Randomized controlled trial of prepregnancy lifestyle intervention to reduce recurrence of gestational diabetes mellitus

Preconception lifestyle intervention holds potential for reducing gestational diabetes mellitus, but clinical trial data are lacking.This study aimed to determine the effects of a prepregnancy weight loss intervention on gestational diabetes mellitus recurrence in women with overweight/obesity and previous gestational diabetes mellitus.A 2-site, randomized controlled trial comparing a prepregnancy lifestyle intervention with educational control was conducted between December 2017 and February 2022. A total of 199 English- and Spanish-speaking adults with overweight/obesity and previous gestational diabetes mellitus were randomized to a 16-week prepregnancy lifestyle intervention with ongoing treatment until conception or educational control. The primary outcome was gestational diabetes mellitus recurrence. Analyses excluded 6 participants who conceived but did not have gestational diabetes mellitus ascertained by standard methods.In the 63 (33%) women who conceived and had gestational diabetes mellitus ascertained (Ns=38/102 [37%] intervention vs 25/91 [28.0%] control; P=.17), those in the intervention group had significantly greater weight loss at 16 weeks compared with controls (4.8 [3.4-6.0] vs 0.7 [-0.9 to 2.3] kg; P=.001) and a greater proportion lost ≥5% of body weight (50.0% [17/34] vs 13.6% [3/22]; P=.005). There was no significant difference in the incidence of gestational diabetes mellitus recurrence between the intervention (57.9% [ns=23/38]) and the control group (44.0% [ns=11/25]; odds ratio, 1.8 [0.59-5.8]). Independent of group, greater prepregnancy weight loss predicted 21% lower odds of gestational diabetes mellitus recurrence (odds ratio, 0.79 [0.66-0.94]; P=.008). A ≥5% weight loss before conception reduced the odds of gestational diabetes mellitus recurrence by 82% (odds ratio, 0.18 [0.04-0.88]; P=.03).Lifestyle intervention produced considerable prepregnancy weight loss but did not affect gestational diabetes mellitus rates. Given that the conception rate was 50% lower than expected, this study was underpowered.