Metformin use in pregnancy brings benefits to mums but also more risks to babies

In pregnant women with polycystic ovary syndrome (PCOS), treatment with metformin appears to protect against infections throughout pregnancy while also contributing to an increased risk of childhood allergies and eczema in the offspring, according to post hoc analyses of two randomized controlled trials and one related follow-up study.

In a cohort of pregnant women with PCOS from the PregMet and PregMet2 studies, those treated with metformin were less likely to contract infections during pregnancy compared with those who received placebo (42 percent vs 52 percent; odds ratio [OR], 0.68, 95 percent confidence interval [CI], 0.50–0.93). This benefit was driven by a reduction in viral infections (OR, 0.71, 95 percent CI, 0.51–0.99), particularly respiratory tract infections (26 percent vs 33 percent).

The incidence of bacterial infections during pregnancy, at delivery, and postpartum did not differ between metformin- and placebo-treated women. The same was true for fungal infections and the use of antibiotics during pregnancy.

Adjusting for baseline BMI did not alter the results. Notably, women treated with metformin had fewer viral infections during pregnancy compared with those in the placebo group across all BMI categories, although the difference was only statistically significant for those with obesity (OR, 0.52, 95 percent CI, 0.30–0.89).

Negative effects on babies

For children whose mothers had PCOS in the PedMet study, exposure to metformin vs placebo in utero was associated with a much higher incidence of allergies (18 percent vs 4.2 percent) and eczema (35 percent vs 18 percent) at around 8 years of age.

Metformin exposure in utero was associated with a more than fourfold greater odds of developing childhood allergies (OR, 4.83, 95 percent CI, 1.47–21.8) and more than twofold greater odds of experiencing eczema (OR, 2.42, 95 percent CI, 1.14–5.33) during the follow-up.

Offspring BMI z-score and maternal infections during pregnancy did not modify the association between in utero metformin exposure and the risk of allergies and eczema in offspring.

The case for caution

This is the first study to show that metformin reduces infections during pregnancy in women with PCOS and that exposure to the drug in utero leads to increased incidence of allergies and eczema in childhood, according to first study author Mariell Ryssdal, a doctoral candidate at the Norwegian University of Science and Technology in Trondheim, Norway, and colleagues.

“The adverse effects observed in the offspring highlight the need for caution when considering metformin treatment during pregnancy,” they said.

Metformin may help fight off infections in pregnant women with PCOS through metabolic and immunologic pathways. Ryssdal and colleagues noted that the pronounced reduction in maternal viral infections observed in those with obesity is suggestive of the significant metabolic effects of the drug. In a previous study, metformin induced a broad maternal immune mobilization by increasing multifunctional serum cytokines. [J Clin Endocrinol Metab 2023;108:e743-e753]

“Maternal health during pregnancy, including exposure to microbes, impacts the offspring and primes the development of the foetal immune system. Therefore, it is likely that metformin-induced changes in the mother can affect the child and contribute to an increased risk of allergic diseases,” Ryssdal and colleagues explained. [Cell 2021;184:3394.e20-3409.e20; Pediatr Res 2015;77:189-195]

A paradoxical phenomenon in which metformin reduced maternal infections but increased the risk of allergic conditions in the offspring suggests that the infections themselves might be a protective factor for the baby, they added. However, the results of a mediator analysis showing a lack of association between maternal infections and offspring outcomes prevented the authors from drawing a definitive conclusion.

“More research is needed to understand the maternal immunological changes during pregnancy and how they affect the offspring,” they said.

The analysis on maternal infections included 634 women from the PregMet and PregMet2, which examined whether metformin reduced pregnancy complications in the presence of PCOS. Among these, 316 received metformin (median age 30 years, median BMI 28.3 kg/m²) and 332 received placebo (median age 30 years, median BMI 26.7 kg/m²). Baseline maternal characteristics and pregnancy outcomes were similar between the treatment groups, except for fewer preterm births in the metformin group.

For the analysis on childhood allergic conditions, 145 children in the PedMet study were included. PedMet was a follow-up of offsprings born to mothers with PCOS who had participated in the Pilot study and the PregMet study. Offsprings exposed to metformin vs placebo in utero had similar gestational age (median, 281 vs 279 days) and birth weight (median, 3,550 vs 3,500 g), but those in the metformin group had a larger head circumference at birth (median, 36 vs 35 cm) and a higher weight z-score (median, 0.48 vs 0.19) and BMI z-score (median, 0.50 vs 0.00) at follow-up.