To evaluate the association of second-hand smoke exposure of pregnant mothers using urine cotinine with the neurodevelopment of their children at 18 months of age in the mother-child cohort in Crete (Rhea Study).

Selected participants were Greek mothers with singleton pregnancies, had never smoked, and had available urine cotinine measurements in pregnancy, and their children for whom a neurodevelopmental assessment was completed. We performed face-to-face interviews twice during pregnancy and postnatally, and assessed children's neurodevelopment at 18 months of age using the Bayley Scales of Infant and Toddler Development, Third Edition. We used linear regression and generalized additive models.

Of 599 mothers, 175 (29%) met the inclusion criteria. Maternal urine cotinine levels were low (mean: 10.3 ng/mL, SD: 11.7 ng/mL). Reported passive smoking from different sources was strongly associated with urine cotinine levels. A negative association was observed between cotinine levels in pregnancy and child's gross motor function (beta = −3.22 per 10 ng/mL, 95% CI −5.09 to −1.34) after adjusting for factors potentially associated with neurodevelopment; results were similar in both sexes. A negative association was also observed for cognitive and receptive communication scales but the effect was small and not statistically significant.

Maternal exposure during pregnancy to second-hand smoke measured through urine cotinine was associated with a decrease in gross motor function among 18-month-old children, even at low levels of exposure.