Tobacco smoke exposure at one month of age and subsequent risk of SIDS--a prospective study.

The aim of this investigation was to identify the sources of postnatal exposure to tobacco smoke at 1 month of age and to examine their relation to sudden infant death syndrome (SIDS). The Tasmanian Infant Health Survey was a prospective cohort study undertaken from 1988 to 1995. It involved 9,826 infants (89% of eligible infants) at higher risk of SIDS. Subsequently 53 eligible infants died of SIDS. Hospital interviews were available on 51 and home interviews on 35 SIDS infants. Urinary cotinine assays were conducted using gas-liquid chromatography (n = 100). Within a predictive model that explained 63% of urinary cotinine variance, the strongest predictor of cotinine and also of SIDS was maternal smoking, though the effects of prenatal and postnatal smoking could not be separated. However, for particular smoking-related behaviors, there was a discordance between prediction of cotinine concentration and prediction of risk of SIDS. If smoking mothers did not smoke in the room with the baby, the cotinine level in the infant's urine was reduced by a little more than a half (p = 0.009), but this was not associated with a reduction in SIDS risk (odds ratio = 1.09, 95% confidence interval 0.47-2.55). Similarly, the presence of other adult resident smokers was associated with a 63% increase in urinary cotinine (p = 0.047) but not with increased SIDS risk (odds ratio = 0.69, 95% confidence interval 0.34-1.40). However, the study lacked the power to detect modest effects, that is, those altering risk less than twofold.