Metformin Shows Limited Impact on Neonatal Outcomes in Diabetes During Pregnancy

Published on: 

Results from the MOMPOD RCT indicate metformin added to insulin did not reduce the frequency of a composite adverse neonatal outcome in pregnant adults with diabetes.

Metformin added to insulin to treat pre-existing type 2 diabetes (T2D), or diabetes diagnosed before 23 weeks’ gestation, did not reduce the frequency of a composite neonatal adverse outcome, compared to placebo, according to new research.1

The double-blinded, randomized Medical Optimization of Management of Overt T2D in Pregnancy (MOMPOD) trial evaluated the effect of metformin among pregnant adults with diabetes aged 18 to 45 years across 17 centers in the United States.1

Among nearly 800 adults enrolled in MOMPOD, the study found no significant differences in prespecified secondary outcomes, including maternal hypoglycemia and neonatal fat mass results, or in prespecified subgroup analyses, including maternal body mass index (BMI) and timing of diabetes diagnosis. Compared to participants receiving placebo, however, those randomized to metformin had a lower likelihood of delivering a large-for-gestational-age infant.

“The effect of reduction in odds of a large-for-gestational-age infant observed after addition of metformin to insulin warrants further investigation,” wrote the investigative team, led by Kim A. Boggess, MD, of the University of North Carolina at Chapel Hill.

Preexisting T2D or gestational diabetes early in pregnancy has each been shown to increase the risk for adverse pregnancy outcomes, intrapartum complications, and neonatal hypoglycemia at birth. Professional medical bodies, including the American College of Obstetricians and Gynecologists (ACOG) and the American Diabetes Association (ADA), recommend insulin as first-line pharmacotherapy for preexisting T2D in pregnancy.2 Metformin use, while studied in pregnancies complicated by T2D, is typically reserved for those who cannot use insulin or decline its use.

Screening for eligibility in MOMPOD occurred between June 2017 to November 2021, with a follow-up completed in May 2022. All individuals were treated with insulin and randomly assigned to either twice-daily metformin 1000 mg or an identical placebo through delivery.

Boggess and colleagues hypothesized the addition of metformin would lower the odds of the primary outcome, a composite of neonatal complications, including perinatal death or severe neonatal complications, compared with insulin alone, without increasing adverse perinatal event risk.

Among 2003 eligible participants, 831 were randomized for the study, of which 794 took ≥1 dose of metformin (n = 397) or placebo (n = 397) and were included in the final analysis. The mean age of the population was 32.9 years, 234 participants were Black, and 412 (52%) were Hispanic. Investigators noted the study was halted at 75% accrual for failure to detect a significant difference in the primary outcome.

Upon analysis, the occurrence of neonatal complications was higher than expected and not significantly different between the metformin group (280 [71%]) and the placebo group (292 [74%] of 39) (adjusted odds ratio [aOR], 0.86; 95% CI, 0.63 - 1.19). Regarding the individual components of the primary outcome, the metformin group had lower odds of large-for-gestational-age infants compared with placebo (OR, 0.63 [95% CI, 0.46 - 0.86).

The analysis of prespecified secondary outcomes also identified that clinically relevant maternal hypoglycemia was similar between the groups (aOR, 1.02 [95% CI, 0.72 - 1.46). Compared with the placebo group, the mean neonatal birthweight and mean birthweight z score were lower among patients in the metformin group. Prespecified subgroup analyses of the primary outcome revealed no significant differences in composite adverse neonatal outcomes in subgroup analyses by maternal BMI of ≥30, timing of diabetes diagnosis, or gestational age at randomization.

The metformin group also showed significantly better glycemic control than placebo (hemoglobin A1c [HbA1c], 5.97% vs. 6.22%. However, HbA1c data were collected only if performed as part of clinical care – approximately 39% of patients had HbA1c data in MOMPOD.

An accompanying editorial from​​ Denice S. Feig, MD, of the department of medicine at the University of Toronto compared the similar findings of MOMPOD to the Metformin in Women With Type 2 Diabetes in Pregnancy (MiTy) trial.3 Both MiTy and MOMPOD showed a lack of significant difference between groups in the primary outcome – however, unlike MiTy, MOMPOD did not identify a reduction in cesarean deliveries, insulin doses, or pregnancy weight gain

“We learn that adding metformin to insulin in those with T2D is safe and results in better glycemic control and reduced births of large-for-gestational-age infants,’ Feig wrote.3 “However, other benefits of metformin noted in MiTy were not seen in the MOMPOD trial, and this finding deserves closer scrutiny.”

Since MOMPOD was stopped for futility in finding a significant difference in the primary outcome, Feig noted this may have limited the ability to learn from secondary analyses. However, Feig indicated the plurality of data supports a role for metformin in individuals with gestational diabetes and T2D during pregnancy, particularly in the short term.

“An individual participant data meta-analysis using MiTy and MOMPOD data is planned, which will hopefully provide a deeper understanding of the effects of metformin in the population of individuals with type 2 diabetes,” Feig wrote.3


  1. Boggess KA, Valint A, Refuerzo JS, et al. Metformin Plus Insulin for Preexisting Diabetes or Gestational Diabetes in Early Pregnancy: The MOMPOD Randomized Clinical Trial. JAMA. 2023;330(22):2182–2190. doi:10.1001/jama.2023.22949
  2. American Diabetes Association Professional Practice Committee. Management of diabetes in pregnancy: standards of medical care in diabetes-2022. Diabetes Care. 2022;45(suppl 1):S232-S243. doi:10.2337/dc22-S015
  3. Feig DS. Metformin for Diabetes in Pregnancy: Are We Closer to Defining Its Role? JAMA. 2023;330(22):2167–2169. doi:10.1001/jama.2023.18589