Résumé :
|
[BDSP. Notice produite par INIST-CNRS R0xd2c4x. Diffusion soumise à autorisation]. Background The concept of priming of childhood obesity by prenatal exposure to maternal smoking is based on a number of consistent studies. A recent paper found similar associations between paternal smoking and childhood obesity, questioning the presumed causal effect attributed to the prenatal exposure. Is the relation to paternal smoking consistent ? Does it explain the effect of maternal smoking before or in pregnancy ? Methods Data from a cross sectional study on 5899 children in the setting of the 2005 school entrance health examinations in Bavaria were analysed. Associations between paternal smoking or maternal smoking before or in pregnancy and childhood obesity were assessed with adjustment for potential confounders. Results The children's mean age was 5.8 years. The unadjusted odds ratio for obesity and paternal smoking was 2.0 (95% CI : 1.5,2.6) and similar to that for maternal smoking before or in pregnancy with 2.3 (95% CI : 1.8,3.1). After adjustment for a number of potential confounders and paternal smoking at interview the odds ratio for maternal smoking before or in pregnancy and childhood obesity was 1.9 (95% CI : 1.3,2.7). There was no evidence for interaction between paternal smoking and maternal smoking before or in pregnancy (P=0.38). Conclusions Although of similar magnitude, the association of paternal smoking could only partially explain the effect of maternal smoking before or in pregnancy on childhood obesity. Whether this persistent association reflects residual confounding or causality is unclear.
|