| Abstract
| - Spiers' review (1) highlights the consistency and strength of the evidence supporting the assertion that maternal smoking is independently associated with sudden infant death syndrome (SIDS) risk and that this may account for a large part of the remaining fraction of deaths not attributable to the prone sleeping position. The review emphasizes the difficulty in unraveling prenatal and postnatal maternal smoking effects and confirms that the question raised by our data about the relative causal importance of postnatal smoking is supported by the inconsistency of evidence from those studies where investigators were able to separate prenatal from postnatal effects. A recent case control study from Scandinavia found again that, although maternal smoking showed a dose response relation, paternal smoking had no effect on SIDS if the mother did not smoke (2). While there is a need for considerably more evidence on this issue as the statistical power from the available studies on isolated postnatal environmental tobacco smoke exposure is limited, there is sufficient justification for a focus on how prenatal smoke exposure may cause SIDS. Spiers does this, and we agree with his conclusion that the effect of prenatal smoke exposure is not mediated solely through birth weight. In our study and as previously reported by others including a recent paper (3), the significant effect of maternal smoking on SIDS risk persisted after adjustment for birth weight. In a recent study (4), maternal smoking was found to potentiate the adverse effect of low birth weight. Multiplicative interactions were observed for heavy maternal smoking and other prenatal factors, such as placental insufficiency, leading the authors to suggest that future SIDS infants may have an increased susceptibility to the adverse effect of tobacco smoke in utero (4). The important work by Bulterys et al. (5) in 1990 using the US Collaborative Perinatal Project is relevant here. A possible multiplicative interaction was found between maternal smoking and low hematocrit during pregnancy and SIDS. The interaction remained after adjustment for several confounders, including birth weight, suggesting that the combined effect of prenatal smoking and anemia was not due to birth weight effects. As Spiers suggests, further detailed analyses of maternal smoking and associated factors in the prenatal setting may assist in deciphering how maternal smoking increases the risk of SIDS.
|
