Maternal Urinary Cotinine Concentrations During Pregnancy Predict Infant BMI Trajectory After Birth: Analysis of 89617 Mother-Infant Pairs in the Japan Environment and Children's Study.

Background: Clinical or epidemiological conclusions remain undecided on the direct effects of active and second-hand smoking during pregnancy on childhood obesity. Urinary cotinine (UC) concentration, an accurate and quantitative marker for smoking, may elucidate the dose-dependent relationship between smoking during pregnancy and childhood obesity. To analyze the relationship between UC concentration and smoking questionnaire (SQ) classes for active and second-hand smoking in pregnant mothers and trajectory of infant Kaup index (body mass index: BMI). Methods: This multicenter prospective cohort study was conducted using a list-wise complete set of 35829 among 89617 mother-infant singleton pairs, recruited between 2011 and 2014, in the Japan Environment and Children's Study (JECS). Pairs were categorized according to UC levels (1 to 4 classes) or SQ (0 to 4 classes). Results: Maternal BMI at delivery was the highest in UC class 4 (highest). Maternal and paternal education of ≥16 years and annual household income were lowest in UC class 4. Infant BMI was lower at birth, but trends in BMI and ΔBMI were higher from six to 36 months step-wise in the UC classes. The above tendency was observed in the list-wise complete dataset but was emphasized after multiple imputations and corrections of cofounders. UC concentration in five SQ classes largely fluctuated, and the relationship between SQ classes and trends in BMI and ΔBMI was not statistically significant. Conclusion: Infants from high UC mothers had a low BMI at birth, increasing from six to 36 months of age. UC concentrations, but not smoking questionnaire classes, predict infant BMI trajectory, suggesting that active and second-hand smoking affect child obesity in a dose-dependent manner.