Type 2 diabetes in pregnancy: effect of metformin added to insulin on neonatal outcome

Optimal management of type 2 diabetes mellitus in pregnancy is uncertain. Our objective was to estimate the effect of metformin added to insulin for maternal type 2 diabetes treatment on composite adverse neonatal outcome. In this multicenter trial, we randomly assigned patients with pre-gestational type 2 diabetes or diabetes diagnosed < 22 weeks’ gestation to insulin/placebo or insulin/metformin until delivery. The primary outcome was a composite of perinatal death or severe neonatal complications; principal secondary outcomes were maternal hypoglycemia and neonatal fat mass at birth. Pre-specified subgroup analyses included patients with BMI > 30 kg/m2 and pre-gestational diabetes. Logistic regression controlling for site, timing of diabetes diagnosis, BMI, and gestational age at enrollment estimated the adjusted odds ratio (aOR) and 95% CI of composite adverse neonatal outcome. 397 were randomized to insulin/placebo and 397 to insulin/metformin. The primary outcome occurred in 71% of infants in the insulin/placebo group and 69% in the insulin/metformin group [relative risk 0.96; 95% CI 0.82-1.12]. Frequency of maternal hypoglycemia was similar between groups. Compared to insulin/placebo, the insulin/metformin group had lower odds of delivery of a large for gestational age (LGA) infant [aOR 0.66, 95%CI 0.50-0.92], but no significance difference in cesarean delivery. There were no significant differences in composite adverse neonatal outcome in subgroup analyses for BMI > 30 kg/m2 or pre-gestational diabetes. There were also no significant differences in maternal/neonatal adverse events or delivery of an SGA infant. Adding metformin to insulin to treat type 2 diabetes in pregnancy did not reduce composite adverse neonatal outcome, reduce maternal hypoglycemia, or infant fat mass. Metformin reduced frequency of an LGA infant without effecting mode of delivery. While the addition of metformin did not demonstrate a decrease in adverse composite neonatal outcome, the reduction in LGA suggests additional studies are needed to measure long term infant metabolic outcomes.