Environmental tobacco smoke and ischaemic heart disease: a case study in applying causal criteria.

ABSTRACT

BACKGROUND: Whether ischaemic heart disease (IHD) is caused by exposure to environmental tobacco smoke (ETS), commonly known as "passive smoking", has been debated from both epidemiological and biological perspectives. METHODS AND RESULTS: In this paper we use Bradford Hill criteria to synthesize results from the biological and epidemiological literature in a formal assessment of the strength of support for such a relationship. Although we find that these criteria, designed for clinical trials, do not give an ideal framework for assessment of epidemiological and biological studies, nevertheless they do provide systematic guidance for this assessment. For the general population, of the nine tests proposed by Hill we find that one (biological plausibility) seems to be supported, though not unarguably; three (strength, consistency. specificity) appear to fail by accepted standards; and the remaining five have insufficient data for a clear evaluation (biological gradient, experimental evidence, temporality, coherence, analogy). Overall, this provides at best weak support for a causal association between ETS and IHD across the general community. Conversely, there appears to be more support, especially in the biology studies, for an association between ETS and IHD for those with preexisting disease, although epidemiological studies are limited in this area. CONCLUSIONS: One of the outcomes of this review is the identification of areas of focus for future epidemiological and biological research. First, we find that stronger associations may be found in the particular subpopulation with pre-existing IHD. In this case, more convincing biological plausibility and experimental evidence indicate a need for relevant epidemiological studies, although individual responses are very variable. Second, we identify the need for further, more detailed evaluations of the nature of vessel wall thickenings occurring in experimental models of ETS exposure. Third, we propose long-term animal studies of initiation of IHD, including direct assessment of effects on the accumulation of lipid in vessel walls, at appropriate ETS exposure levels.