Street-level ambient fine particulate matter and tuberculosis: A machine learning and Bayesian spatiotemporal...

Area of Science:

• Environmental epidemiology
• Public health
• Spatial analysis

Background:

• Tuberculosis (TB) remains a significant global infectious disease threat.
• Limited evidence exists on the link between short-term fine particulate matter (PM2.5) exposure and TB incidence at a granular spatial level.

Purpose of the Study:

• To investigate the street-level association between short-term PM2.5 exposure and active TB onset in Jinan, China.
• To identify high-risk subgroups and seasonal variations in PM2.5's impact on TB incidence.

Main Methods:

• Utilized daily street-level PM2.5 concentrations estimated via an extreme gradient boosting (XGBoost) model and GIS.
• Employed a Bayesian hierarchical spatiotemporal model to quantify exposure-response relationships for 8,291 TB cases (2015-2019).
• Conducted stratified analyses by age, sex, season, and residential location.

Main Results:

• A 10 μg/m³ increase in PM2.5 at a 3-day lag was associated with a 2% rise in TB incidence (IRR=1.02, 95% CI: 1.01-1.03).
• Significant effects persisted up to a 6-day lag (IRR=1.01, 95% CI: 1.00-1.02).
• Elevated risks were observed in the elderly (≥65 years), females, rural residents, and during cold seasons (November-April).

Conclusions:

• Short-term PM2.5 exposure is a significant risk factor for TB incidence.
• Public health interventions should prioritize protecting high-risk subgroups and implementing pollution controls during cold seasons.
• Findings highlight the need for targeted strategies to mitigate the TB burden associated with air pollution.