A critical review of PM10-mortality time-series studies.

While the mortality effects of particulate matter (PM) have been obvious during extreme historical pollution episodes (e.g., the London Fog of 1952), evaluating effects at more routine pollution levels has required the use of complex statistical modeling approaches. This paper critically reviews available time-series studies on PM10 mortality to provide a common basis for an evaluation of the PM10-mortality association. These PM10 studies confirm that an acute pollution-mortality association can occur at routine ambient levels, and suggest that such effects extend below the present United States air quality standards, especially for susceptible subpopulations. Furthermore, these new PM10 studies are consistent with the hypothesis noted in past studies that PM is a causal agent in the mortality impacts of air pollution. The relative risks (RRs) for PM10 mortality, however, were found to vary across studies. Variation probably was caused by differences in PM10 composition and in the PM10 averaging period employed in the analysis, as well as differences in whether other pollutants were considered simultaneously in the mortality-PM10 model. Overall, the RR estimates derived from available PM10-total mortality studies suggest a 24-h average, 100 micrograms/m3 PM10 acute exposure effect on the order of RR approximately 1.05-1.10 in the general population. Higher PM10 RRs were indicated for the elderly and for those with preexisting respiratory conditions, both of which represent subpopulations who appear to be especially at risk for the mortality implications of acute exposures to air pollution. A key research question remaining involves a determination of the component or components of PM10 (e.g., fine particles, sulfates, acid aerosols, or ultrafine particles) that are most important to the noted acute PM-mortality associations.