Effects of the modified Physiologically Equivalent Temperature on mortality and morbidity in the Mediterranean, a case study for Cyprus

Human health faces a significant threat from climate change, especially in regions like the Mediterranean, and the Middle East and North Africa, which have experienced extreme temperatures in recent decades. Extreme heatwaves are becoming common in these areas, impacting enclosed populations more severely. However, hot weather is experienced in different ways by different demographics. This is an aspect that must be accounted for in any analysis aiming at studying the human health related effects of heat exposure. For this, here, were use the modified Physiologically Equivalent Temperature (mPET) – an advanced human-biometeorological index, which is capable of  considering the variability of thermo-physiological responses among different population groups. mPET in particular combines temperature, humidity, wind and radiation loads with anthropometric (e.g. age), activity and clothing factors to assess human thermal comfort for a wide variety of sub-populations. Focusing on Cyprus, a Mediterranean island, mPET was computed in the present study based on population-weighted climate data derived from Copernicus European Regional Reanalysis (CERRA), at 5.5 km spatial resolution and at 1 h temporal resolution, using the RayMan Pro Model. The mPET estimates were specifically provided for the five districts of Cyprus and for six different population groups: female and male children, adults, and seniors. Regarding epidemiological data, we considered daily mortality counts and daily morbidity records associated with cardiovascular and respiratory diseases (ICD10 codes: I00-I99 and J00-J99), spanning nearly two decades (2004-2019). To understand how the thermo-physiologically relevant heat stress affects mortality and morbidity rates, we applied Distributed Lag non-Linear models (DLNMs) within the general framework of Generalized Additive Models (GAMs). This modelling framework produces estimates of the temporally distributed effect of mPET on mortality/morbidity rates. We demonstrate how the health risk varies across the various population sub-groups and districts in Cyprus and that using mPET rather than single heat-related meteorological metrics enables such differences to be better highlighted. This is the first time that health risks arising from heat were assessed for various demographics in Cyprus, providing crucial information for better understanding the climate change impact in the country.