Early Metformin in Gestational Diabetes: A Randomized Clinical Trial

Abstract

Importance: Gestational diabetes is a common complication of pregnancy and the optimal management is uncertain.

Objective: To test whether early initiation of metformin reduces insulin initiation or improves fasting hyperglycemia at gestation weeks 32 or 38.

Design, setting, and participants: Double-blind, placebo-controlled trial conducted in 2 centers in Ireland (one tertiary hospital and one smaller regional hospital). Participants were enrolled from June 2017 through September 2022 and followed up until 12 weeks' postpartum. Participants comprised 510 individuals (535 pregnancies) diagnosed with gestational diabetes based on World Health Organization 2013 criteria.

Interventions: Randomized 1:1 to either placebo or metformin (maximum dose, 2500 mg) in addition to usual care.

Main outcomes and measures: The primary outcome was a composite of insulin initiation or a fasting glucose level of 5.1 mmol/L or greater at gestation weeks 32 or 38.

Results: Among 510 participants (mean age, 34.3 years), 535 pregnancies were randomized. The primary composite outcome was not significantly different between groups and occurred in 150 pregnancies (56.8%) in the metformin group and 167 pregnancies (63.7%) in the placebo group (between-group difference, -6.9% [95% CI, -15.1% to 1.4%]; relative risk, 0.89 [95% CI, 0.78-1.02]; P = .13). Of 6 prespecified secondary maternal outcomes, 3 favored the metformin group, including time to insulin initiation, self-reported capillary glycemic control, and gestational weight gain. Secondary neonatal outcomes differed by group, with smaller neonates (lower mean birth weights, a lower proportion weighing >4 kg, a lower proportion in the >90% percentile, and smaller crown-heel length) in the metformin group without differences in neonatal intensive care needs, respiratory distress requiring respiratory support, jaundice requiring phototherapy, major congenital anomalies, neonatal hypoglycemia, or proportion with 5-minute Apgar scores less than 7.

Conclusion and relevance: Early treatment with metformin was not superior to placebo for the composite primary outcome. Prespecified secondary outcome data support further investigation of metformin in larger clinical trials.

Trial registration: ClinicalTrials.gov Identifier: NCT02980276; EudraCT: 2016-001644-19.

Figure 1.. Flow of Patients With Gestational Diabetes Randomized to Metformin vs Placebo

^aNumeric values indicating reasons for exclusion are approximate. ^bSee eBox in Supplement 2 for more information regarding exclusion criteria. ^cFollow-up information until delivery is not shown for 2 pregnancies because delivery took place outside of Ireland.

Figure 2.. Time to Insulin Initiation and Fasting Blood Glucose Between Metformin and Placebo Groups

A, Time to first insulin was computed for each patient as the difference in days between the date of randomization and the earlier of the dates of initiation and delivery. Patients who did not receive insulin over this time period were censored at their date of delivery. Significantly fewer patients in the metformin group required insulin over the course of the study (P value calculated using the log-rank test). B, Fasting glucose (approximate distribution) levels were compared between the metformin and placebo groups at randomization, 32 weeks’ gestation, 38 weeks’ gestation, and 12 weeks’ postpartum. The horizontal lines in the boxes indicate the mean, lower and upper ends of the boxes indicate the first and third quartiles, whiskers indicate values within 1.5× the IQR from the upper or lower quartile, and data more extreme than the whiskers are plotted individually as circles. Mean fasting glucose at randomization was very similar between groups. However, the mean fasting glucose level decreased faster in the metformin group than the placebo group resulting in significant differences at 32 and 38 weeks. These differences were small compared with the standard deviation/IQR of fasting glucose at each time point. To convert glucose levels to mg/dL, divide values by 0.0555.