Particulate air pollution and daily mortality: who is at risk?

Major episodes of air pollution, such as the London fog incident in 1952, have shown conclusively that relatively high levels of particulate air pollution can lead to serious adverse health effects, including death. Recent data suggest that exposure to respirable particulates [with an aerodynamic diameter < or = 10 microns (PM10)] under the current American standard (150 micrograms/m3 for 24 h) is associated with increased daily mortality. These results have sparked an intense debate regarding the veracity of the reported associations and whether they should be interpreted causally. Most investigators have relied on a time-series approach to investigate the short-term effects of air pollution on daily mortality. Concerns have been expressed that the results may vary with the type of statistical methods used and the potential for uncontrolled and residual confounding effects. Of perhaps greater importance is that cause-specific mortality, temporal aspects of the exposure-disease relationship, and subgroups of the population have not been studied adequately. Under the hypothesis that only persons with impaired physiologic systems should be at risk, the identification of susceptible subgroups is of great importance in discerning causal mechanisms and in setting public health policies. In this paper, the main methodologic issues are discussed and a novel design to identify at-risk subgroups is presented.