Robust evidence for specific nutrition therapy for gestational diabetes (GDM) is lacking. Preliminary data suggested our CHOICE diet, higher in complex carbohydrate (60%) and lower in fat (25%) reduced fasting glucose (FBG), free fatty acids (FFA), and newborn adiposity (NB%fat). We tested the hypothesis that 7-8 wks of CHOICE would improve insulin resistance, reduce FFA and NB%fat (2-wk PeaPod; 1° powered outcome) vs. a conventional low-carbohydrate (40%), higher fat (45%) diet (LC/CONV). After diagnosis (∼28-30 wks), 59 BMI-matched diet-controlled GDM women (mean±SD; BMI 32±5) were randomized to a eucaloric CHOICE or LC/CONV diet (7.2±1 wks; ALL MEALS PROVIDED). At baseline, a 2-hr 75g OGTT (with insulins) was performed and diet initiated. On day-4, a breakfast meal was given (30% of total calories) with fasting and hourly blood drawn x5 (gluc, insulin, FFA, triglycerides[TG]) for area-under-the curve (AUC). Measures were repeated at 36-37 wks. Of 59, 13 met exclusions (4 diet failures, 2/group). By ANCOVA (n=23/group), total and weight gain during diet were similar (CHOICE 1.9 vs. LC/CONV 1.8 kg) as was delivery wk (39.2 vs. 39.3 wks). At 37 wks, the meal gluc (p=0.001) and insulin AUCs (p=0.013) were lower for LC/CONV, though fasting gluc/insulin were similar. TG increased similarly. The FFA AUC decreased from 30-37 wks on CHOICE but rose on LC/CONV (p=0.016), and was lower for CHOICE at 37 wks (p=0.009). By the 37-wk OGTT, FBG decreased within both groups (CHOICE -7.2, LC/CONV -3.5 mg/dL, both p<0.01) but CHOICE led to improved (p=0.001) and lower gluc AUC (p<0.05)(similar insulin AUCs). Birthweight (3293 vs. 3303 g), anthropometrics, NB%fat (10.8±4 vs. 10.3±4%), and cord blood gluc, C-peptide, FFA and TG were similar. This RCT shows that complex carbohydrate can be liberalized by 20% above conventional recommendations and may improve gluc tolerance and similarly normalize fetal growth, expanding nutrition options in GDM.


T.L. Hernandez: None. S.S. Farabi: None. J. Arbet: None. N. Hirsch: None. E.Z. Dunn: None. K.P. Heiss: None. J. Martin Carli: None. N.F. Krebs: None. T.L. Powell: None. R. Reynolds: None. R.E. Van Pelt: None. J.E. Friedman: None. L.A. Barbour: None.


National Institutes of Health (R01DK101659, UL1TR002535); Janssen Research & Development, LLC

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at