Metformin helps lower-insulin pregnancies

Adjunctive metformin improved neonatal outcomes only in women needing less than 30 insulin units daily.
*Secondary subgroup analysis of randomized controlled trial; Level 2 (OCEBM).

Citation

Shrestha KS, Jensen CE, Boggess KA, Ramos GA, Battarbee AN. The effect of metformin on perinatal outcomes in pregnant women with type 2 diabetes differs by baseline insulin requirements. American Journal of Obstetrics & Gynecology. 2026;966-973. doi:10.1016/j.ajog.2026.04.043

Background

Insulin is first-line drug treatment for type 2 diabetes in pregnancy. Prior large trials found that adding metformin to insulin did not improve overall neonatal outcomes, but benefits may vary by baseline insulin dose.

Patients

785 pregnant women with preexisting type 2 diabetes or diabetes diagnosed early in pregnancy, singleton nonanomalous fetus, enrolled at 10 to 22 weeks and 6 days. Excluded here: no baseline insulin data or no study drug taken.

Intervention

Metformin, up to 1000 mg twice daily, added to insulin.

Control

Placebo added to insulin, the current standard treatment.

Outcome

Composite adverse neonatal outcome: death, low glucose, severe acidemia, shoulder dystocia injury, jaundice treatment, preterm birth, large or small size for gestational age, or low birthweight.

Follow-up Period

Randomization through delivery and 28 neonatal days.

Results

Metformin’s effect differed by baseline insulin dose. Significant findings are shown.

Subgroup Outcome Effect
<30 insulin units/day Composite adverse neonatal outcome RR 0.76 (95% CI 0.59 to 0.98); NNT 6
<30 insulin units/day Preterm birth RR 0.50 (95% CI 0.28 to 0.90); NNT 7
<30 insulin units/day Large for gestational age RR 0.55 (95% CI 0.34 to 0.90); NNT 5
<30 insulin units/day Neonatal intensive care admission RR 0.47 (95% CI 0.28 to 0.80); NNT 6
>60 insulin units/day Change in insulin dose Mean difference −13 units (95% CI −26 to −1)
>90 insulin units/day Change in insulin dose Mean difference −27 units (95% CI −50 to −5)

RR, relative risk; CI, confidence interval; NNT, number needed to treat.

Results used a modified intention-to-treat approach. No neonatal benefit appeared in higher-dose insulin groups. The low-dose composite benefit was supported by reductions in preterm birth and large-for-gestational-age infants.

Limitations

Secondary subgroup analysis with smaller groups and multiple comparisons, so false-positive findings are possible. No direct insulin-resistance measures, limited late-pregnancy glucose data, and no long-term child outcomes.

Funding

Not reported; authors reported no conflicts.

Clinical Application

Do not add metformin routinely; consider it selectively in insulin-treated pregnant patients needing low baseline insulin doses.