Tracheal, bronchus, and lung cancer mortality and air pollution exposure in Tuscany, Italy: Bayesian Health Impact Assessment and Global Sensitivity Analysis on a sub-regional scale

Abstract

Outdoor air pollution is a significant risk factor for tracheal, bronchus, and lung (TBL) cancer. This study employs a Bayesian approach to evaluate TBL cancer mortality due to air pollution in Tuscany, Central Italy, in 2023. Using locally validated data, we assessed the impact of fine particulate matter (PM₁₀ and PM_2.5) and nitrogen dioxide (NO${}_2$) in terms of attributable deaths and years of life lost (YLL). Our three-step methodology included: (1) Bayesian modeling to derive posterior distributions for life expectancy, pollution levels, mortality rates, and exposure-response functions (inputs); (2) Monte Carlo simulations to propagate uncertainty from the inputs to the impact metrics (outputs); and (3) Global Sensitivity Analysis (GSA) to quantify the influence of each input on the outputs. The largest impact was estimated for PM_2.5, with 432 deaths (50% CrI: 174;705) and 6,500 YLL (50% CrI: 2,624;10,613) in the region due to annual average concentrations exceeding the WHO threshold of $5\mu g$/m${}^3$. Central districts, with higher exposure levels, were particularly affected, reporting 14 attributable deaths and 207 attributable YLL per 100,000 inhabitants. The GSA indicated that uncertainty in exposure-response functions and annual average concentrations of air pollutants significantly affected outcomes, highlighting the need to strengthen the regional air quality network and conduct local studies to address effects heterogeneity. Our findings highlight the value of high-quality local health assessments for identifying critical areas, setting intervention priorities, and informing context-specific action plans.