Abstract
OBJECTIVE: To evaluate the effect of metformin on immunological outcomes in pregnant women with polycystic ovary syndrome (PCOS) and their offspring. DESIGN: Post hoc analyses of two randomised controlled trials (PregMet and PregMet2) and one follow-up study (PedMet). SETTING: Women followed at multiple hospitals in Norway, Sweden and Iceland, and offspring followed at multiple hospitals in Norway. POPULATION OR SAMPLE: Pregnant women with PCOS, randomised to metformin or placebo from the first trimester to delivery, and offspring exposed to metformin or placebo in utero. METHODS: Maternal infections and allergic diseases in offspring were compared using logistic regression. Maternal body mass index (BMI), offspring BMI z-score and maternal infections were evaluated as effect modifiers or mediators. MAIN OUTCOME MEASURES: Incidence of maternal infections during pregnancy, delivery, and postpartum, and allergic diseases in offspring at 8-year follow-up. RESULTS: Altogether 634 women and 145 offspring were included. Women treated with metformin experienced fewer overall infections during pregnancy (OR = 0.68, 95% CI: 0.50-0.93), particularly viral infections (OR = 0.71, 95% CI: 0.51-0.99). Offspring exposed to metformin in utero had a higher incidence of allergies (OR = 4.83, 95% CI: 1.47-21.8) and eczema (OR = 2.42, 95% CI: 1.14-5.33). Maternal BMI did not modify the effect of metformin, and offspring BMI z-score or maternal infections did not mediate the relationship between metformin treatment and increased allergies and eczema in offspring. CONCLUSIONS: Metformin treatment in pregnant women with PCOS reduced maternal infections during pregnancy and increased the incidence of allergies and eczema in offspring at 8-year follow-up. TRIAL REGISTRATION: ClinicalTrials.gov identifiers: NCT03259919, NCT00159536 and NCT01587378.