Abstract

Cause-specific deaths by day for the years 1973 to 1980 in Philadelphia, Pennsylvania, were extracted from National Center for Health Statistics mortality tapes. Death from accidents (International Classification of Disease, Revision 9 ⩾ 800) and deaths outside of the city were excluded. Daily counts of deaths were regressed using Poisson regression on total suspended particulate (TSP) and/or SO2 on the same day and on the preceding day, controlling for year, season, temperature, and humidity. A significant positive association was found between total mortality (mean of 48 deaths/day) and both TSP (second highest daily mean, 222 µg/m3) and SO2 (second highest daily mean, 299 µg/m3). The strongest associations were found with the mean pollution of the current and the preceding days. Total mortality was estimated to increase by 7% (95% CI, 4 to 10%) with each 100-µ/m3 increase in TSP, and 5% (95% CI, 3 to 7%) with each 100-µg/m3 increase in SO2. When both pollutants were considered simultaneously, the SO2 association was no longer significant. Mortality increased monotonically with TSP. The effect of 100 µg/m3 TSP was stronger in subjects older than 65 yr of age (10% increase) compared with those younger than 65 yr of age (3% increase). Cause-specific mortality was also associated with a 100-µg/m3 increase in TSP: chronic obstructive pulmonary disease (ICD9 490-496), +19% (95% CI, 0 to 42%), pneumonia (ICD9 480-486 & 507), +11% (95% CI, −3 to +27%), and cardiovascular disease (ICD9 390-448), +10% (95% CI, 6 to 14%). These results are somewhat higher than previously reported associations, and they add to the body of evidence showing that particulate pollution is associated with increased daily mortality at current levels in the United States.