Supervisors info:
Κλέα Κατσουγιάννη, Καθηγήτρια, Ιατρική σχολή, ΕΚΠΑ
Ευαγγελία Σαμόλη, Aν. Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ
Κωνσταντίνα Δημακοπούλου, Μεταδιδάκτωρ, Ιατρική Σχολή, ΕΚΠΑ
Summary:
Introduction: The impact of weather on human health is a matter of increasing concern, especially in light of the observed and predicted climate change. Optimal health protection requires an understanding of the nature of the effect of weather conditions on health at the level of geographic region. The primary purpose was to investigate the relationship between weather and mortality in 10 greek cities: Athens, Thessaloniki, Patra, Volos, Larisa, Heraklion, Ioannina, Kavala, Lamia and Xalkida. Another purposes was to explore heterogeneity among the cities and summarize the results. Data and Methods: The database included daily counts of all-cause mortality, mean temperature and mean relative humidity for each city assembled for 8 years between 2008 and 2016. City-specific Poisson regression models were fitted separately for warm (April–September) and cold (October–March) periods as it is proved that the effect of high temperatures on mortality follow a different pattern in comparison to the effect of the low ones. The analysis was carried out in 2 stages. In the first stage, individual-city data were analyzed and city-specific effect estimates were obtained, which were subsequently used in a second-stage analysis to investigate heterogeneity and provide overall estimates for cold and warm season, respectively. First step was the investigation of the delayed effect of mean temperature on mortality for each city, seperately for cold and warm period. The lag structures for each period and for each city were determined using Distributed Lag Nonlinear Models (DLNM). Second step was to determine the shape of the relationship between mean temperature and mortality for each city for cold and warm period, respectively, using Generalized Additive Models (GAM). A final model was specified for each city, taking into account potential confounders: holidays; day of the week; calendar month; and linear and quadratic terms for time to discern the potentially remaining longterm trend. In the second-stage analysis, fixed- and random- effects meta- analysis was perfomed to combine the city-specific effect estimates of mean temperature on mortality. Results: The lag structure determined by Distributed Lag Nonlinear Models (DLNM) for the cold period was the average exposure of mean temperature for the same and 13 previous days (lag 0-13) while for the warm period was the average exposure of mean temperature for the same and 3 previous days (lag 0-3). The results of the Generalized Additive Models (GAM) comfirmed that the relationship between mean temperature and mortality were V shaped for all the cities with a change-point that varied among cities. This change-point is the value of mean temperature associated with the minimum mortality rate and it is called city-specific threshold. A 1oC decrease in mean temperature (average of lags 0-13) was associated with a 2.2% (95% confidence interval (CI): 1.9, 2.5) increase in the daily number of total natural deaths. The meta- analytic estimate of the threshold was 25.1°C (95% confidence interval (CI): 24, 26.1) for all the cities. A 1oC increase in mean temperature (average of lags 0-3) above the meta-analytic estimate of the threshold was associated with a 5.7% (95% confidence interval (CI): 3.7, 7.6) increase in the daily number of total natural deaths. Conclusions: The findings provide evidence that high and low values of mean temperature are inversely associated with mortality. Cold-related mortality is also an important public health problem across Greece and it should not be underestimated by public health authorities because of the recent focus on heat-wave episodes. The results also confirm that the effects of temperature are more prolonged in cold period compared to the warm period. This study suggests that the nature of the effects of temperature and humidity on mortality vary between cities for unknown reasons which require further investigation but may relate to city-specific population, socioeconomic, and environmental characteristics. This may have consequences on epidemiological studies and local temperature-related warning systems. Concluding, considering the results, prevention programs should be planned by authorities specifically on days with very high or low temperature in order to reduce the burden of heat- and cold-related mortality.
Keywords:
Biostatistics, Time series, Mortality, Temperature, Meteorological factors
