Air pollution and daily mortality in seven major cities of Korea, 1991-1997.

The relationship between ambient air pollution and daily mortality in seven major cities of Korea for the period 1991-1997 was examined. These cities account for half of the Korean population (about 22 million). The observed concentrations of sulfur dioxide (SO(2), mean=23.3 ppb), ozone (O(3), mean=23.7 ppb), and total suspended particulates (TSP, mean=77.9 microg/m(3)) during the study period were at levels below Korea's current ambient air quality standards. Generalized additive models were applied to allow for the highly flexible fitting of seasonal and long-term time trends in air pollution as well as nonlinear associations with weather variables, such as air temperature and relative humidity. In city-specific analyses, an increase of 50 ppb of SO(2) corresponded to 1-12% more deaths, given constant weather conditions. The risk of all-cause mortality was estimated to increase by 0.5-4%, with an increase in the 2-day moving average of TSP levels equal to 100 microg/m(3). In multipollutant models with pooled data, we found that the estimated risk of death by SO(2) was notably unaffected by adding the other two pollutants (TSP and O(3)) to the model and was statistically significant in various regression models. The rate ratio (RR) for SO(2) remained elevated, indicating an excess mortality of 3% 50 ppb (RR=1.03; 95% CI, 1.01-1.05). TSP's effect on mortality maintained its significance with O(3), but not with SO(2). This implies that there may be collinearity problems where TSP and SO(2) are included in the same model or that TSP may function less than SO(2) as a surrogate for fine particles in the ambient air of Korea. In conclusion, increased mortality was associated with air pollution at SO(2) levels below the current recommendation for air quality. Further research is needed to clarify the relationship between SO(2) and fine particles in Korea.