The Impact of Major Heat Waves on the Global and Cause Specific Mortality in France from 1971 to 2003

Abstract

TM4-O-03 Introduction: Few studies have systematically described the medical causes of death occurring during heat waves. The aim of this study is to identify the major heat waves that have occurred from 1971 to 2003 in France and describe their impact on global and cause-specific mortality. Material and Methods: The underlying causes of death were regrouped in 18 categories, isolating heatstroke, hyperthermia, and dehydration in a “heat-related causes” (HRC) category. The deaths certified as HRC, on underlying or contributing causes, were also identified. Using a Poisson log-linear model, the number of deaths observed (O) during identified heat episodes were compared with those expected (E) on the basis of the mortality rates observed on the last 3 preceding years. Two main indicators were used: the mortality ratio (O/E) and the excess mortality (O − E). Heat waves were identified in 2 stages: first, their temporal bounds were fixed on meteorologic criteria and then, on the selected periods, their geographic extents were identified on mortality criteria. Results: Five major heat waves were identified in France with a close chronologic concordance between the daily temperature and mortality evolutions during the heat waves. Mortality ratios were increasing with age for subject over 55, and they were from 4% to 24% higher for women over the age of 65. Considering the underlying causes of death, among the 5 heat waves: Excess mortality were significant for almost all causes; Mortality ratios were the highest for HRC, respiratory diseases, nervous system diseases, mental disorders, infectious diseases, and endocrine and nutritional diseases; Excess deaths were the highest for cardiovascular diseases, respiratory system diseases, neoplasms, HRC, external, and ill-defined conditions. From 1975 to 2003, the proportion of nervous system diseases and mental disorders in the global excess death during heat waves has been increasing. During the 2003 heat wave, HRCs were more marked (as underlying or contributing cause) while neoplasms and cardiovascular diseases were in declining proportions. Discussion and Conclusions: Heat waves associated with a mortality excess are not very rare events in France. The mortality ratios have common characteristics for the 5 heat waves. The cause of death structure of the excess mortality is mainly following that of the usual mortality. This is a step to the identification of medical vulnerability factors.