A study of the association between daily mortality and ambient air pollutant concentrations in Pittsburgh, Pennsylvania

We have studied the possible association of daily mortality with ambient po llutant concentrations (PM10, CO, O-3, SO2, NO2, and fine [PM2.5] and coars e PM) and weather variables (temperature and dew point) in the Pittsburgh, PA, area for two age groups-less than 75, and 75 and over-for the 3-year pe riod of 1989-1991. Correlation functions among pollutant concentrations sho w important seasonal dependence, and this fact necessitates the use of seas onal models to better identify the link between ambient pollutant concentra tions and daily mortality. An analysis of the seasonal model results for th e younger-age group reveals significant multicollinearity problems among th e highly correlated concentrations of PM10, CO, and NO2 (and O-3 in spring and summer), and calls into question the rather consistent results of the s ingle- and multi-pollutant non-seasonal models that show a significant posi tive association between PM10 and daily mortality. For the older-age group, dew point consistently shows a significant association with daily mortalit y in all models. Collinearity problems appear in the multi-pollutant season al and non-seasonal models such that a significant, positive PM10 coefficie nt is accompanied by a significant, negative coefficient of another ambient pollutant, and the identity of this other pollutant changes with season. T he PM2.5 data set is half that of PM10. Identical-model runs for both data sets reveal instability in the pollutant coefficients, especially for the y ounger age group. The concern for the instability of the pollutant coeffici ents due to a small signal-to-noise ratio makes it impossible to ascertain credibly the relative associations of the fine- and coarse-particle modes w ith daily mortality. In this connection, we call for caution in the interpr etation of model results for causal inference when the models use fully or partially estimated PM values to fill large data gaps.