Temperature Variability and Mortality: A Multi-Country Study

Abstract

Background: The evidence and method are limited for the associations between mortality and temperature variability (TV) within or between days.

Objectives: We developed a novel method to calculate TV and investigated TV-mortality associations using a large multicountry data set.

Methods: We collected daily data for temperature and mortality from 372 locations in 12 countries/regions (Australia, Brazil, Canada, China, Japan, Moldova, South Korea, Spain, Taiwan, Thailand, the United Kingdom, and the United States). We calculated TV from the standard deviation of the minimum and maximum temperatures during the exposure days. Two-stage analyses were used to assess the relationship between TV and mortality. In the first stage, a Poisson regression model allowing over-dispersion was used to estimate the community-specific TV-mortality relationship, after controlling for potential confounders. In the second stage, a meta-analysis was used to pool the effect estimates within each country.

Results: There was a significant association between TV and mortality in all countries, even after controlling for the effects of daily mean temperature. In stratified analyses, TV was still significantly associated with mortality in cold, hot, and moderate seasons. Mortality risks related to TV were higher in hot areas than in cold areas when using short TV exposures (0-1 days), whereas TV-related mortality risks were higher in moderate areas than in cold and hot areas when using longer TV exposures (0-7 days).

Conclusions: The results indicate that more attention should be paid to unstable weather conditions in order to protect health. These findings may have implications for developing public health policies to manage health risks of climate change.

Citation: Guo Y, Gasparrini A, Armstrong BG, Tawatsupa B, Tobias A, Lavigne E, Coelho MS, Pan X, Kim H, Hashizume M, Honda Y, Guo YL, Wu CF, Zanobetti A, Schwartz JD, Bell ML, Overcenco A, Punnasiri K, Li S, Tian L, Saldiva P, Williams G, Tong S. 2016. Temperature variability and mortality: a multi-country study. Environ Health Perspect 124:1554-1559; http://dx.doi.org/10.1289/EHP149.

Figure 1

Locations of study areas and their mean values of 0–1 days’ temperature variability (°C). The map is freely downloaded from the “maps” package of R software. TV, Temperature variability

Figure 2

Percent change (95% confidence interval) in mortality associated with an interquantile (for each community) increase in temperature variability (°C) on different exposure days, (A) after controlling for the effect of daily mean temperature, (B) without controlling for the effect of temperature.

Figure 3

Percent change (95% confidence interval) in mortality associated with an interquantile range (for each community) increase in temperature variability (°C) on different exposure days in cold, moderate cold, moderate hot and hot areas, after controlling for the effect of daily mean temperature.