Associations of Long-Term Exposure to Fine Particulate Matter and Its Constituents with Gynecologic Cancer Incidence: A Prospective Cohort Study from Beijing

Jinting Guo,Lei Yang,Ning Kang,Ning Wang,Xi Zhang,Shuo Liu,Huichao Li,Lili Cao,Min Gao,Tao Xue,Tong Zhu,Jiafu Ji

DOI: https://doi.org/10.1016/j.medp.2024.100058

2024-01-01

Abstract:Background There is limited evidence of the association between long-term exposure to particulate matter with aerodynamic diameter ≤2.5 μm (fine particulate matter, PM2.5) and gynecologic cancer incidence, particularly in developing countries with high air pollution levels. The present study aimed to investigate the associations of long-term exposure to PM2.5 and its constituents with the incidence of gynecologic cancer, and to identify the primary constituent. Methods A total of 85,500 women from the National Urban Cancer Screening Program in Beijing (2013–2019) were included. 5-year average concentrations of PM2.5 mass and its constituents, namely carbon black (BC), ammonium (NH4+), nitrate (NO3−), organic matter (OM), and sulfate (SO42−), were obtained from the Tracking Air Pollution in China (TAP) dataset. Cox proportional hazard models were used to analyze the association between long-term exposure to PM2.5 and its constituents and gynecologic cancer incidence. Mixture exposure models, including quantile-based g-computation (Qgcomp) and explanatory machine learning model (SHapley Additive exPlanations, SHAP), were used to assess the effect of each constituent. Results During a mean follow-up of 6.22 years, 524 participants were diagnosed with gynecologic cancer. The adjusted hazard ratios (HRs) for gynecologic cancer incidence associated with 1 standard deviation (SD) increment was 1.10 (95% CI: 1.00–1.22) for PM2.5, 1.12 (95% CI: 1.01–1.23) for BC,1.11 (95% CI: 1.00–1.23) for NH4+, 1.12 (95% CI: 1.01–1.23) for OM, 1.12 (95% CI: 1.01–1.23) for SO42−, and 1.08 (95% CI: 0.98–1.19) for NO3−. The Qgcomp model indicated the association between a mixture of PM2.5 constituents and gynecologic cancer incidence, with positive weights assigned to BC (61.6%), SO42− (35.9%), and NO3− (2.5%). The SHAP model showed that SO42− was the most important constituent, with the mean |SHAP| value of 0.0322, followed by BC (0.0270), OM (0.0232), NO3− (0.0212) and NH4+ (0.0190). Conclusions Long-term exposure to PM2.5 was associated with gynecologic cancer incidence, with SO42− and BC contributing significantly to the association.