The Concentration-Response Curve Relation Between Air Pollution and Respiratory Hospital Admissions in Children Under 5 Years Old in Taiyuan, Shanxi, China

Abstract

PP-31-016 Background/Aims: To explore the shape of concentration-response relation between air pollution and hospital admissions for respiratory diseases among children. Methods: Electronic medical record abstracts were collected from 22 major hospitals in Taiyuan, and daily air pollution data including particulate matter (PM)10, SO2, NO2, and CO were obtained from local environmental monitoring stations. We built Poisson regression models that included nonparametric adjustments for longer term trends, weather variables, and day-of-week with the logarithm of the expected value of daily hospital admissions as a function of air pollution using natural cubic splines with 2 knots. Results: The strongest association of PM10 with all respiratory diseases was found during nonheating period, while it was in heating period for 3 gaseous pollutants. For PM10 and NO2, their main effects were cumulative effects, whereas the lag effects of SO2 and CO were their main effects. All 4 pollutants were curvedly associated with children's respiratory hospital admissions, with a threshold only for CO lag effect (2120 μg/m3). The effect of NO2 was stronger than PM10, SO2, and CO. For 10 μg/m3 increase in the concentration of NO2 (Avg06, heating period), PM10 (Avg06: 2120 μg/m3, heating period), the percentage increased was 6.04% (95% CI: 2.09%, 10.13%). Air pollutions were more strongly associated with admissions for chronic lower respiratory diseases and acute lower respiratory infections than for acute upper respiratory infections and pneumonia. Conclusion: The study confirmed the acute effect of air pollution on children's respiratory hospital admissions, and suggested the larger excess risks in a high level of air pollution setting.