This individual participant data meta-analysis pooled ten double-blind placebo-controlled trials of metformin in pregnancies without pre-existing diabetes, obtaining participant-level data for 92.2% of what was available and analysing 2,297 pregnancies with adjustment for maternal age, BMI, gestational age at commencement and baseline glucose.
Metformin did not reduce gestational diabetes. Adjusted odds ratios were 1.00 under both WHO 1999 and NICE 2015 criteria, with only the IADPSG threshold analysis reaching significance after adjustment (adjusted odds ratio 0.71, 95% CI 0.52-0.98). Fasting glucose was marginally lower, by 0.06 mmol/L, with no difference in two-hour values.
The unexpected findings were obstetric rather than metabolic. Gestation was longer by 0.30 weeks (95% CI 0.06-0.54), preterm birth was less likely (adjusted odds ratio 0.64, 95% CI 0.47-0.89), and neonatal head circumference was larger by 2.43 percentiles. Gastrointestinal side effects were commoner. None of this justifies prescribing metformin to prevent gestational diabetes, which it plainly does not do. The preterm birth signal is interesting enough to warrant a trial designed to test it directly, and it should be read as hypothesis-generating from a secondary outcome rather than as an indication.
- No reduction in gestational diabetes under WHO or NICE criteria (adjusted OR 1.00)
- Significant only under IADPSG thresholds after adjustment (aOR 0.71)
- Gestation longer by 0.30 weeks; preterm birth aOR 0.64
- A signal to test directly, not a reason to prescribe
The statistics, in plain English
That the result depends on which diagnostic criteria are applied is the most instructive part: the same pregnancies show no effect under two definitions and a significant effect under a third, which is a reminder that gestational diabetes is a threshold on a continuum rather than a fixed disease. Testing multiple definitions also raises the chance that one reaches significance by accident. The preterm birth finding was not a primary outcome, so it carries the same caution.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for obstetrics & gynaecology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free