O efeito do calor e do frio na mortalidade não-acidental da população residente nas regiões metropolitanas brasileiras ao longo do período de 2000-2014: estudo de série temporal

Abstract

The increase in the relative risk of mortality from various causes is associated with extreme ambient temperatures. Factors at the individual and community level can influence this association and may increase or decrease its effects. These can be variable over time, which could cause this association to fluctuate over the years. This thesis investigated the short-term effects of ambient temperature on mortality in the general population and the elderly in Brazil, as well as the modification of this association over time. The first manuscript verified which effect modifiers modulate the short-term association of temperature and mortality. The second manuscript verified whether there was a change in the association between temperature and mortality over three sequential periods. Non-accidental mortality of the general and elderly population was analyzed, and for these it was subdivided into circulatory, respiratory and other causes. Both studies used a generalized additive model combined with distributed lag non-linear models and a lag of 21 days, to estimate the temperature-mortality association for each metropolitan region, group and period. The studies included a test of heterogeneity between locations and combined estimates for Brazil and geographic regions via meta-analysis. Meta-regression models including factors at the local level were used to analyze possible modifiers of the association. The results of the first manuscript show an effect of cold and heat on non-accidental mortality in the general and elderly population, as well as for cardiovascular, respiratory and other causes of the elderly in Brazil, especially in the metropolitan and geographic regions of the Southeast and South of Brazil. Brazil, with low temperatures being generally the promoters of higher relative risks of death. The geographical factors were responsible for explaining most of the heterogeneity between the locations, with emphasis on the amplitude of the average temperature. The results of the second manuscript show specific estimates of the relative risk of mortality associated with cold and heat differently every five years, with this fluctuation of different patterns according to the type of cause and the Brazilian region. The heterogeneity of the temperature-mortality association between sites increased over the periods for all groups. In the non-accidental and circulatory mortality groups, in the three periods, the temperature amplitude is the variable that best explains this heterogeneity. In the respiratory group, the daily temperature range was a strong factor in the first two periods. The results of this thesis suggest the effect of ambient temperature on mortality in Brazil, with a greater focus on the South and Southeast regions. Thus, action planning with a focus on adaptation to extreme temperatures, both high and low, is necessary and immediately focus on the South and Southeast regions.