Study finds no link between prenatal metformin and child developmental delays

Children whose mothers took metformin during pregnancy do not face higher risks of developmental delays when they enter full-time schooling, according to new research published in JAMA Network Open. The study addresses growing concerns about the long-term effects of this widely prescribed medication on fetal development.

Why researchers examined prenatal exposure

Metformin crosses the placenta, resulting in fetal blood concentrations similar to those of the mother. While short-term safety data are reassuring, the impact on long-term neurodevelopment has remained unclear. Previous animal studies suggested potential effects on cortical development, though human research has not consistently linked early exposure to neurodevelopmental delays in younger children.

Conflicting findings regarding childhood adiposity have persisted due to small sample sizes and varied treatment indications. Researchers noted that developmental differences might become more apparent after preschool age, necessitating larger studies focused on school-aged children.

How the Australian cohort was analyzed

The team conducted a cohort study involving singleton births in Victoria, Australia, between January 1, 2009, and December 31, 2020. Data were linked from multiple sources, including the Victorian Perinatal Data Collection (VPDC), National Diabetes Service Scheme (NDSS), Pharmaceutical Benefits Scheme (PBS), Australian Early Development Census (AEDC), and Victorian Admitted Episodes Dataset. The study adhered to STROBE guidelines for observational research.

Exposure was defined as dispensing at least one prescription between the last normal menstrual period and birth. The AEDC, a teacher-completed assessment for children aged 4 to 6 years, measured developmental vulnerability. This was defined as scoring below the 10th percentile in at least two of five domains: physical health, social competence, emotional maturity, language and cognitive skills, and communication and general knowledge.

What the data revealed about outcomes

Among 871,627 singleton births, 177,409 children had linked developmental outcome data. Of these, 1,095 children (0.6%) had antenatal metformin exposure. Most exposed children were assessed between 5 years 1 month and 6 years of age. Within the exposed group, 64.2% received their first prescription in the first trimester, 12.9% in the second, and 22.9% in the third.

Women exposed to metformin had higher mean body mass index (BMI) rates and were more likely to conceive using assisted reproductive technology (ART). They also more commonly had polycystic ovary syndrome, type 2 diabetes, or gestational diabetes. Their infants were more likely to be born preterm, via cesarean delivery, or with birth weights above the 97th percentile. Birth weights below the 10th percentile were similar between groups.

Does adjustment eliminate observed disparities?

Initial unadjusted data showed developmental vulnerability in 18.2% of exposed children compared with 13.9% of unexposed children. However, this difference disappeared after adjusting for confounding factors such as maternal age, BMI, education, Socio-Economic Indexes for Areas (SEIFA), diabetes status, and ART use. Adjusted analyses found no association between metformin exposure and any of the five individual AEDC domains.

Sensitivity analyses confirmed these results. Among women with gestational or type 2 diabetes, vulnerability occurred in 20.7% of exposed children versus 15.9% of unexposed children, but this gap also vanished after adjustment. Excluding children with special educational needs or incomplete data yielded similar findings, with no increased risk observed in any developmental domain.

Limitations and clinical implications

As an observational study, residual confounding could not be ruled out. Limitations included the inability to verify actual medication intake, lack of data on specific doses or treatment indications, and insufficient statistical power for fully adjusted models regarding multiple prescriptions or later-trimester initiation.

The findings offer reassurance to clinicians and pregnant women considering metformin, showing no association with developmental vulnerability in the first year of school. These results held consistent across all five AEDC domains and in subgroups restricted to diabetic pregnancies or first-trimester exposure. Longer-term research remains necessary to assess impacts on children’s growth and cardiometabolic health.