Environmental drivers of tuberculosis transmission in Guangdong, China: Integrating generalized additive models and dynamic simulations.

Tuberculosis (TB) remains a critical global public health challenge, particularly in high-burden regions like Guangdong Province, China. This study develops an integrated framework combining generalized additive models (GAM) and non-autonomous dynamical modeling to elucidate the synergistic effects of environmental and socioeconomic factors on TB transmission dynamics. Utilizing weekly TB case data, air quality index (AQI), absolute humidity (AH), and holiday indicators from Guangdong (2014-2019), GAM quantified nonlinear lagged effects of environmental exposures (AQI, AH) and aperiodic drivers (holidays) on incidence. Results revealed that a 10-unit increase in AQI elevated TB risk by 3.8 % (95 % CI: 1.2-6.5 %), while AH exhibited a negative regulatory effect on transmission. Holiday-related population aggregation amplified case fluctuations by 37 % (p < .01), with post-holiday rebounds up to 68 %. These time-varying parameters were incorporated into a non-autonomous SEIR model with recurrence mechanisms. The basic reproduction number R0 was estimated at 1.9 (95 % CI: 1.2-2.6). Bifurcation analysis confirmed global stability of the disease-free equilibrium when R0 < 1 and endemic persistence when R0 > 1. Sensitivity analysis identified infection rate and relapse probability as dominant drivers of transmission intensity. The model predicted a declining long-term trend (-2.6 % annually) but persistent winter-spring seasonality. This hybrid approach providing a quantitative tool for optimizing intervention strategies. Key recommendations include reducing airborne pollutants, enhancing surveillance, and targeting relapse prevention to mitigate endemic persistence.