Generalized additive model of air pollution to daily mortality

The association of air pollution with daily mortality due to cardiovascular disease, respiratory disease, and old age (65 or older) in Seoul, Korea was investigated in 1999 using daily values of TSP, PM10, O3, SO2, NO2, and CO. Generalized additive Poisson models were applied to allow for the highly flexible fitting of daily trends in air pollution as well as nonlinear association with meteorological variables such as temperature, humidity, and wind speed. To estimate the effect of air pollution and weather on mortality, LOESS smoothing was used in generalized additive models. The findings suggest that air pollution levels affect significantly the daily mortality. (orig.)