Particulate air pollution and daily mortality in Bangkok

This study was designed to assess the association between PM10 and visibility, and to determine whether the variations in daily mortality were associated with fluctuations in daily PM10 and visibility levels, in Bangkok during 1992--1997. Mortality data were extracted from death certificates, provided by the Bureau of Registration Administration. PM10 data were obtained from three monitoring stations operated by the Pollution Control Department, and visibility data were obtained from two monitoring stations operated by the Department of Meteorology.; PM10 was regressed on visibility using multiple regression. Inverse and significant association was found between PM10 and visibility, after controlling for relative humidity, minimum temperature, and winter indicator variable. Positive association was found between total mortality and PM10, in Poisson regression model while controlling for long-term trends, season, and variations in weather. Five-day moving average of PM10 was significantly and most strongly associated with total mortality from non-external causes; a 2.3% (95% CI = 1.3, 3.3) increase in mortality was estimated for one interquartile range (30 mug/m3) increase in PM10. When PM10 was replaced with visibility, a 1.3% (95% CI = 0.4, 2.3) increase in mortality was estimated for one interquartile range (1.5 km) decrease in visibility. Lagged effects up to three day lags prior to death with similar patterns were observed for both PM10 and visibility.; The findings suggest the possibility of using visibility as a surrogate for fine particulate matter. This approach is feasible because visibility data are usually routinely recorded at airports throughout the world. On the other hand, given the large number of population living in Bangkok, the small but significant percent excess deaths attributable to airborne particle exposure is an important public health concern.