Gestational Diabetes Mellitus and School-Age Cognitive and Health Outcomes in New Zealand: A Matched Cohort Study.

Abstract

To determine if children exposed in utero to gestational diabetes mellitus (GDM) have poorer school-age neurocognitive and cardiometabolic function than matched children not exposed to GDM. In a matched, cohort study, we compared 186 children of women diagnosed with GDM from a trial cohort with 186 children born at the same hospital whose mothers did not have GDM, matched for child sex, maternal body mass index (BMI), ethnicity, socio-economic status, and gestational age. Both GDM-exposed and nonGDM-exposed children were assessed at 6-7 years' corrected age for neurocognitive function (NIH Toolbox), numeracy (Checkout Game), body composition, and peripheral/central blood pressure. The primary outcome was the Composite Cognitive Standard Score. Analysis was by generalized linear models, adjusted for matching criteria; exposure effects were estimated as OR or mean difference (MD), with 95% confidence intervals. Exposure groups were similar for birthweight z-score, corrected age at assessment, matching criteria, and hypertensive disorders of pregnancy. There was no difference between groups in mean cognitive score (104[15] vs 104[15]; MD = 2[-2, 5]) or its components. GDM-exposed children were less likely to have motor impairment (6% vs 11%, aRD = -6[-11, -1] but more likely to have low numeracy (10% vs 4%, aRD = 6[2, 11]) and hypertension (14% vs 6%, aRD = 8[1, 14]), and had higher mean BMI z-scores (0.95[1.61] vs 0.51[1.50], MD = 0.39[0.13, 0.65]). Children exposed to GDM are not at increased risk of poorer school-age cognitive function compared with controls matched for potential confounders. However, GDM exposure was associated with increased childhood BMI and hypertension, which may affect subsequent cardiometabolic health.