Associations of temperature, particulate matter, and greenness with human papillomavirus positivity at screening: an analysis of Guangdong Province (2016-2022) using the treed distributed lag nonlinear model.

INTRODUCTION: Human papillomavirus (HPV) is a double-stranded DNA virus that may cause cervical, vulvar, and anal cancers in women. Environmental factors including temperature, air pollution, and greenness may be associated with HPV positivity; however, evidence remains limited. METHODS: This study analysed health examination data of 1,287,412 women in Guangdong Province, China (2016-2022). Ten covariates were collected, and temperature, particulate matter 2.5 (PM2.5), and normalized difference vegetation index (NDVI) were assigned by examination date. Tree distributed lag non-linear models were applied to evaluate lagged associations with HPV positivity at screening. RESULTS: Using the Q50 of biweekly mean temperature, PM2.5, and NDVI (25.594°C, 20.300 μg/m³, and 0.544, respectively) as references, low temperature (Q2.5, 13.300°C) was associated with higher cumulative odds of HPV positivity at Lag Weeks 3-8 (cOR = 1.303, 95% CrI, 1.194, 1.428), whereas high temperature (Q97.5, 29.851°C) showed a longer-lag association at Lag Weeks 9-14 (cOR = 1.063, 95% CrI, 1.007, 1.122). High PM2.5 (Q97.5, 38.021μg/m³) and low NDVI (Q2.5, 0.229) were associated with higher odds of HPV positivity over Lag Weeks 1-16 (cOR = 1.172, 95% CrI, 1.051, 1.283) and Lag Weeks 5-18 (cOR = 1.318, 95% CrI, 1.131, 1.566), respectively. Stratified analyses suggested stronger low-temperature associations among middle-aged and older women and among women in the Pearl River Delta and Western Guangdong. DISCUSSION: In conclusion, temperature showed lag-dependent associations with screening-detected HPV positivity, with stronger associations for low temperature at shorter lags and high temperature at longer lags. High PM2.5 and low NDVI were also associated with higher odds in the primary 24-week analyses. Temporal sensitivity analyses attenuated the PM2.5 association and weakened the NDVI association under the 36-week specification, whereas the low-temperature association was comparatively stable and uncertainty increased for high temperature. These findings identify pre-screening environmental correlates of HPV positivity and warrant confirmation in longitudinal studies with repeated HPV testing and more comprehensive control of temporal, spatial, behavioral, and socioeconomic factors.