A multicity study of association between air pollution and CHD mortality in China by using time series threshold poisson regression model

There are few multicity studies to address the effect of short-term effect of particulate matter air pollution on daily Coronary Heart Disease (CHD) mortality in developing countries, much fewer to further discuss its threshold and seasonal effect. This study investigates the season-varying association between particulate matter less than or equal to 10 ?m in aerodynamic diameter (PM10) and daily CHD mortality in seven cities of China. Time series threshold Poisson regression model is specified to estimate the health effect for four cities with the threshold effect, and conventional linear Poisson model is used to analyze the effect for three cities without threshold. We apply the Bayesian hierarchical model to pool the city-specific estimates into overall level. On average, a 10?g/m3 increase of the moving average concentrations of current-day and previous-day PM10 is associated with an increase of 0.81% (95% Posterior Interval, PI: -0.04%, 1.67%) in daily CHD mortality for all the cities as a whole. The associations are smaller than reported in developed countries or regions with lower polluted level, which is consistent to the findings in the literature. The hazardous effect are higher in hot summer and cold winter (1.15% and 0.89%) but lower in relative warm spring and fall (0.85% and 0.69%). In summary, we found significant associations between short-term exposure to PM10 and CHD mortality in China. The sensitivity analyses in the study support the robustness of our results.