA1c units(95% CI 20.54% to 20.40% [25.92 to24.38 mmol/mol], P , 0.001) for thetest diet and 20.31% (23.44 mmol/mol)absolute HbA1c units (95% CI20.38% to20.25% [24.17 to 22.71 mmol/mol],P , 0.001) for the control diet. The rel-ative HbA1c reduction for the test dietwas 20.16% (21.71 mmol/mol) (95%CI 20.25% to 20.06% [22.77 to 20.65mmol/mol], P = 0.002) (Table 2) and re-mained statistically significant after ad-justment for body weight change (P =0.010). The body weight reductionswere similar at 22.1 kg and 21.6 kgfor both the test and the control diets,respectively (Table 2). There was no sig-nificant treatment difference inwaist cir-cumference, although, as with bodyweight, both treatments were associ-ated with a reduction (waist circumfer-ence 21.8 vs. 22.4 cm for test andcontrol diets, respectively) (Table 2).

Serum LipidsLipid-lowering medications were de-creased in one participant on the testdiet and three on the control diet, withno significant treatment difference in

Figure 2—Changes from baseline in HbA1c (percent absolute HbA1c units) during canola low-GL (test) and high wheat-fiber (control) diets. Dietresults in participants with lower (A) or higher baseline systolic blood pressure (SBP) (B) than themetabolic syndrome cut points. HbA1c was reducedmore for the test diet than for the control diet in those with higher baseline SBP (P = 0.003).

care.diabetesjournals.org Jenkins and Associates 1809

medication use (P = 0.620). The test pro-duced significant falls within treatmentin total cholesterol, LDL cholesterol(LDL-C), triglycerides, and the ratios oftotal cholesterol:HDL-C and LDL-C:HDL-C(Table 2). Relative to the control diet,the test diet resulted in significant reduc-tions in total cholesterol (20.34 mmol/L[95% CI 20.46 to 20.23], P , 0.0001),LDL-C (20.25 mmol/L [95% CI 20.34

to 20.15], P , 0.0001), triglycerides(20.14 mmol/L [95% CI 20.26 to20.03], P = 0.018), and HDL-C (20.03mmol/L [95% CI 20.06 to 0.00], P ,0.041), albeit with still significant reduc-tions in the ratios of total cholesterol:HDL-C (20.21 [95% CI20.32 to20.11],P , 0.0001) and LDL-C:HDL-C (20.16[95% CI 20.24 to 20.07], P = 0.001)(Table 2).

Blood Pressure, Heart Rate, andReactive Hyperemia IndexNo significant treatment differenceswere seen in blood pressure or heartrate (Table 2). There was a nonsignifi-cant reduction in vascular reactivity forthe test diet but a nearly significant risefor the control diet, resulting in a rela-tive increase in the reactive hyperemiaindex for the control diet (20.24 [95% CI20.42 to 20.06], P = 0.009) (Table 2).

CVD RiskThe Framingham risk score for CVD wasreduced for both treatments but signif-icantly more for the test diet (20.6 [95%CI 21.1 to 20.2], P = 0.008) (Table 2).

Effect of Baseline Metabolic SyndromeComponents and Framingham RiskScore Components on HbA1c

ResponseTo determine whether participants athigher risk benefited more or less fromthe intervention, we assessed the HbA1ctreatment effect for those with higherversus lower baseline measures forcomponents of the metabolic syndromeand Framingham risk score. In general,the effect size and degree of significancewas greatest in those whose baselinemeasures were elevated (Supplemen-tary Table 2). However, by multiple im-putation for missing data, only for thosewith higher systolic blood pressure($130 mmHg) was the treatment differ-ence significantly different from thosewith lower systolic blood pressure(,130 mmHg). In participants with sys-tolic blood pressure .130 mmHg, thetest diet HbA1c reduction was substan-tial at 20.62% (26.79 mmol/mol) (95%CI 20.77% to 20.47% [28.40 to 25.19mmol/mol], P , 0.001) (Fig. 2 and Sup-plementary Table 2). The treatment dif-ference in HbA1c in those with systolicblood pressure .130 mmHg (20.41%[24.45 mmol/mol] [95% CI 20.62% to20.19% (26.80 to 22.09 mmol/mol)],P = 0.001) was more than five times thetreatment difference (P = 0.003) seen inthose with systolic blood pressure,130mmHg (20.07% [20.81 mmol/mol][95% CI 20.20% to 0.06% (22.22 to0.60 mmol/mol)], P = 0.253) (Supple-mentary Table 2).

To identify possible confounders, bi-variate regression of HbA1c change onbaseline components of the metabolicsyndrome and Framingham risk score in-dicated that only age was a significant

Table 1—Baseline (week 0) characteristics of study participants

Participants

Characteristic* Control diet (n = 71) Test diet (n = 70)

Age (years) 59 6 10 59 6 10

SexFemale 32 (45) 32 (46)Male 39 (55) 38 (54)

Race/ethnicityAfrican 2 (3) 4 (6)East Indian 13 (18) 21 (30)European 29 (41) 24 (34)Far Eastern 8 (11) 4 (6)Other white/Caucasian 13 (18) 9 (13)Other 6 (8) 8 (11)

Weight (kg) 84 6 19 85 6 20

BMI (kg/m2) 31 6 6 30 6 5

Waist (cm) 106 6 14 104 6 13

Current smokers 0 0

Duration of diabetes (years) 7.5 6 5.4 7.6 6 6.9

Glucose (mmol/L) 7.5 6 1.6 7.7 6 1.5

HbA1c (%) 7.2 6 0.6 7.4 6 0.6HbA1c (mmol/mol) 55.7 6 6.8 57.1 6 6.9Participants #7.1% 34 (48) 31 (44)Participants .7.1% 37 (52) 39 (56)

Total cholesterol (mmol/L) 3.99 6 1.00 4.15 6 1.12

LDL-C (mmol/L) 2.13 6 0.85 2.25 6 0.90

HDL-C (mmol/L) 1.16 6 0.28 1.20 6 0.30

Triglycerides (mmol/L) 1.52 6 0.80 1.54 6 0.76

Systolic blood pressure (mmHg) 122 6 11 121 6 12

Diastolic blood pressure (mmHg) 72 6 8 71 6 8

Heart rate (bpm) 73 6 10 73 6 11

Absolute CVD risk score† 10.3 6 5.1 9.6 6 3.7

Relative CVD risk score 1.3 6 0.7 1.3 6 0.5

RHI ratio 1.73 6 0.36 1.86 6 0.50

Antihyperglycemic medications 71 (100) 70 (100)Metformin 67 (94) 65 (93)Sulfonylurea 18 (25) 22 (31)Thiazolidinedione 4 (6) 8 (11)Dipeptidyl peptidase-4 inhibitors 12 (17) 12 (17)Meglitinides (nonsulfonylurea) 2 (3) 1 (1)a-Glucosidase inhibitors 0 (0) 1 (1)Injectable GLP-1 analog (Victoza) 0 (0) 1 (1)Combination (Janumet) 2 (3) 2 (3)

Cholesterol-lowering medications 51 (72) 50 (71)

Blood pressure medications 43 (61) 39 (56)

Data aremean6 SD or n (%). RHI, reactive hyperemia index. *No significant differences in baseline(week 0) characteristics were seen between treatments. †CVD risk score was calculated by usingthe Framingham CVD predictive equation by Anderson et al. (22).

1810 Canola Oil and Glycemic Control in Diabetes Diabetes Care Volume 37, July 2014

independent predictor of HbA1c change(P = 0.024), but the effect of the diet onHbA1c remained significant after con-trolling for age. No baseline measurescontributed .10% to the HbA1c effect.

Adverse EventsThere were no serious adverse eventsthat required hospitalization. Five par-ticipants (three on the test diet andtwo on the control diet) were examinedeither by their family physician or at alocal hospital emergency departmentfor events unrelated to the diet. Fivesubjects had repeated HbA1c values.8.5% (69 mmol/mol) (three on thecontrol diet and two on the test diet).Five participants (three on the controldiet and two on the test diet) reportedexperiencing hypoglycemic episodes.

CONCLUSIONS

Increased MUFA and ALA (canola oil)consumption as part of a canola low-GLdiet modestly lowered HbA1c but to a clin-ically significant extent in participantswith raised blood pressure. Togetherwith the reduction in Framingham riskscore, these data support the use of ca-nola oil in type 2 diabetes.

This study is the first to our knowl-edge to combine three dietary strate-gies (n-3 [ALA], MUFA, and low-GLdiets) to manage diabetes that in thelonger term have been associated withreduced CVD risk both in people withand without diabetes (3,4,9,10,27).

Previous meta-analyses of low-GIstudies in type 2 diabetes have dem-onstrated a 0.43% reduction in HbA1c(28), and large studies have reported0.4%–0.5% (4.4–5.5 mmol/mol) HbA1c

reductions in their low-GI or -GL arm(25) similar to that seen in the cur-rent study. Recently, a major Spanishtrial demonstrated a 30% CVD riskreduction after monounsaturated fator nut (including n-3 [ALA]–rich wal-nuts) supplementation in high-risk trialparticipants, including those with type2 diabetes (27). Furthermore, threemeta-analyses of cohort studies indi-cated cardioprotective properties oflow-GL diets in women without diabe-tes (4,29,30). In other studies, partici-pants with increased BMI and insulinresistance but without diabetes demon-strated greater effects of low-GL diets oncardiovascular outcomes and weight loss,respectively (3,31). The current study also

Table

2—Changesfro

mbase

linein

studymeasu

rements

onthebasis

ofraw

data

andsig

nifica

nce

oftre

atm

entdiffe

rences

forraw

andmultip

leim

putatio

n

Contro

ldiet

Testdiet

Betw

eendiets

Week

0(n

=71) b

Change

a

with

indiet

Week

0(n

=70) b

Change

a

with

indiet

Change

aPvalu

e(raw

)Pvalu

e(M

I)

Weigh

t(kg)

84.4(79.9,88.9)

21.6

(22.0,2

1.3)*

84.5(79.7,89.4)

22.1

(22.5,2

1.7)*

20.5

(21.0,0.0)

0.0700.458

Waist

(cm)

106(103,110)

22.4

(22.9,2

1.9)*

104(101,108)

21.8

(22.4,2

1.3)*

0.6(2

0.2,1.3)0.143

0.065

HbA1c(%

HbA1cunit)

7.2(7.1,7.4)

20.31

(20.38,2

0.25)*

7.4(7.2,7.5)

20.47

(20.54,2

0.40)*

20.16

(20.25,2

0.06)0.002

0.016

HbA1c(m

mol/m

ol)

55.7(54.1,57.3)

23.44

(24.17,2

2.71)57.1

(55.4,58.8)25.15

(25.92,2

4.38)21.71

(22.77,2

0.65)

Fastingglu

cose

(mmol/L)

7.5(7.1,7.9)

20.30

(20.48,2

0.12)*

7.7(7.3,8.0)

20.37

(20.56,2

0.18)*

20.07

(20.33,0.19)

0.5910.491

Cholestero

l(mmol/L)

4.0(3.8,4.2)

0.04(2

0.03,0.12)4.1

(3.9,4.4)20.30

(20.38,2

0.22)*

20.34

(20.46,2

0.23)0.000

0.000

LDL-C

(mmol/L)

2.1(1.9,2.3)

0.04(2

0.02,0.11)2.2

(2.0,2.5)20.20

(20.27,2

0.13)*

20.25

(20.34,2

0.15)0.000

0.000

HDL-C

(mmol/L)

1.2(1.1,1.2)

0.00(2

0.02,0.02)1.2

(1.1,1.3)20.03

(20.05,2

0.01)*

20.03

(20.06,0.00)

0.0410.164

Triglycerides

(mmol/L)

1.5(1.3,1.7)

20.01

(20.09,0.07)

1.5(1.4,1.7)

20.15

(20.24,2

0.07)*

20.14

(20.26,2

0.03)0.018

0.085

Totalch

olestero

l/HDL-C

3.6(3.3,3.8)

0.02(2

0.05,0.10)3.6

(3.3,3.8)20.19

( 20.27,2

0.11)*

20.21

(20.32,2

0.11)0.000

0.000

LDL-C

/HDL-C

1.9(1.7,2.1)

0.03(2

0.03,0.09)1.9

(1.7,2.1)20.13

(20.19,2

0.07)*

20.16

(20.24,2

0.07)0.001

0.000

Systolic

BP(m

mHg)

122(120,125)

25.1

(26.7,2

3.5)*

121(118,124)

24.7

(26.4,2

2.9)*

0.4(2

1.9,2.8)0.718

0.892

Diasto

licBP(m

mHg)

72(70,74)

23.3

(24.2,2

2.3)*

71(69,73)

23.0

(24.1,2

2.0)*

0.2(2

1.2,1.7)0.740

0.763

Heart

rate(bpm)

73(71,76)

22.6

(23.6,2

1.6)*

73(70,75)

22.3

(23.4,2

1.3)*

0.2(2

1.2,1.7)0.770

0.898

Absolute

CVDrisk

c(10-year

%)

10.3(9.1,11.5)

20.53

(20.84,2

0.22)*

9.6(8.8,10.5)

21.16

(21.49,2

0.82)*

20.63

(21.09,2

0.17)0.008

0.079

Relative

CVDrisk

1.3(1.2,1.5)

20.07

(20.12,2

0.03)*

1.3(1.2,1.4)

20.16

(20.20,2

0.11)*

20.08

(20.15,2

0.02)0.007

0.049

RHIratio

1.7(1.6,1.8)

0.13(0.00,0.25)

1.9(1.7,2.0)

20.12

(20.24,0.01)

20.24

(20.42,2

0.06)0.009

0.015

Data

aremean

(lower

confiden

celim

it,upper

confiden

celim

it).Physicalan

dbiochem

icalmeasu

reswere

obtain

edatw

eek0,rep

resentin

gbaselin

e,andweeks

8,10,and12

2baselin

e,represen

tingchange

from

baselin

e.BP,bloodpressu

re;MI,m

ultip

leim

putatio

n;RHI,reactive

hyperem

iaindex.*Sign

ifican

tdifferen

cefro

mbaselin

e(P,

0.05).aM

ean,co

nfiden

celim

its,andPvalu

esdeterm

ined

usin

grep

eated-m

easures

leastsquares

mean

sinPROCMIXED

ofSA

S9.2

with

allavailabledata.

bContro

l:n=71

atbaselin

eand64,60,an

d59

atweeks

8,10,and12,resp

ectively.Test:n=70

atbaselin

eand54

atweeks

8,10,and12.

cCVD

riskcalcu

latedusin

gtheFram

ingham

CVDpred

ictiveeq

uatio

nbyAnderso

net

al.(22).

care.diabetesjournals.org Jenkins and Associates 1811

supports the conceptof agreater effective-ness of low-GL diets in insulin-resistantstates, including central adiposity, lowHDL-C, and hypertension (24).Despite the relatively low statin-

treated LDL-C baseline concentrationsof 2.17–2.22 mmol/L, canola oil con-sumption was associated with a signifi-cant additional reduction in LDL-C.According to statin dose-response stud-ies, the observed LDL-C reduction couldtranslate into an extra 7% reduction inCVD events or an additional 20 mg ator-vastatin (32). Earlier studies demon-strated reduced triglyceride and VLDLcholesterol levels with increased MUFAintake in type 2 diabetes (33). To ourknowledge, the current study is one ofthe first to assess the effect on serumlipids and glycemic control of an ALA-rich oil in type 2 diabetes. The effectsof walnuts, as sources of ALA, havebeen studied in type 2 diabetes, and de-spite no effect on HbA1c, they wereshown to reduce LDL-C (34) and improvevascular reactivity (35). In nondiabeticstudy participants, walnut consumptionhas also been associated with a reduc-tion in LDL-C (36).Increased whole-grain intake has con-

sistently been associated with reducedCVD events in cohort studies (12,13)without a clear mechanism for this ben-efit. Whole-wheat fiber is nonviscous,and unlike viscous fibers from oats, bar-ley, and other sources, it does not lowerserum cholesterol (15,37) or reducepostprandial glycemia (38). However,there is evidence that whole-wheatproducts may reduce insulin resistance(39). Thus, this finding together with thepossible improvement in vascular reac-tivity seen here after wheat bran intakemay be part of the explanation for thereduced CVD risk among whole-grainconsumers (11–14,40).A study limitation is the relatively

small effect size of HbA1c, the primaryoutcome, of 0.5% (5.1 mmol/mol) com-pared with the larger than previouslyseen reduction for the control diet of0.3% (3.4 mmol/mol). However, in par-ticipants at increased risk for adverseoutcomes, a clinically significant effectwas observed, especially in those withhypertension, where the HbA1c reduc-tion for the test diet was 0.62% (6.79mmol/mol) and the relative HbA1c re-duction was 0.41% (4.45 mmol/mol)and, therefore, in the range of 0.3%–

0.4% and above that set by Food andDrug Administration guidelines for dia-betes drug development (1). Further-more, the study participants werealready taking one or more oral antihy-perglycemic agents, and 40% of the par-ticipants had HbA1c levels at the clinicaltarget of #7.0% (53 mmol/mol).

The strengths of this study include theparticipant numbers and frequency ofblood sampling that allowed small treat-ment differences to be detected. Fur-thermore, because the baseline HbA1cand blood lipid levels were close to tar-get, it is likely that there may be greaterreductions in participants with higherlevels commonly seen in clinical practice.

The significance of differences havebeen provided for both the raw data,using repeated measures in the mixedmodel, and also where missing valueshave been derived by multiple imputa-tion. Both approaches were similar inidentifying significant differences. Theraw data, however, also show significanttreatment differences, favoring higherHDL-C, lower triglycerides, and lowerabsolute coronary heart disease riskfor the test diet and indicate that olderand more centrally obese (increasedwaist circumference) individuals re-sponded better to the high-canola–low-GL diet. These data support theview that patients at greatest risk bene-fit most (3,24,31). The assessment usingmultiply imputed data failed to reachsignificance for these differences.

In conclusion, the reduction of GL byincreasing the intake of MUFA and ALA(e.g., canola oil) to displace dietary car-bohydrates and reduce the GL improvedglycemic control, particularly in partici-pants at high risk for diabetes complica-tions, and reduced LDL-C, a feature notseen with similar low-GI diets (25). Bycontrast, whole-grain cereals appear toimprove vascular reactivity, possiblyhelping to explain their benefit in CVDrisk reduction.

Acknowledgments. The authors thank SheilaWest, Departments of Nutritional Sciences andBehavioral Health, Pennsylvania State Univer-sity, for advice on EndoPAT use and helpfulcomments and Quang Dieu from KensingtonNatural Bakers (Toronto, ON, Canada) for pro-viding the study breads.Funding. This work was supported by theCanola Council of Canada, Agriculture andAgri-Food Canada, and Loblaw Companies,

Canada. D.J.A.J. has received salary supportas a Canada Research Chair from the federalgovernment of Canada and has received variousfunding from the Canadian Institutes of HealthResearch, Canada Foundation for Innovation,Ontario Research Fund, Canola Council ofCanada, The International Tree Nut CouncilNutrition Research & Education Foundation, Al-pro Foundation, and Peanut Institute. C.W.C.K.has received support from the American Pista-chio Growers, Canadian Institutes of Health Re-search, Canola Council of Canada, andInternational Tree Nut Council. R.J.d.S. is a re-cipient of a postdoctoral research fellow-ship from the Canadian Institutes of HealthResearch.Duality of Interest. This work was supportedby Loblaw Companies. All authors have com-pleted and submitted the International Com-mittee of Medical Journal Editors Form forDisclosure of Potential Conflicts of Interest.D.J.A.J. has served on the scientific advisoryboard of Unilever, Sanitarium Company, Cali-fornia Strawberry Commission, Loblaw Super-market, Herbal Life International, NutritionalFundamentals for Health, Pacific Health Labora-tories, Metagenics, Bayer Consumer Care, Orafti,Dean Foods, Kellogg’s, Quaker Oats, Procter &Gamble, The Coca-Cola Company, NuVal GriffinHospital, Abbott, Pulse Canada, and Saskatche-wan Pulse Growers; has received honorariafor scientific advice from the Almond Board ofCalifornia, Barilla, Unilever Canada, Solae, Old-ways, Kellogg’s, Quaker Oats, Procter & Gam-ble, The Coca-Cola Company, NuVal GriffinHospital, Abbott, Dean Foods, California Straw-berry Commission, and Haine Celestial; hasbeen on the speakers panel for the AlmondBoard of California; has received researchgrants from Loblaw Brands Ltd., Unilever, Bar-illa, Almond Board of California, Solae, HaineCelestial, Sanitarium Company, and Orafti;and has received travel support to meetingsfrom the Almond Board of California, Unilever,Canola Council of Canada, Barilla, Oldways,and the Nutrition Foundation of Italy. V.V. andD.J.A.J.’s wife are part owners of Glycemic IndexLaboratories, Inc., a contract research organiza-tion, Toronto, ON, Canada. C.W.C.K. has receivedresearch grants, travel funding, consultant fees,or honoraria or has served on the scientific advi-sory board for Abbott, Advanced Food MaterialsNetwork, Almond Board of California, AmericanPeanut Council, Barilla, California StrawberryCommission, Danone, General Mills, Haine Celes-tial, Kellogg’s, LoblawBrands Ltd., Oldways,Orafti,Paramount Farms, Pulse Canada, SaskatchewanPulse Growers, Solae, and Unilever. L.C. holds acasual clinical research coordinator position atGlycemic Index Laboratories (Toronto, ON, Can-ada). R.J.d.S. is a coapplicant on unrestricted re-search grants awarded to D.J.A.J. from TheCoca-Cola Company and the Calorie ControlCouncil. J.L.S. has received research supportfrom the Canadian Institutes of Health Research,Calorie Control Council, The Coca-Cola Company(investigator initiated, unrestricted), Dr. PepperSnapple Group (investigator initiated, unre-stricted), Pulse Canada, and The InternationalTree Nut Council Nutrition Research & EducationFoundation. He has received travel funding,speaker fees, and/or honoraria from theAmerican

1812 Canola Oil and Glycemic Control in Diabetes Diabetes Care Volume 37, July 2014

Heart Association, American College of Physi-cians, American Society for Nutrition, NationalInstitute of Diabetes and Digestive and KidneyDiseases of the National Institutes of Health, Ca-nadian Diabetes Association, Canadian NutritionSociety, University of South Carolina, Universityof Alabama at Birmingham, Oldways Preserva-tion Trust, Nutrition Foundation of Italy, CalorieControl Council, Diabetes and Nutrition StudyGroup of the European Association for the Studyof Diabetes (EASD), International Life SciencesInstitute North America, International Life Scien-ces Institute Brazil, Abbott Laboratories, PulseCanada, Canadian Sugar Institute, Dr. PepperSnapple Group, The Coca-Cola Company, andCorn Refiners Association. He is on the ClinicalPractice Guidelines Expert Committee for Nutri-tion Therapy of both the Canadian Diabetes As-sociation and EASD, as well as on an AmericanSociety for Nutrition writing panel for a scientificstatement on the metabolic and nutritional ef-fects of fructose, sucrose, and high-fructose cornsyrup. He is a member of the International Car-bohydrate Quality Consortium and Board Mem-ber of the Diabetes and Nutrition Study Groupof the EASD. He serves as an unpaid scientificadvisor for the International Life SciencesInstitute North America, Food, Nutrition, andSafety Program. His wife is an employee of Uni-lever Canada. No other potential conflicts of in-terest relevant to this article were reported.Author Contributions. D.J.A.J. contributed tothe study supervision, concept, and design; dataanalysis and interpretation, drafting of themanuscript; and critical revision of the manu-script for important intellectual content. C.W.C.K.obtained funding and contributed to the studysupervision, concept, and design; data analysisand interpretation; and critical revision of themanuscript for important intellectual content.V.V. contributed to concept, design, data acqui-sition, and critical revision of the manuscriptfor important intellectual content. D.F. contrib-uted to the study supervision, data acquisition,and critical revision of the manuscript forimportant intellectual content. L.S.A.A. contrib-uted to the data acquisition, analysis, and in-terpretation and critical revision of themanuscriptfor important intellectual content. S.M. contrib-uted to the data acquisition and critical revisionof the manuscript for important intellectualcontent. C.I. contributed administrative, techni-cal, and material support and to the dataacquisition, analysis, and interpretation and sta-tistical analysis. K.S. contributed to the dataanalysis and interpretation, statistical analysis,drafting of the manuscript, and critical revision ofthe manuscript for important intellectual con-tent. A.M. contributed administrative, technical,and material support and to the data acquisitionand critical revision of the manuscript for impor-tant intellectual content. L.C. contributed admin-istrative, technical, and material support and tothe data acquisition and critical revision of themanuscript for important intellectual content.S.B.M. contributed to the data acquisition andcritical revision of the manuscript for importantintellectual content. S.N. contributed to thedata acquisition and critical revision. S.S.-P. andJ.C. contributed to the data acquisition. D.P.contributed to the data acquisition. B.B. con-tributed administrative, technical, and material

support. E.V. contributed to the data analysisand interpretation and statistical analysis. R.J.d.S.contributed to the data analysis and interpreta-tion, statistical analysis, and critical revision ofthe manuscript for important intellectual con-tent. J.L.S., R.G.J., and L.A.L. contributed to thecritical revision of the manuscript for importantintellectual content. D.J.A.J., L.S.A.A., C.I., E.V.,and R.J.d.S. are the guarantors of this work and,as such, had full access to all the data in the studyand take responsibility for the integrity of thedata and the accuracy of the data analysis.Prior Presentation. Parts of this study werepresented at the 74th Scientific Sessions of theAmerican Diabetes Association, San Francisco,CA, 13–17 June 2014.

References1. U.S. Food and Drug Administration. Guid-ance for industry: diabetes mellitusdevaluat-ing cardiovascular risk in new antidiabetictherapies to treat type 2 diabetes [article on-line], 2008. Available from http://www.fda.gov/downloads/Drugs/GuidanceCompliance-RegulatoryInformation/Guidances/ucm071627.pdf. Accessed 30 July 20132. Gerstein HC, Miller ME, Byington RP, et al.;Action to Control Cardiovascular Risk in Diabe-tes Study Group. Effects of intensive glucoselowering in type 2 diabetes. N Engl J Med2008;358:2545–25593. Liu S, Willett WC, Stampfer MJ, et al. A pro-spective study of dietary glycemic load, carbo-hydrate intake, and risk of coronary heartdisease in US women. Am J Clin Nutr 2000;71:1455–14614. Mirrahimi A, de Souza RJ, Chiavaroli L, et al.Associations of glycemic index and load withcoronary heart disease events: a systematic re-view and meta-analysis of prospective cohorts.J Am Heart Assoc 2012;1:e0007525. Livesey G, Taylor R, Livesey H, Liu S. Is therea dose-response relation of dietary glycemicload to risk of type 2 diabetes? Meta-analysisof prospective cohort studies. Am J Clin Nutr2013;97:584–5966. Canadian Diabetes Association Clinical Prac-tice Guidelines Expert Committee. Clinicalpractice guidelines: definition, classificationand diagnosis of diabetes, prediabetes andmetabolic syndrome. Can J Diabetes 2013;37:S8–S117. Evert AB, Boucher JL, CypressM, et al.; Amer-ican Diabetes Association. Nutrition therapyrecommendations for the management ofadults with diabetes. Diabetes Care 2013;36:3821–38428. Cefalu WT. American Diabetes Association-European Association for the Study of Diabetesposition statement: due diligence was con-ducted. Diabetes Care 2012;35:1201–12039. de Lorgeril M, Renaud S, Mamelle N, et al.Mediterranean alpha-linolenic acid-rich diet insecondary prevention of coronary heart dis-ease. Lancet 1994;343:1454–145910. Brouwer IA, KatanMB, Zock PL. Dietary a-li-nolenic acid is associated with reduced risk offatal coronary heart disease, but increased pros-tate cancer risk: a meta-analysis. J Nutr 2004;134:919–92211. Schulze MB, Liu S, Rimm EB, Manson JE,Willett WC, Hu FB. Glycemic index, glycemic

load, and dietary fiber intake and incidence oftype 2 diabetes in younger and middle-agedwomen. Am J Clin Nutr 2004;80:348–35612. Jacobs DR Jr, Meyer KA, Kushi LH, FolsomAR. Is whole grain intake associated with re-duced total and cause-specific death rates inolder women? The IowaWomen’s Health Study.Am J Public Health 1999;89:322–32913. Liu S, Stampfer MJ, Hu FB, et al. Whole-grain consumption and risk of coronary heartdisease: results from the Nurses’ Health Study.Am J Clin Nutr 1999;70:412–41914. Mellen PB, Walsh TF, Herrington DM.Whole grain intake and cardiovascular disease:a meta-analysis. Nutr Metab Cardiovasc Dis2008;18:283–29015. Jenkins DJ, Kendall CW, Augustin LS, et al.Effect of wheat bran on glycemic control andrisk factors for cardiovascular disease in type2 diabetes. Diabetes Care 2002;25:1522–152816. Jenkins DJ, Kendall CW, Banach MS, et al.Nuts as a replacement for carbohydrates in thediabetic diet. Diabetes Care 2011;34:1706–171117. Jenkins DJ, Kendall CW, Augustin LS, et al.Effect of legumes as part of a low glycemic indexdiet on glycemic control and cardiovascular riskfactors in type 2 diabetesmellitus: a randomizedcontrolled trial. Arch Intern Med 2012;172:1653–166018. Skulas-Ray AC, Kris-Etherton PM, HarrisWS, Vanden Heuvel JP, Wagner PR, West SG.Dose-response effects of omega-3 fatty acidson triglycerides, inflammation, and endothelialfunction in healthy persons with moderatehypertriglyceridemia. Am J Clin Nutr 2011;93:243–25219. U.S. Department of Agriculture. Composi-tion of Foods, Agriculture Handbook No. 23:The Agriculture Research Service. Washington,DC, U.S. Dept of Agriculture, 201020. Atkinson FS, Foster-Powell K, Brand-MillerJC. International tables of glycemic index andglycemic load values: 2008. Diabetes Care2008;31:2281–228321. International Standards Organisation. ISO26642-2010. Food Products - Determination ofthe Glycaemic Index (GI) and Recommenda-tion for Food Classification. Geneva, Switzer-land, International Standards Organisation,201022. Anderson KM,Wilson PW, Odell PM, KannelWB. An updated coronary risk profile. A state-ment for health professionals. Circulation 1991;83:356–36223. SAS Institute Inc. SAS 9.2 ETL Studio: User’sGuide. Cary, NC, SAS Institute Inc., 200424. Grundy SM, Brewer HB Jr, Cleeman JI, SmithSC Jr, Lenfant C; American Heart Association;National Heart, Lung, and Blood Institute. Defi-nition of metabolic syndrome: report of theNational Heart, Lung, and Blood Institute/Amer-ican Heart Association conference on scientificissues related to definition. Circulation 2004;109:433–43825. Jenkins DJ, Kendall CW, McKeown-EyssenG, et al. Effect of a low-glycemic index or ahigh-cereal fiber diet on type 2 diabetes: a ran-domized trial. JAMA 2008;300:2742–275326. Borm GF, Fransen J, LemmensWA. A simplesample size formula for analysis of covariance in

care.diabetesjournals.org Jenkins and Associates 1813

randomized clinical trials. J Clin Epidemiol 2007;60:1234–123827. Estruch R, Ros E, Salas-Salvado J, et al. Primaryprevention of cardiovascular disease with a Med-iterranean diet. N Engl JMed 2013;368:1279–129028. Brand-Miller J, Hayne S, Petocz P, Colagiuri S.Low-glycemic index diets in the management ofdiabetes: a meta-analysis of randomized con-trolled trials. Diabetes Care 2003;26:2261–226729. Dong JY, Zhang YH, Wang P, Qin LQ. Meta-analysis of dietary glycemic load and glycemicindex in relation to risk of coronary heart dis-ease. Am J Cardiol 2012;109:1608–161330. Fan J, Song Y, Wang Y, Hui R, Zhang W. Di-etary glycemic index, glycemic load, and risk ofcoronary heart disease, stroke, and stroke mor-tality: a systematic review with meta-analysis.PLoS One 2012;7:e5218231. Ebbeling CB, Leidig MM, Feldman HA,LoveskyMM, LudwigDS. Effects of a low-glycemicload vs low-fat diet in obese young adults: a ran-domized trial. JAMA 2007;297:2092–2102

32. Shepherd J, Barter P, Carmena R, et al. Ef-fect of lowering LDL cholesterol substantiallybelow currently recommended levels in pa-tients with coronary heart disease and diabetes:the Treating to New Targets (TNT) study. Diabe-tes Care 2006;29:1220–122633. Garg A, Bantle JP, Henry RR, et al. Effects ofvarying carbohydrate content of diet in patientswith non-insulin-dependent diabetes mellitus.JAMA 1994;271:1421–142834. Tapsell LC, Gillen LJ, Patch CS, et al. Includ-ing walnuts in a low-fat/modified-fat diet im-proves HDL cholesterol-to-total cholesterolratios in patients with type 2 diabetes. DiabetesCare 2004;27:2777–278335. Ma Y, Njike VY, Millet J, et al. Effects ofwalnut consumption on endothelial functionin type 2 diabetic subjects: a randomized con-trolled crossover trial. Diabetes Care 2010;33:227–23236. Sabate J, Fraser GE, Burke K, Knutsen SF,Bennett H, Lindsted KD. Effects of walnuts on

serum lipid levels and blood pressure in normalmen. N Engl J Med 1993;328:603–60737. Grundy SM, Cleeman JI, Merz CN, et al.;National Heart, Lung, and Blood Institute; Amer-ican College of Cardiology Foundation; Ameri-can Heart Association. Implications of recentclinical trials for the National Cholesterol Edu-cation Program Adult Treatment Panel III guide-lines. Circulation 2004;110:227–23938. Jenkins DJ, Wolever TM, Leeds AR, et al.Dietary fibres, fibre analogues, and glucosetolerance: importance of viscosity. BMJ1978;1:1392–139439. PereiraMA, Jacobs DR Jr, Pins JJ, et al. Effectof whole grains on insulin sensitivity in over-weight hyperinsulinemic adults. Am J Clin Nutr2002;75:848–85540. Ye EQ, Chacko SA, Chou EL, Kugizaki M, LiuS. Greater whole-grain intake is associated withlower risk of type 2 diabetes, cardiovascular dis-ease, and weight gain. J Nutr 2012;142:1304–1313

1814 Canola Oil and Glycemic Control in Diabetes Diabetes Care Volume 37, July 2014