Metformin in pregnancy and risk of abnormal growth outcomes at birth: a register-based cohort study

We previously reported an increased risk of being small for gestational age (SGA) and a decreased risk of being large for gestational age (LGA) after in utero exposure to metformin compared with insulin exposure. This follow-up study investigated if these observations remain when metformin exposure (henceforth, metformin cohort) is compared with non-pharmacological antidiabetic treatment of gestational diabetes mellitus (GDM; naïve cohort), instead of insulin. RESEARCH DESIGN AND METHODS : This was a Finnish population register-based cohort study from singleton children born during 2004-2016. Birth outcomes from metformin cohort (n=3964) and the naïve cohort (n=82 675) were used in the main analyses. Additional analyses were conducted in a subcohort, restricting the metformin cohort to children of mothers with GDM only (n=2361). Results were reported as inverse probability of treatment weighted OR (wOR), with the naïve cohort as reference. RESULTS  : No difference was found for the outcome of SGA between the cohorts in the main analyses (wOR 0.97, 95% CI 0.73 to 1.27) or in the additional analyses (wOR 1.01, 95% CI 0.75 to 1.37). No difference between the cohorts was found for the risk of LGA (wOR 0.91, 95% CI 0.75 to 1.11) in the main analyses but a decreased risk was observed in the additional analyses (wOR 0.72, 95% CI 0.56 to 0.92). CONCLUSIONS : This follow-up study found no increase in the risk of SGA or LGA after in utero exposure to metformin, compared with drug-naïve GDM. The decreased risk of LGA in mothers with GDM may suggest residual confounding. The lack of increased SGA risk aligns with findings from studies using metformin in non-diabetic pregnancies. In contrast, lower birth weight and increased SGA birth risk were observed in GDM pregnancies for metformin versus insulin. Metformin should be avoided with emerging growth restriction in utero. The interplay of intrauterine hyperglycemia and pharmacological treatments needs further assessment.