Air pollution and coronary heart disease–related hospital visits in Beijing, China: time-series analysis using a generalized additive model
Yuan Gao,Weixuan Sheng,Yongtao Yang
DOI: https://doi.org/10.1007/s11356-022-24803-x
IF: 5.8
2022-12-24
Environmental Science and Pollution Research
Abstract:To investigate correlations between environmental and meteorological factors and frequency of presentation for coronary heart disease (CHD) in Beijing. Daily measurements of levels of six atmospheric pollutants were made, data relating to meteorological conditions collected, and CHD-related outpatient visits recorded from January 2015 to December 2019 in Beijing. A time-series analysis was made, using a generalized additive model with Poisson distribution, and R 3.6.3 software was used to estimate relationships among levels of atmospheric pollutants, ambient temperature, and visits occasioned by CHD. Results were controlled for time-dependent trend, other weather variables, day of the week, and holiday effects. Lag-response curves were plotted for specific and incremental cumulative effects of relative risk (RR). The aim was to correlate meteorological-environmental factors and the daily number of CHD-related hospital visits and to quantify the degree of correlation to identify any pathological associations. Response diagrams and three-dimensional diagrams of predicted exposure lag effects were constructed in order to evaluate relationships among the parameters of air pollution, temperature, and daily CHD visits. The fitted model was employed to predict the lag RR and 95% confidence interval (95% CI) for specific and incremental cumulative effects of random air pollutants at random concentrations. This model may then be used to predict effects on the outcome variable at any concentration of any defined pollutant, giving flexibility for public health purposes. The overall lag-response RR curves for the specific cumulative effects of the pollutants, particulate matter (PM)2.5, PM10, SO 2 , CO, and NO 2 , were statistically significant and for PM2.5, PM10, CO, and NO 2 , the overall lag-response RR curves for the incremental cumulative effect were statistically significant. When PM2.5, PM10, SO 2 , CO, and NO 2 concentrations were above threshold values and the temperature was below 45 °F (reference value 70 °F), the number of CHD-related hospital visits increased with a time lag effect. The outpatient volume of CHD was predicted by the model to guide the flexible distribution of medical resources. Elevated PM2.5, PM10, SO 2 , CO, and NO 2 concentrations in the atmosphere combined and low ambient temperature increased the risk of CHD with a time lag effect.
environmental sciences