The objectives of this article were (i) to find the association between extreme temperatures and respiratory emergency department (ED) visits and (ii) to explore the added effects of heat waves and cold spells on respiratory ED visits in Beijing from 2009 to 2012. A quasi-Poisson generalised linear model combined with a distributed lag non-linear model was performed to quantify this association. The results indicated that (i) ambient temperature related to respiratory ED visits exhibited a U-shaped association. The minimum-morbidity temperature was 21.5°C. (ii) the peak relative risk (RR) of cold spells on respiratory ED visits was observed in relatively mild cold spells with a threshold below the 3rd percentile for 4days (RR=1.885, 95% CI: 1.300-2.734), and there was a reduction in risk during extremely chilly cold spells (RR=1.811, 95% CI: 1.229-2.667). However, the risk of heat waves increased with the thresholds, and the greatest risk was found for extremely hot heat waves (RR=1.932, 95% CI: 1.461-2.554). (iii) the added effect of heat waves was small, and we observed that the added heat wave effect only introduced additional risk in females (RR=1.166, 95% CI: 1.007-1.349). No added effect of cold spells was identified. In conclusion, the main effects of heat waves and cold spells on respiratory ED visits showed different change trends. In addition, the added effects of extreme temperatures on respiratory ED visits were small and negligible.
Keywords: Extreme temperatures; Main and added effects; Respiratory morbidity.
Copyright © 2017 Elsevier B.V. All rights reserved.