Statistical studies of mortality and air pollution: Multiple regression analyses stratified by age group

Multiple regression analyses are presented relating age-specific mortality rates for up to 158 U.S. cities for 1969–1971 to community air pollution and socioeconomic variables, including an index of cigarette smoking. The results showed that excess mortality was associated with concentrations of airborne manganese and, to a lesser extent, total suspended particulates (TSP). The associations tended to be stronger for the older age groups (>65).In exploring these associations further, it was shown that the regression results were not inconsistent with a lag effect, i.e., that pollution from a 10 year or more earlier time period was implicated. Analysis for pollutant thresholds indicated none for manganese, but the suggestion of a no-effect threshold for TSP at 85–130 μg/m³.In contrast to some previous studies by other authors, no consistent (positive) associations were found between excess mortality and suspended sulfates.