High summer temperatures and mortality in the Czech Republic in 1982–2000

Overview

The heat-stress-related mortality, which isamong the main impacts of periods of high summertemperature on society, was reported in many European countries, but analyses focusing oncentral European population have been rare. Results of the analysis for the period of 1982–2000in the Czech Republic indicate that heat stress leads to a considerably increased all-causesmortality and mortality due to cardiovascular diseases. Periods with the highest deviations ofthe daily number of deaths from a baseline (in all-year data) are influenza epidemics and heatwaves; the distribution of days with the highest excess mortality in a year is clearly bimodal, showing a main peak in winter and a secondary one in summer. Summer days with a considerablyincreased mortality are almost entirely days with a positive temperature deviation fromthe seasonal course. Deviations of mortality from the baseline exceed 100 deaths daily (morethan 30% relative increase) in heat wave peaks, and the excess total mortality during the severe1994 heat waves was +456 deaths (+10.3%) for June 17 to 30, and +598 deaths (+12.3%) for July24 to August 8. The relative increase in mortality due to cardiovascular diseases was evenmorepronounced. The mortality displacement effect played an important role, since it was estimatedto account for as much as 52% of the total number of victims for the June 1994 heat wave and48% for the July–August heat wave. People who would die soon without oppressive weatherconditions make about half of the total number of deaths, which is a larger value compared towhat other studies reported. The increased mortality is observed at maximum (average,minimum) daily temperatures higher than 25 °C (18 °C, 14 °C) and their anomalies from meanseasonal courses larger than 3 °C. The same values hold for both the total and cardiovascularmortality. The mortality response at high temperatures is more pronounced in females thanmales. Correlations between mortality and temperature variables (including heat index) arepositive and statistically significant (p=0.01), stronger for deviations of meteorological variablesfrom seasonal courses than for raw values, in females than males, and for heat index than forany temperature variable and summer simmer index. The unlagged correlations are strongerthan correlations with lags 1–3 days; positive values of correlation coefficients hold for lags 0to 3 days only while at lags of 4 to 25 days, the link is negative (mostly statistically significant)which demonstrates the mortality displacement effect and its time extent.

Key words:
mortality – air temperature – heat stress – Czech Republic.