Hair Biomarkers as Measures of Maternal Tobacco Smoke Exposure and Predictors of Fetal Growth

Introduction: Most biomarker studies of the effects of maternal smoking on fetal growth have been based on a single blood or urinary cotinine value, which is inadequate in capturing maternal tobacco exposure over the entire pregnancy. We used hair biomarkers to compare the associations of maternal self-reported smoking, hair nicotine, and hair cotinine with birth weight for gestational age (BW for GA) among active and passive smokers during pregnancy. Methods: We collected maternal hair in the immediate postpartum period and measured nicotine and cotinine concentrations averaged over the pregnancy in 444 term controls drawn from 5,337 participants in a multicenter nested case-control study of preterm birth. BW for GA Z-score and small for gestational age (SGA) were based on Canadian population-based standards. Results: The addition of hair nicotine to multiple linear regression models containing self-reported active smoking, hair cotinine, or both explained significantly more variance in the BW for GA Z-score (p = .01, .03 and .04, respectively). Similarly, women with hair nicotine, but not cotinine, at or above the median value had a significant increase in the risk of SGA birth (odds ratio: 3.07, 95% CI: 1.25-7.52). No significant association was observed between maternal passive smoking and BW for GA based on hair biomarkers. Conclusions: Hair nicotine is a better predictor of reductions in BW for GA than either hair cotinine or self-report. Our negative results for passive smoking suggest that previously reported small but significant effects may be explained by misclassification of active smokers as passive smokers based on self-report.