Estimating mortality effects of fine particulate matter in Japan: A comparison of time-series and case-crossover analyses

Small particles can reach alveoli without being trapped in the upper respiratory tract and may have a greater impact on health than larger particles. Given the limited knowledge on health effects of fine particulate matter (PM_2.5) in Japan, the short-term effects of PM_2.5 on daily mortality using the generalized additive model (GAM), generalized linear model (GLIM), and time-stratified case-crossover analysis were estimated. Daily mortality data were obtained from the Ministry of Health, Labor, and Welfare. Air pollution and meteorological data in 20 areas were obtained from the National Institute for Environmental Studies and the Japan Meteorological Agency, respectively. The three methods were compared, adjusting for meteorological variables and co-pollutants, for areaspecific analyses and combined area-specific results using meta-analysis with a random-effects model. Daily mortality for elderly aged 65 and over varied from 0.5 to 127.3 by area. The 24-hr mean concentration of PM_2.5 ranged from 11.8 to 22.8 μg/m3. Area-specific analyses revealed regional heterogeneity. Furthermore, combined results showed that a 10-μg/m3 increase in PM_2.5 for the singlepollutant model at lag1 was associated with a 0.53, 0.77, and 0.88% increase in all-cause mortality for the GAM, GLIM, and case-crossover analysis, respectively. These findings provide robust evidence for the short-term effects of air pollutants on daily mortality in Japan and suggest that differences in estimates obtained from different statistical models should be considered when multipollutant models are used.