Background

Wildfires are increasing in size, frequency, and severity around the world. Wildfire smoke exposure has been linked to numerous adverse health outcomes, but few studies have investigated potential effects on infectious respiratory diseases. We assessed the effect of wildfire smoke exposure on tuberculosis (TB) diagnoses in Brazil, a high-TB-burden country heavily impacted by wildfires.

Methods

We conducted a time-stratified, bidirectional case-crossover analysis using the national TB case registry in Brazil (SINAN) from 2015-2022, anchored on the diagnosis date. Daily concentrations of wildfire-attributable fine particulate matter (PM2.5) were estimated for the municipality of residence using the 3D chemical transport model GEOS-Chem. Primary analyses assessed the count of days of exposure to wildfire-attributable PM2.5 with concentration >35µg/m3 (the EPA 24-hour exposure standard) using a one-year hazard window immediately preceding the date of TB diagnosis and two commensurate-length bidirectional control periods. Secondary analyses assessed 180-day hazard/control windows.

Results

We identified 743,563 individuals with TB from 2015-2022 (Figure). On average, participants experienced 0.54 days with wildfire-attributable PM2.5 >35µg/m3 in the year preceding diagnosis with TB (range:0-95). Using one-year hazard and control periods, each additional day of exposure in the hazard period was associated with a 0.6% increase in the odds of TB diagnosis (OR: 1.006, 95% CI: 1.004, 1.007). Using 180-day windows, each additional day of exposure in the hazard period was associated with a 4% increase in the odds of TB diagnosis (OR: 1.037, 95% CI: 1.035, 1.040).

Conclusion

Among individuals with TB in Brazil, residential exposure to high concentrations of wildfire-attributable PM2.5 was more frequent during risk periods immediately preceding diagnoses compared to more distal referent periods.