Active Fire Detections and Hospital Admissions for Circulatory Diseases in Brazil: An Ecological Panel Study (2013-2024).

Abstract

Background: Wildfire smoke has been associated with cardiovascular outcomes, yet annual counts of fire detections capture a different exposure pattern from the short-term smoke concentrations on which much of that evidence rests. We examined whether fire detection density was associated with hospital admissions for circulatory diseases across Brazil between 2013 and 2024.

Methods: The ecological panel covered 27 federative units over 12 years, giving 324 unit-year observations. Admissions for International Classification of Diseases, 10th Revision, Chapter IX diseases, came from the hospital information system/unified health system; fire detections from the Brazilian National Institute for Space Research; and population and territorial area from the Brazilian Institute of Geography and Statistics. We calculated Spearman correlations and then fitted negative binomial models of admission counts with a population offset and state fixed effects, adding year effects, temperature, poverty, and primary care coverage in successive specifications. Robustness was examined through three temporal subsets and a state-clustered standard-error analysis.

Results: The period comprised 13,801,934 admissions and 2,318,883 fire detections. Ischemic heart disease was the largest individually specified diagnostic group, with 3,301,354 admissions (23.9%). All three correlations lay close to zero: between states ρ = -0.03 (p = 0.87), in the pooled panel ρ = -0.02 (p = 0.70), and after within-state centering ρ = 0.01 (p = 0.89). Across five regression specifications, rate ratios per additional 1,000 detections/100,000 km²/year ranged from 0.979 to 1.020, with the full model at 0.998 (conventional 95% confidence interval (CI), 0.970-1.028; p = 0.913). CIs included unity in every temporal subset and under state-clustered inference as well.

Conclusions: No statistically significant association was detected at the annual state level. These findings do not establish the absence of cardiovascular effects from wildfire smoke; rather, the indirect exposure measure, temporal aggregation, crude outcome rates, and methodological changes in population and healthcare indicators each limit what the data can show.