Our aim was to ascertain whether metformin can reduce insulin requirement without compromising glycaemic control during pregnancy in women with type 1 diabetes.

A total of 126 pregnant women with type 1 diabetes were recruited for a randomised, double-blind, placebo-controlled multicentre study. The primary outcome was total insulin change, defined as the difference between baseline and third trimester maximum insulin dose (IU).

Fifty women in the placebo group and 51 women in the metformin group completed the study. A predetermined sample size of 200 participants was not achieved. There was no significant difference in the primary outcome, that is, in the change of total insulin requirement (33 vs. 27 IU, p = 0.193). However, the metformin group showed a significantly lower increase in the prandial insulin change, with 24 versus 14 IU (p = 0.014) and 0.3 versus 0.2 IU/kg (p = 0.048). In the exploratory subgroup analysis, metformin attenuated prandial insulin increase in women with high BMI (> 25 kg/m²) or high baseline insulin requirement (> 40 IU) (25 vs. 15 IU, p = 0.028, 30 vs. 14 IU, p = 0.007). Weight gain remained more often within target in the metformin group (20% vs. 40%, p = 0.029). A similar weight benefit was observed in subgroups (BMI> 25 kg/m² 8% vs. 32%, p = 0.005, insulin requirement> 40 IU 6% vs. 34%, p = 0.004). No differences were seen in glycaemic control or neonatal outcome between the groups.

Metformin was not shown to affect total insulin change but reduced the prandial insulin change and improved weight gain control especially in insulin-resistant subgroups. These findings warrant further studies on metformin as an adjunctive medicine.