Short-term exposure to air pollution and psoriasis outpatient visits in a NW China arid heavy industrial city: a time-series study with DLNM and ARIMAX models

The influence of ambient air pollution on psoriasis is acknowledged; however, data from inland northwestern China is scarce. We correlated 21,458 outpatient consultations for psoriasis from two tertiary medical facilities in Shihezi, China, between 2015 and 2024 with daily air quality and weather data. A distributed lag non-linear model (DLNM) served as the principal analytical framework to evaluate the immediate lagged and cumulative relationships of PM 2.5 , PM 10 , NO 2 , SO 2 , CO, and O₃ with outpatient visits across a 7-day lag interval. Complementary autoregressive integrated moving average model (ARIMAX) were employed as robustness checks to address temporal autocorrelation and significant external interventions. In the initial DLNM evaluations, PM 2.5 , PM 10 , NO 2 , and SO 2 were positively associated with outpatient consultations for psoriasis. The strongest cumulative association was found for NO 2 for lag 0–1 (relative risk 1.082, 95% confidence interval 1.023–1.145). The exposure-response relationships were non-linear, and all associations became non-significant after co-pollutant correction in 2-pollutant models. In the supplementary ARIMAX evaluations, solely the weighted cumulative SO 2 exposure exhibited statistical significance in the main model (relative risk 1.046, 95% confidence interval 1.002–1.092), although this association lacked robustness in various alternative ARIMAX model configurations. These results offer localized evidence that brief exposure to various combustion-related pollutants may elevate psoriasis outpatient visits in Shihezi, while the ARIMAX-derived SO 2 finding should be regarded with caution, as it is model-dependent rather than a strong independent association.