Life-course epidemiology Parental smoking and childhood obesity: higher effect estimates for maternal smoking in pregnancy compared with paternal smoking—a meta-analysis

Background: Some studies reported similar effect estimates for the impact of maternal smoking in pregnancy and paternal smoking on childhood obesity, whereas others suggested higher effects for maternal smoking. We performed a meta-analysis to compare the effect of in utero exposure to maternal smoking and that of paternal or household smoking exposure in utero or after birth with mutual adjustment. Methods: Meta-analysis of observational studies identified in MEDLINE, EMBASE and Web of Knowledge published in 1900–2013. Study inclusion criterion was assessment of the association of maternal smoking during pregnancy and paternal or household smoking (anyone living in the household who smokes) at any time with childhood overweight and obesity. The analyses were based on all studies with mutually adjusted effect estimates for maternal and paternal/household smoking applying a random-effects model. Results: Data for 109 838 mother/child pairs were reported in 12 studies. The pooled odds ratios (ORs) for overweight 1.33 [95% confidence interval (CI) 1.23;1.44] (n1⁄4 6, I1⁄4 0.00%) and obesity 1.60 (95% CI 1.37;1.88) (n1⁄4 4, I1⁄4 32.47%) for maternal smoking during pregnancy were higher than for paternal smoking: 1.07 (95% CI 1.00;1.16) (n1⁄4 6, I1⁄4 41.34%) and 1.23 (95% CI 1.10;1.38) (n1⁄44, I1⁄4 14.61%), respectively. Similar estimates with widely overlapping confidence limits were found for maternal smoking during pregnancy and childhood overweight and obesity: 1.35 (95% CI 1.20;1.51) (n1⁄4 3, VC The Author 2014; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association 1593 International Journal of Epidemiology, 2014, 1593–1606 doi: 10.1093/ije/dyu150 Advance Access Publication Date: 29 July 2014 Original article I1⁄4 0.00%) and 1.28 (95% CI 1.07;1.54) (n1⁄4 3, I1⁄40.00%) compared with household smoking 1.22 (95% CI 1.06;1.39) (n1⁄4 3, I1⁄472.14%) and 1.31 (95% CI 1.15;1.50)] (n1⁄4 3, I1⁄4 0.00%). Conclusions: Higher effect estimates for maternal smoking in pregnancy compared with paternal smoking in mutually adjusted models may suggest a direct intrauterine effect.