The effect of air temperature and human thermal indices on mortality in Athens, Greece

This paper investigates whether there is any association between the daily mortality for the wider region of Athens, Greece and the thermal conditions, for the 10-year period 1992-2001. The daily mortality datasets were acquired from the Hellenic Statistical Service and the daily meteorological datasets, concerning daily maximum and minimum air temperature, from the Hellinikon/Athens meteorological station, established at the headquarters of the Greek Meteorological Service. Besides, the daily values of the thermal indices Physiologically Equivalent Temperature (PET) and Universal Thermal Climate Index (UTCI) were evaluated in order to interpret the grade of physiological stress. The first step was the application of Pearson's chi (2) test to the compiled contingency tables, resulting in that the probability of independence is zero (p = 0.000); namely, mortality is in close relation to the air temperature and PET/UTCI. Furthermore, the findings extracted by the generalized linear models showed that, statistically significant relationships (p < 0.01) between air temperature, PET, UTCI and mortality exist on the same day. More concretely, on one hand during the cold period (October-March), a 10A degrees C decrease in daily maximum air temperature, minimum air temperature, temperature range, PET and UTCI is related with an increase 13%, 15%, 2%, 7% and 6% of the probability having a death, respectively. On the other hand, during the warm period (April-September), a 10A degrees C increase in daily maximum air temperature, minimum air temperature, temperature range, PET and UTCI is related with an increase 3%, 1%, 10%, 3% and 5% of the probability having a death, respectively. Taking into consideration the time lag effect of the examined parameters on mortality, it was found that significant effects of 3-day lag during the cold period appears against 1-day lag during the warm period. In spite of the general aspect that cold conditions seem to be favourable factors for daily mortality, the air temperature and PET/UTCI exceedances over specific thresholds depending on the distribution reveal that, very hot conditions are risk factors for the daily mortality.