Differences in the impact of heat waves according to urban and peri-urban factors in Madrid.

Aside from climatic factors, the impact of heat waves on mortality depends on the demographic and socio-economic structure of the population as well as variables relating to local housing. Hence, this study's main aim was to ascertain whether there might be a differential impact of heat waves on daily mortality by area of residence. The study is a time-series analysis (2000-2009) of daily mortality and minimum and maximum daily temperatures (°C) in five geographical areas of the Madrid region. The impact of such waves on heat-related mortality due to natural causes (ICD-10: A00- R99), circulatory causes (ICD-10: I00-I99) and respiratory causes (ICD-10: J00-J99) was obtained by calculating the relative risk (RR) and attributable risk (AR), using GLM models with the Poisson link and controlling for trend, seasonalities and the autoregressive nature of the series. Furthermore, we also evaluated other external variables, such as the percentage of the population aged over 65 years and the percentage of old housing. No heat-related mortality threshold temperature with statistical significance was detected in the northern and eastern areas. While the threshold temperatures in the central and southern areas were very similar and close to the 90th percentile, the threshold in the western area corresponded to the 97th percentile. Attributable mortality proved to be highest in the central area with 85 heat wave-related deaths per annum. External factors found to influence the impact of heat on mortality in Madrid were the size of the population aged over 65 years and the age of residential housing. Demographic structure and the percentage of old housing play a key role in modulating the impact of heat waves. This study concludes that the areas in which heat acts earliest are those having a higher degree of population ageing.