The Role of Air Pollution as a Potential Confounder of the Association between Temperature and Mortality in Cyprus

Purpose: This study is the first to examine the effect of extreme weather on mortality in Cyprus. It aims to investigate the individual effect of meteorological indicators on mortality, as well as the role of air pollution as a potential confounder of heat related health effects. Methods: The analysis concentrated on the warm periods (April to September) of the years 2007-2009. A Generalized Linear Model (GLM) with quasi-Poisson regression included a temperature function and was adjusted for relative humidity and seasonality. The temperature function was developed under a newlydeveloped framework of Distributed Lag Non-Linear Models, to capture non-linearities and lag effects simultaneously. The GLM was extended to examine the confounding effect of air pollution. Results: High temperatures had a significant effect on mortality, independent of humidity and seasonality. Mortality risk increased steeply above a temperature threshold at 33.7oC. A direct heat effect was shown, with high risk on the current and next two days of a severe heat event. PM10 was not found to be significant: the strength of the temperature-mortality relation remained after the inclusion of air pollution in the model. Conclusions: The study indicated that high temperatures result in increased mortality rates in Cyprus, independent of humidity and seasonality, whereas air pollution does not have a confounding effect on the mortality-temperature relation.