Abstract

The association between ambient coarse particulate matter (PM2.5–10) and mortality in multi-drug resistant tuberculosis (MDR-TB) patients has not yet been studied. The modifying effects of temperature and humidity on this association are completely unknown. To evaluate the effects of long-term PM2.5–10 exposures, and their modifications by temperature and humidity on mortality among MDR-TB patients. A Chinese cohort of 3469 MDR-TB patients was followed up from diagnosis until death, loss to follow-up, or the study's end, averaging 2567 days per patient. PM2.5–10 concentrations were derived from the difference between PM10 and PM2.5. Cox proportional hazard models estimated hazard ratios (HRs) per 3.74 μg/m3 (interquartile range, IQR) exposure to PM2.5–10 and all-cause mortality for the full cohort and individuals at distinct long-term and short-term temperature and humidity levels, adjusting for other air pollutants and potential covariates. Exposure-response relationships were quantified using smoothed splines. Hazard ratios of 1.733 (95 % CI, 1.407, 2.135) and 1.427 (1.114, 1.827) were observed for mortality in association with PM2.5–10 exposures for the full cohort under long-term and short-term exposures to temperature and humidity. Modifying effects by temperature and humidity were heterogenous across sexes, age, treatment history, and surrounding living environment, measured by greenness and nighttime light levels. Nonlinear exposure-response curves suggested a cumulative risk of PM2.5–10-related mortality starting from a low exposure concentration around 15 μg/m3. Long-term exposure to PM2.5–10 poses significant harm among MDR-TB patients, with effects modified by temperature and humidity. Immediate surveillance of PM2.5–10 is crucial to mitigate the progression of MDR-TB severity, particularly due to co-exposures to air pollution and adverse weather conditions.