ASSOCIAÇÃO DE LONGA DEPENDÊNCIA ENTRE MORTALIDADE E SÉRIES CLIMÁTICAS

Abstract

Public health policy strategies for reducing mortality from respiratory and cardiovascular problems are important, particularly, in the context of climate and pollution factors. However, there is a need for increasingly sophisticated statistics for the diagnoses of these relationships, thus reflecting on social well-being. The relationship between climatic factors, pollution and health has been observed in several studies, in the majority of these are used regression models and generalized linear models. This relationship is characterized by the dependence between near and distant observations (in time). The HAR  (Heterogeneous autoregressive model) allows the fit of this dependence, considering series of daily, weekly and monthly averages of the variables under study, which answers the question: "How long does it  take for these variables to lead to death?". In this study, the HAR model, a new methodology in the medical research, was used to analyze the association between mortality (respiratory and cardiovascular problems), climatic series (minimum temperature and humidity) and series of pollutants (PM10 and CO) in  the city of São Paulo, Brazil. There was an association between long-term dependence between climatic  factors and mortality due to respiratory and cardiovascular problems. However, this association was not observed for the pollution data.