Part 3. Estimating the effects of air pollution on mortality in Bangkok, Thailand.

While the effects of particulate matter (PM*) on mortality have been well documented in North America and Western Europe, considerably less is known about its effects in developing countries in Asia. Existing air pollution data in Bangkok, Thailand, indicate that airborne concentrations of PM < or = 10 pm in aerodynamic diameter (PM10) are as high or higher than those experienced in most cities in North America and Western Europe. At the same time, the demographics, activity patterns, and background health status of the population, as well as the chemical composition of PM, are different in Bangkok. It is important, therefore, to determine whether the effects of PM10 on mortality occurring in this large metropolitan area are similar to those in Western cities. The quality and completeness of Bangkok mortality data have been recently enhanced by the completion of a few mortality studies and through input from monitors currently measuring daily PM10 in Bangkok. In this analysis, we examined the effects of PM10 and several gaseous pollutants on daily mortality for the years 1999 through 2003. Our results suggest strong associations between several different mortality outcomes and levels of PM10 and several of the gaseous pollutants, including nitrogen dioxide (NO2), nitric oxide (NO), and ozone (O3). In many cases, the effect estimates were higher than the approximately 6% per 10 microg/m3 typically reported in Western industrialized nations-based on reviews by the U.S. Environmental Protection Agency (U.S. EPA) and the World Health Organization (WHO) (Anderson et al. 2004). For example, the excess risk (ER) for mortality due to all natural causes was 1.3% (95% confidence interval [CI], 0.8 to 1.7), with higher ERs for cardiovascular and respiratory mortality of 1.9% (95% CI, 0.8 to 3.0) and 1.0% (95% CI, -0.4 to 2.4), respectively. Of particular note, for this warm, tropical city of approximately 6 to 10 million people, is that there is no covariation between pollution and cold weather, with its associated adverse health problems. Multiday averages of PM10 generated even higher effect estimates. Our analysis of age- and disease-specific mortality indicated elevated ERs for young children, especially infants with respiratory illnesses, children less than 5 years of age with lower respiratory infections (LRIs), and people with asthma. Age-restricted analyses showed that the associations between mortality due to all natural causes and PM10 concentration increased with age, with the strongest effects among people aged 75 years and older. However, associations between increases in PM10 concentration and mortality were observed for all of the other age groups. With a few exceptions, relatively similar results were observed for several of the other pollutants-sulfur dioxide (SO2), NO2, O3, and NO, which were highly correlated with PM10. However, many of the effects from gaseous pollutants were attenuated in multipollutant models, while effects from PM10 appeared to be most consistent. In addition, there was some evidence of an independent effect of O3 for certain health outcomes. We conducted substantial sensitivity analyses to examine whether our results were robust. The results indicated that our core model was generally robust to the choice of model specification, spline model, degrees of freedom (df) of time-smoothing functions, lags for temperature, adjustment for autocorrelation, adjustment for epidemics, and adjustment for missing values using centered data (see the description of the centering method used in the Common Protocol found at the end of this volume). Finally, the concentration-response functions for most of the pollutants appear to be linear. Thus, our sensitivity analyses results suggest an impact of pollution on mortality in Bangkok that is fairly consistent. They also provide support for the extrapolation of results from health effects studies conducted in North America and Western Europe to other parts of the world, including developing countries in Asia.