Work

The prevalent trajectory of early menopause associated with PM2.5 exposure across 1956–2018 extrapolated from LightGBM algorithm

Year: 2024

Type: article

Abstract: Fine particulate matter (PM 2.5 ) is noxious to female reproductive development and facilitates the occurrence of subsequent diseases. Early menopause is initiative factor of female aging. But due to the lack of historical exposure of PM 2.5 , we could not gain insight into the linkage between ambient PM 2.5 exposure and early menopause. We conducted a community-based retrospective cross-sectional study and pooled 1173 postmenopausal women. The machine learning algorithm of LightGBM was processed to derive the historical concentrations of PM 2.5 based on aerography of 1956–2022. The quantile g-computation and binary logistic regression were employed to estimate the mixed and single associations between PM 2.5 and early menopause. The visibility topped the most important feature for derivations of historical PM 2.5 concentrations. The R 2 of 10-fold cross-validation and predictive capability during processing were all above 0.8. The prevalence of early menopause was 7.3 %. Each 10 µg/m 3 PM 2.5 increased the prevalence of early menopause during prior 2 years exposure (OR: 1.49, 95 %CI: 1.03–2.16) and spring and autumn (OR: 1.28, 95 %CI: 1.07–1.54). After adjusting the reverse effects of temperature, the prior 2 years exposure of PM 2.5 remained positively associated with early menopause in the fourth quantile vs the first quantile (OR: 3.36, 95 %CI: 1.53–7.36) in the spring and autumn. The higher BMI (OR: 1.40, 95 %CI: 1.14–1.72), waistline (OR: 1.42, 95 %CI: 1.09–1.85) and unfavourable dietary habits of less meat (OR: 1.72, 95 %CI: 1.11–2.68), more fried food (OR: 2.39, 95 %CI: 1.15–4.99) elevated the prevalence of early menopause. The accurate environmental exposure assessment of historical PM 2.5 vigorously promoted the researches on the relationship between PM 2.5 and early menopause. It sounds the alarm on female infertility menace associated with particulate matter especially during the turbulent 2 years before menopause. • Algorithms based on visibility have driven the historical PM 2.5 for 50 years. • The independent associations of PM 2.5 with early menopause were remarkable. • Susceptibility of PM 2.5 -early menopause could trace back to 2 years previously. • Body index, social-economics and diet modified PM 2.5 -early menopause associations.

Source: Ecotoxicology and Environmental Safety

DOI: 10.1016/j.ecoenv.2024.117107