Impact of Pre- and Postnatal Exposure to Environmental Tobacco Smoke (Ets) on Sudden Infant Death Syndrome (Sids)

Abstract

ISEE-0399 Background and Objective: The aim of this review is to critically evaluate the impact of prenatal and postnatal exposure to environmental tobacco smoke (ETS) on the sudden infant death syndrome (SIDS) separately. The main challenge of this analysis is a possible correlation between maternal smoking during pregnancy and postnatal ETS exposure. Methods: We identified eligible studies by means of a systematic literature search. For prenatal exposure we included studies with information about maternal smoking during pregnancy. For postnatal exposure we restricted our evaluation to studies that quantified the risk for postnatal exposure either by adjusting their analysis for maternal smoking during pregnancy or by conducting stratified analysis in non-smoking mothers. Pooled relative risks were obtained by meta-analysis. Results: From 736 potentially eligible studies, 13 studies on maternal smoking during pregnancy were identified and seven case-control studies on postnatal exposure. The pooled relative risk associated with paternal smoking after birth was RR = 1.62 (95% Confidence Interval: 1.34-1.94). For maternal smoking after pregnancy, the pooled relative risk for SIDS adjusted for smoking during pregnancy was 2.28 (95% CI: 1.41-3.69). The relative risk of maternal smoking during pregnancy on SIDS was 2.49 (95% CI: 1.84-3.38) based on 5 studies with prospective exposure assessment and 2.72 (95% CI: 2.28-3.23) based on eight studies with retrospective exposure assessment. Only two studies on prenatal exposure adjusted their analysis for postnatal ETS exposure: their pooled RR was 3.21 (2.42, 4.26). Conclusion: We found evidence that postnatal ETS exposure is a risk factor for SIDS independent of maternal smoking behaviour during pregnancy. Nevertheless, maternal smoking during pregnancy seems to be a somewhat higher risk factor. Interestingly, separate analyses of prospective and retrospective studies do not indicate the presence of recall bias in these studies.