Comparison of four case-crossover study designs to analyze the association between air pollution exposure and acute myocardial infarction

The case-crossover design is frequently used for analyzing the acute health effects of air pollution. Nevertheless, only a few studies compared different methods for selecting control periods. In this study, the bidirectional method and three time-stratified methods were used to estimate the association between air pollution and acute myocardial infarction (AMI) in Charleroi, Belgium, during 1999-2008. The strongest associations between air pollution and AMI were observed for PM10 and NO2 during the warm period, OR = 1.095 (95% CI: 1.003-1.169) and OR = 1.120 (95% CI: 1.001-1.255), respectively. The results of this study reinforce the evidence of the acute effects of air pollution on AMI, especially during the warm season. This study suggests that the different methods of case-crossover study design are suitable to studying the association between acute events and air pollution. The temperature-stratified design is useful to exclude temperature as a potential confounder.