Summary: In pregnant women without diabetes receiving betamethasone for threatened preterm birth, a 48-hour metformin course lowered maternal glucose and roughly halved neonatal hypoglycaemia (21% vs 40%) compared with no treatment, with mild gastrointestinal effects in 14%.
PICO Summary
| Element | Detail |
|---|---|
| Population | 169 pregnant women without diabetes at 24.0–36.5 weeks receiving betamethasone for preterm-birth risk. |
| Intervention | Metformin 425 mg three times daily before meals plus 850–1700 mg at 10 pm, for up to 48 hours after the first betamethasone dose. |
| Comparison | No treatment after betamethasone (open-label). |
| Outcome | Lower maternal total (121 vs 127 mg/dL; P=0.01) and postprandial glucose (129 vs 138 mg/dL; P=0.009). Neonatal hypoglycaemia 21% vs 40% (P=0.04; RR 0.53; 95% CI 0.28–0.99). Mild GI effects in 14%. |
Metformin after antenatal steroids
Open-label RCT · non-diabetic pregnancy · 48 h
A 48-hour metformin course after betamethasone lowered maternal glucose and roughly halved neonatal hypoglycaemia (21% vs 40%). The result is statistically fragile and open-label, so it warrants placebo-controlled replication before routine use.
Expert Commentary
This is a clever trial that targets a problem I have always managed reactively, by watching maternal glucose after steroids and bracing for neonatal hypoglycaemia, rather than trying to prevent it. The idea of a cheap oral agent blunting that iatrogenic glucose surge is genuinely appealing, and what strikes me most is the leverage: a maternal glucose difference of only six to nine mg/dL translated into roughly halving neonatal hypoglycaemia, a vivid reminder of how sensitive the fetus is to the hyperglycaemic peak. My enthusiasm is real but measured, because the neonatal result is statistically fragile, with an upper confidence bound of 0.99 sitting right against the null, and the trial was open-label against no treatment rather than placebo. One supportive single-centre study with a non-standard dosing schedule is not yet practice. Can I use this with my patients? Not as routine, but it is exactly the kind of finding I would raise within a local protocol discussion, and I would not abandon neonatal glucose monitoring whatever we decide. I want this replicated, ideally placebo-controlled, before metformin prophylaxis becomes standard around antenatal steroids.
References
Yefet E, Massalha M, Talmon G, et al. Metformin, maternal glycemic control, and neonatal hypoglycemia after antenatal steroids: a randomized clinical trial. JAMA Netw Open. 2026;9(1):e2552807. doi:10.1001/jamanetworkopen.2025.52807