Work

Refining source-specific lung cancer risk assessment from PM2.5-bound PAHs: Integrating component-based potency factors and machine learning in Ningbo, China

Year: 2025

Type: article

Abstract: The component-based potency factor approach, along with benzopyrene (BaP) unit risk values from the World Health Organization (WHO), is commonly used to assess lung excess cancer risk (LECR) from polycyclic aromatic hydrocarbons (PAHs). However, concerns exist regarding potential overestimation of PAH-related LECR using this method. We utilized BaP unit risk values provided by the WHO and the Environmental Protection Agency (EPA) to estimate LECR, considering the underlying assumptions. Integrating PAHs with high carcinogenic potential into the component-based potency factor approach and WHO unit risk factors leads to an overestimation of LECR more than tenfold. Our analysis indicated a moderate PAH exposure risk level (>1.0×10−6) in Ningbo, China. We employed advanced machine learning (ML) algorithms—specifically random forest (RF), extremely randomized trees (ERT), and extreme gradient boosting (XGBoost)—to accurately estimate source-specific LECR associated with PAH exposure. Our findings identified ERT as the most robust ML algorithm, revealing that industrial emissions, coal combustion, and gasoline engine exhaust are the primary contributors to elevated LECR in Ningbo. The results of this study highlight the need for accurate estimation of source-specific LECR to develop effective interventions for the mitigation of PAH pollution and associated lung cancer risk.

Source: Ecotoxicology and Environmental Safety

DOI: 10.1016/j.ecoenv.2025.118174