OBJECTIVE
To assess the risk of major congenital malformations with metformin versus insulin in pregnancies with type 2 diabetes.

RESEARCH DESIGN AND METHODS
This cohort study used four Nordic countries’ nationwide registers of live and stillborn infants exposed to metformin or insulin during first trimester organogenesis. Main exclusion criteria were type 1 diabetes, polycystic ovary syndrome, fertility treatment, and exposure to other diabetes drugs. Adjusted risk ratios (RRs) and 95% CIs were estimated for any and cardiac malformations.

RESULTS
Of 3,734,125 infants in the source population, 25,956 were exposed to metformin or insulin in the first trimester, and 4,023 singleton infants were included. A malformation was diagnosed in 147 (4.7%) of 3,145 infants with exposure to any metformin (alone or in addition to insulin) and 50 (5.7%) of 878 infants with exposure to insulin alone (RR 0.84, 95% CI 0.46–1.54). Among 2,852 infants exposed to metformin alone and 293 infants exposed to metformin in addition to insulin 127 (4.4%) and 20 (6.8%), respectively, had a malformation. The adjusted risk was not increased for either metformin alone (0.83, 0.44–1.58) or both metformin and insulin (0.98, 0.56–1.69) versus insulin alone. Corresponding RRs for cardiac malformations were 1.01 (0.55–1.84) for any metformin, 0.92 (0.47–1.81) for metformin alone, and 1.72 (0.76–3.91) for both metformin and insulin.

CONCLUSIONS
No evidence of an increased malformation risk with metformin versus insulin in the first trimester was found. Results should be interpreted with caution since information on glycemic control was missing.