Back to Search Start Over

Ambient temperature and mortality: an international study in four capital cities of East Asia

Authors :
Yasushi Honda
Ho Kim
Xiaochuan Pan
Joo-Youn Chung
Yun-Chul Hong
Yue-Leon Guo
Source :
The Science of the total environment. 408(2)
Publication Year :
2009

Abstract

Extreme ambient temperature has been associated with increased daily mortality across the world. We describe the ambient temperature-mortality association for four capital cities in East Asia, Seoul, Beijing, Tokyo, and Taipei, and identify a threshold temperature for each city and the percent increase in mortality. We adapted generalized linear modeling with natural cubic splines (GLM+NS) to examine the association between daily mean apparent temperature (AT) and total mortality, as well as mortality due to respiratory (RD) and cardiovascular (CVD) causes in a threshold model. We conducted a time-series analysis adjusting for day of the week and long-term time trend. The study period differed by city. The threshold temperature for all seasons was estimated to be 30.1-33.5 degrees C, 31.3-32.3 degrees C, 29.4-30.8 degrees C, and 25.2 degrees -31.5 degrees C for Seoul, Beijing, Tokyo, and Taipei, respectively, on the same day. For the mean daily AT increase of 1 degrees C above the thresholds in Seoul, Tokyo, and Taipei, estimated percentage increases in daily total mortality were 2.7 (95% confidence interval (CI)=2.2-3.1), 1.7 (95% CI=1.5-2.0), and 4.3 (95% CI=2.9-5.7), respectively. Beijing provided no total mortality counts. Estimated percentage increases were 2.7-10.5 for RD mortality, 1.1-9.3 for CVD mortality in 4 cities. This study identified increased mortality due to exposure to elevated AT. The importance of effects of AT and city-specific threshold temperatures suggests that analyses of the impact of climate change should take regional differences into consideration.

Details

ISSN :
18791026
Volume :
408
Issue :
2
Database :
OpenAIRE
Journal :
The Science of the total environment
Accession number :
edsair.doi.dedup.....f10fbca28d2d86975ed301c2da8dabdd