Reversal of the seasonality of temperature-attributable mortality from respiratory diseases in Spain

Abstract

A growing number of epidemiological studies have recently assessed temporal variations in vulnerability and/or mortality attributable to hot and cold temperatures. However, the eventual changes in the seasonal distribution of temperature-attributable mortality remain unexplored. Here, we analyse countrywide daily time-series of temperature and mortality counts from respiratory diseases by sex, age group and province of residence during the period 1980-2016 in Spain. We show the complete reversal of the seasonality of temperature-attributable mortality, with a significant shift of the maximum monthly incidence from winter to summer, and the minimum monthly incidence from early and late summer to winter. The reversal in the seasonal distribution of the attributable deaths is not driven by the observed warming in both winter and summer temperatures, but rather by the very large decrease in the risk of death due to cold temperatures and the relatively much smaller reduction due to hot temperatures. We conclude that the projected decrease in the number of moderate and extreme cold days due to climate warming will not contribute to a further reduction of cold-attributable respiratory deaths.

Fig. 1. Relative risk (unitless) of death from respiratory diseases in Spain.

The minimum mortality temperature in 1980–1994 is used as the centring temperature for the two subperiods (i.e. invariant centring temperature assumption, see red vertical lines).

Fig. 2

Minimum mortality temperature (°C) corresponding to respiratory diseases.

Fig. 3. Monthly attributable fraction (unitless) of respiratory disease mortality in Spain.

The minimum mortality temperature in 1980–1994 is used as the centring temperature for the two subperiods (i.e. invariant centring temperature assumption). The attributable fractions (AF) are estimated with the corresponding 95% empirical CI.

Fig. 4. Evolution of the monthly attributable fraction (unitless) of respiratory disease mortality in Spain.

The minimum mortality temperature in 1980–1994 is used as the centring temperature for all the subperiods (i.e. invariant centring temperature assumption).

Fig. 5. Trends in monthly attributable fraction (1 per decade) of respiratory disease mortality in Spain.

The minimum mortality temperature in 1980–1994 is used as the centring temperature for all the subperiods (i.e. invariant centring temperature assumption). Data are presented as mean values ± 95% empirical CI.