Studies Conducted Since 2010 Assessing the Glycemic Impact of the Mediterranean, Vegetarian, and DASH Eating Patterns

StudyDesignnInterventionBlood Glucose Impact
Mediterranean eating pattern
Elhayany et al., 2010 (15)12-month RCT259 (179 completers)Comparison of LCM (35% CHO, 45% fat), TM, and 2003 ADA diet (both TM and ADA diets 50–55% CHO, 30% fat, and 15–20% protein)A1C reduction was significantly greater in the LCM diet than in the ADA diet (−2.0 and −1.6%, respectively; P < 0.022); no statistically significant difference with TM (−1.8%) compared with LCM and ADA diets
Itsiopoulos et al., 2011 (16)12-week randomized crossover clinical trial31 (27 completers)Comparison of Mediterranean diet (40% fat, 44% CHO, 12% protein, 4% alcohol via red wine) ad libitum provided by study to usual careA1C decreased from 7.1% on usual diet to 6.8% on Mediterranean intervention diet (−0.3%, P = 0.012)
Toobert et al., 2011 (17)24-month RCT280 (190 completers)Comparison of a culturally modified Mediterranean eating pattern to enhanced usual care (usual care + option of one free class) in a Latina populationBaseline A1C 8.4% in both groups; at 6 months, A1C improved in intervention (7.8%) compared with usual care (8.4%) (P < 0.05); improvements not sustained at 24 months (intervention 8.4%, usual care 7.8%)
Esposito et al., 2014 (18)4-year follow-up RCT215 (n = 108 for LCM and n = 107 for low-fat)Comparison of LCM to a low-fat diet (≤ 30% kcal from fat and ≤ 10% kcal from saturated fat); 1,500 kcal/day for women and 1,800 kcal/day for men for both groupsDuring first year, LCM group had greater A1C improvement than low-fat group (MD −0.5%, P < 0.001); after 4 years, cumulative incidence of participants requiring diabetes medications was 44% in LCM group vs. 70% in low-fat group (P < 0.001); need for diabetes medications reached in all participants after follow-up of 6.1 years in low-fat group and 8.1 years in LCM group
Carter et al., 2014 (19)Meta-analysis of eight trialsFBG: n = 972; A1C: n = 487Comparison of Mediterranean diets to usual care, low-fat, or Paleolithic diets (lean meat, fish, fruits, leafy and cruciferous vegetables, root vegetables with restricted potato intake, eggs, and nuts)None of the interventions were significantly better than the others at lowering A1C or FBG; compared with usual care, Mediterranean diets with education (WMD −0.31%) and Paleolithic diets with education (WMD −0.21%) significantly reduced A1C, whereas low-fat diets did not; Mediterranean diets did not reduce A1C significantly more than Paleolithic diets
Huo et al., 2015 (20)Meta-analysis of nine trials1,178Comparison of Mediterranean-style diet to control dietsCompared with control diets, Mediterranean-style diet led to greater reductions in A1C (MD −0.30%, P = 0.001) and FBG (MD −0.72 mmol/L, P = 0.007)
Vegetarian eating pattern
Kahleova et al., 2011 (21)12-week RCT74 (68 completers) first 12 weeks of diet onlyComparison of vegetarian diet to conventional diabetes diet; both kcal restricted (500/day)Vegetarian group: A1C change of −0.68%; control group: −0.59%; no statistically significant differences between groups
Yokoyama et al., 2014 (22)Meta-analysis of six trials225Comparison of vegetarian and vegan diet patterns to conventional diabetes diets with animal proteinVegetarian diets associated with a significant reduction in A1C (−0.39% P = 0.001); nonsignificant reduction in FBG concentration (−0.36 mmol/L, P = 0.301)
Viguiliouk et al., 2019 (23)Meta-analysis of nine trials664Comparison of vegetarian and vegan diet patterns to conventional diabetes diets with animal proteinSignificant reduction in A1C and FBG observed for vegetarian diet patterns compared with conventional diets (A1C MD −0.29%, P = 0.0006; FBG MD −0.56 mmol/L, P = 0.01)
Vegan eating pattern
Mishra et al., 2013 (24)18-week RCT291 (211 completers)Comparison of low-fat vegan diet to control group (no dietary changes)A1C reductions of 0.7 and 0.1% in the intervention and control groups, respectively (P < 0.01)
Lee et al., 2016 (25)12-week RCT93Comparison of vegan diet (excluding animal-based food including fish) to a conventional 2011 Korean Diabetes Association dietBoth groups showed significant reductions in A1C, and reductions were significantly different between groups: −0.5 vs. −0.2% for vegan and Korean diets, respectively (P = 0.017 for interaction)
Barnard et al., 2018 (26)20-week randomized translational study45 (40 completers)Comparison of vegan diet to a portion-controlled eating plan (typically restricted by 500 kcal/day)A1C improved significantly in both groups (−0.40%); no significant differences between eating plans (P = 0.68)
DASH eating pattern
Azadbakht et al., 2011 (9)8-week randomized crossover clinical trial44 (31 completers)Comparison of DASH eating pattern to control diet (50–60% CHO, 15–20% protein, <30% fat, and <5% of daily kcal from simple sugars)After following the DASH eating pattern, participants reduced A1C by 1.7% and FBG by 29.4 mg/dL compared with usual care reductions of A1C by 0.5% and FBG by 12.8 mg/dL (P = 0.04)
Paula et al., 2015 (27)4-week RCT40Comparison of DASH eating pattern plus physical activity to 2015 ADA guidelines with usual physical activityA1C change −0.6% for intervention (P = 0.002), −0.5% for control (P = 0.001); no significant difference between groups (P = 0.944); changes in FBG not significant
Chiavaroli et al., 2019 (28)Meta-analysis of two controlled trials for A1C and 10 controlled trials for FBGA1C: n = 65; FBG: n = 826A1C: Comparison of DASH to usual diet FBG: Comparison of DASH to usual diet and a variety of other interventions (weight loss, behavioral interventions, and other diets)DASH resulted in A1C reduction of 0.53% (P <0.001); FBG change not significant
  • CHO, carbohydrate; FBG, fasting blood glucose; kcal, calories; LCM, low-carbohydrate Mediterranean; MD, mean difference; TM, traditional Mediterranean; WMD, weighted mean difference.