Particulate air pollution and daily mortality: a synthesis

Abstract

Episodes of high concentrations of particulate-based fog were associated with substantial increases in daily mortality earlier in this century. The sharp rise and fall of mortality with equally dramatic changes in air pollution left little doubt about causality. Other studies have shown associations of daily fluctuations in particulate-based air pollution and daily mortality in long time series. These associations continued to quite low concentrations of air pollution, although the optical-based measures do not readily allow direct comparisons with current pollution measurement techniques. This paper reports the results of seven studies examining the relationship between gravimetrically measured airborne particles and daily mortality. In Poisson regressions controlling for weather, time trends, and serial correlation, TSP or PM10 were significantly associated with daily mortality. The association with particles was independent of SO2, but not vice versa. The similarity of the regression coefficients among the studies was striking, and suggests that a 100 micrograms/m3 increase in daily TSP concentrations is associated with about a 6% increase in mortality. The same coefficient also suffices to explain the more than two-fold increase in mortality in London in 1952. The use of covariate adjusted quintile plots showed no evidence of a threshold down to concentrations of 40 micrograms/m3 or less. The relationship was significant in St. Louis, where daily concentrations of PM10 never exceeded 97 micrograms/m3. Given the similar results in all six studies, and the strong associations recently reported between particles and morbidity, the association is likely causal.