OBJECTIVES: To analyze the morbidity of patients with respiratory diseases and the effect of exposure to particulate matter (PM2.5) on the occurrence of these diseases in children from one to four years of age and adults aged to 65 years and older in municipalities of the state of Mato Grosso, Brazil, in 2004. METHODS: Ecological study in two stages: The first was an investigation of the global space autocorrelation of the prevalence of hospital admissions due to respiratory diseases and percentage of critical hours of concentration of particulate matter in the municipalities of the state of Mato Grosso, using the Empirical Bayes Smoothing function to minimize random fluctuations of indicators and Global Moran's I. The second, a Multiple Spatial Regression Analysis, had the prevalence of hospital admissions due to respiratory diseases as the outcome variable and the annual percentage of critical hours as independent variable. In order to fit the model, proxy variables of air pollution, population healthcare and living conditions variables were used. RESULTS: No global spatial patterns were observed for hospital admissions due to respiratory diseases in the municipalities; however, a high spatial dependence of the annual percentage of critical hours was detected. In the Multiple Spatial Regression, the models presented significant associations between the annual percentage of critical hours and numbers of hospital admissions due to respiratory diseases. CONCLUSION: We noted that intense production of particulate matter resulting from burning in the 'Legal Amazon' can increase the occurrence of respiratory diseases in vulnerable groups in the municipalities of the state of Mato Grosso, which belongs to the Brazilian Amazon.
Particulate matter; Biomass burning emission; Respiratory diseases; Brazilian Amazon
