Infants exposed to metformin during pregnancy showed no developmental delays related to the medication by age 5, according to a recent study published in JAMA Network Open.

The findings address a longstanding safety concern for pregnant women with diabetes seeking an affordable, oral alternative to daily insulin injections. Because metformin crosses the placenta, lingering questions about its long-term impact on offspring have fueled clinical hesitancy, even as the American College of Obstetricians and Gynecologists lists it as a reasonable alternative to insulin for those with diabetes. 

"There's a bit of hesitancy using metformin in pregnancy because we know that it crosses the placenta, and the long-term impacts for the child have been less well described in the past," said Hannah Gordon, MD, a perinatal epidemiology researcher at Mercy Hospital for Women in Heidelberg, Australia, who led the study. "We found there was no difference in outcomes across all of the domains assessed." 

About 1 in 9 women have diabetes in the U.S., with most cases being type 2. Additionally, up to 9% of pregnancies are complicated by gestational diabetes. 

The drug is a preferred course of treatment for many women opposed to insulin, which is generally more expensive and requires self-administered injections up to four times a day, Gordon said. 

In the retrospective population study, Gordon and her team analyzed state-level data of 177,409 children (51.1% male) in Victoria, Australia, born between 2009 and 2020. Among 1095 children with in utero metformin exposure, the primary indication for maternal use was gestational diabetes (39.2%), followed by polycystic ovary syndrome (22.8%) and type 2 diabetes (17.9%). 

The research team compared metformin exposure to national data assessing children's development during their first year of full-time school between ages of 4 and 6. Children were tested in 2015, 2018, or 2021, with more than three-quarters being age 5. 

Children had been evaluated on physical health, social competence, emotional maturity, school-based language and cognitive skills, and communication. The researchers defined developmental vulnerability as a score below the 10th percentile in two or more areas. 

"Ages 4 to 6 is a sweet spot in that we've got high-quality data that is validated, but is also predictive of future educational outcomes," Gordon said. 

After controlling for confounding variables such as maternal age, body mass index, education, and socioeconomic status, metformin use during pregnancy showed no association with the children's neurodevelopment (adjusted risk ratio [ARR], 0.97; 95% CI, 0.74-1.29). 

The researchers also looked at women with gestational diabetes or type 2 diabetes (n = 16 ,121). In this subsample, 20.7% of children exposed to metformin were considered developmentally vulnerable, compared to 15.9% of children without exposure. But no statistically significant association was found after adjustment (ARR, 1.08; 95% CI, 0.86-1.35). 

Many pregnant patients worry about the long-term effects of starting or continuing metformin, said Lydia Shook, MD, a maternal-fetal medicine specialist at Mass General Hospital in Boston. 

Smaller, randomized trials and observational studies had shown longer-term safety for offspring and mothers, and the current research adds population-level outcomes, she said.  

"Now we have long-term safety data to support its use," Shook said. "It's helpful to have a large, population-based study to say, 'We don't think there's a signal towards harm.' " 

Gordon and Shook reported having no financial conflicts of interest. The research received grant funding from Norman Beischer Medical Research Foundation and the Australasian Diabetes in Pregnancy Society.

Kelsey Mesmer, PhD, is a freelance journalist and journalism professor at Saint Louis University in St. Louis.