Influence of relatively low level of particulate air pollution on hospitalization for COPD in elderly people

To assess the association between relatively low levels of size-fractioned particulate matter (PM) and hospitalization for chronic obstructive pulmonary disease (COPD), we conducted a time-series analysis among elderly people 65 yr of age or more living in Vancouver between June 1995 and March 1999. Measures of thoracic PM (PM10), fine PM (PM2.5), coarse PM (PM10-2.5), and coefficient of haze (COH) were examined over periods varying from 1 to 7 days prior to hospital admissions. Generalized additive models (GAMS; general linear models, GLMs) were used, and temporal trends and seasonal and subseasonal cycles in COPD hospitalizations were removed by using GLM with parametric natural cubic splines. The relative risks were calculated based on an incremental exposure corresponding to the interquartile range of these measures, and were adjusted for daily weather conditions and gaseous pollutants. PM measures had a positive effect on COPD hospitalization, especially 0 to 2 days prior to the admissions, before copollutants were accounted for. For 3-day average levels of exposure the relative risk estimates were 1.13 (95 % confidence interval: 1.05-1.21) for PM10, 1.08 (1.02-1.15) for PM2.5, 1.09 (1.03-1.16) for PM10-2.5, and 1.05 (1.01-1.09) for COH. The associations were no longer significant when NO2 was included in the models. We concluded that the particle-related measures were significantly associated with COPD hospitalization in the Vancouver area, where the level of air pollution is relatively low, but the effects were not independent of other air pollutants.