Temperature-Related Mortality by Season in Seoul, South Korea, From 1994 to 2005

Abstract

Listen

Translate article

ISEE-491 Objective: Heat-related mortality studies have been largely reported. In particular, the temperature thresholds that mortality risk begins to rise as the temperature increases or decreases were suggested in an USA study. However, these temperatures varied by location. In this study, we evaluated the shape of temperature-mortality associations in 4 seasons of Seoul and analyzed the effects of temperature on mortality, especially among the vulnerable people. Material and Methods: We obtained mortality and meteorological data between 1994 and 2005 from the governmental agencies. After adjusting for the time-trend, day of the week and holidays, relative humidity, and air pressure, we explored various aspects of the association between mortality and temperature by season in Seoul, Korea. First, we fitted nonparametric smoothing regression models with lagged temperature variables to check the associations and then fitted threshold models (including 2 different slopes in a model) to estimate the thresholds and the effects. These thresholds were estimated by the AIC criteria. Results: We evaluated the shape of the temperature-mortality association on the lag structure according to seasons. But, we could see the graphical associations between temperature and mortality only in summer. We confirmed the statistical significance of the effect above the threshold in summer. The threshold values with moving average temperature of current and lagged 1 day were estimated as 27.8°C for 65+, as 28.3°C for all-aged population, and as 29.2°C for 64−. Thresholds for cardiovascular mortality were estimated as 28.3°C, 28.4°C, and 29.3°C for 65+, all aged, and 64−, respectively. Estimated percent increases as 1°C temperature increase above the thresholds were, respectively, 13.3%, 13.0%, and 12.2% in total mortality, 20.1%, 17.4%, and 12.2% in cardiovascular mortality. In respiratory mortality, thresholds and percent increases were estimated as 26.3°C, 6.6% for all aged. Conclusions: According to our findings, we suggest that Korean public health programs to prevent temperature-related mortality should be targeted at the vulnerable groups particularly in summer contrary to other countries.