Mortality risks from a spectrum of causes associated with wide-ranging exposure to fine particulate matter: A case-crossover study in Beijing, China

Background: Exposure to fine particulate matter (<= 2.5 mu m in aerodynamic diameter; PM2.5) has been shown to be associated with an increased risk of mortality due to cardiovascular, respiratory, and other pulmonary diseases. However, fewer studies have investigated the relationship between ambient PM2.5 and human mortality for a wider range of causes of death, or for more specific causes of death within these broader categories, especially at the high PM2.5 concentrations currently experienced in Chinese megacities. Beijing, China, has a very large population and a wide range of PM2.5 exposures, allowing a prime opportunity to estimate such risks across a broad spectrum of causes, including rarer causes of death. Objective: To estimate the relative risk of cause-specific mortality associated with PM2.5 for a spectrum of causes of death, as well as characterize the time course of cause-specific mortality following PM2.5 exposure, in a location where PM2.5 concentrations are representative of common exposures in Chinese megacities. Methods: We collected daily data on mortality counts of Beijing residents and Beijing weather and air pollution measurements for January 1, 2009 to December 31, 2012. We used a time-stratified case-crossover study design to estimate the association between ambient PM2.5 concentrations and risk of death from several broad causes of death and from more refined specific causes within these broader categories. Primary results were estimated for risks the day of and the day following exposure (lag 0-1), but the time pattern of associated risk was also explored up to seven days following exposure. Results: Increased concentrations of PM2.5 were associated with increased risks at lag days 0-1 of all-cause mortality (0.26% increase per 10 mu g/m(3); 95% confidence interval [CI]: 0.12%-0.39%), non-accidental deaths (0.25%; 95% CI: 0.11%-0.38%), circulatory deaths (0.39%; 95% CI: 0.21%-0.59%), respiratory deaths (0.43%; 95% CI: 0.05%-0.81%), intentional self-harm deaths (1.94%; 95% CI: 0.19%-3.73%) and nervous system deaths (0.9%; 95% CI: -0.2%-2%), although the observed increase was not statistical significant for the final one rarer cause of death. In addition to these five broad death outcomes, risk also increased following PM2.5 exposure at lag days 0-1 for deaths from several specific causes, including most of the specific circulatory causes considered. The largest observed increased risk by far was for one of the rarest causes of death considered, extrapyramidal and movement disorders (2.35%; 95% CI: 0.03%-4.72%). Conclusions: This study indicates that exposure to PM2.5 in a study location more representative of exposures in developing cities is associated with an increased risk of mortality from broad range of causes of death, including some causes rarely studied previously in association with PM2.5 exposure.