Nonlinear association between body mass index and psoriasis risk: A cross-sectional study based on the NHANES database

Body mass index (BMI) is a widely used anthropometric measure reflecting overall adiposity. Emerging evidence suggests a significant association between obesity and psoriasis pathogenesis. This study aimed to investigate the nonlinear relationship between BMI and psoriasis risk, with particular focus on exploring BMI thresholds. Data were derived from the National Health and Nutrition Examination Survey (NHANES) database, spanning 2 nonconsecutive periods (2003–2006 and 2009–2014). The analytical sample comprised 19,892 participants aged ≥ 20 years who had complete information on BMI, psoriasis status, and all relevant covariates. The study utilized logistic regression to assess the association between psoriasis and BMI. The nonlinear relationship between BMI and psoriasis was examined using a generalized additive model (GAM) for smooth curve fitting. Additionally, a 2-piecewise linear regression model was fitted using a smoothing algorithm to identify potential critical BMI thresholds at which the association with psoriasis risk changed significantly. To explore potential effect modification, we conducted stratified analyses by several factors, including age, sex, and comorbidities. Curve fitting revealed a nonlinear positive association between BMI and psoriasis risk ( P for nonlinearity < .001). A threshold effect was identified at BMI = 26.74 kg/m 2 : compared with BMI < 26.74 kg/m 2 , the odds ratio (OR) for psoriasis was 1.11 (95% CI: 1.05–1.16, P = .0002); for BMI ≥ 26.74 kg/m 2 , the magnitude of risk increase was attenuated (OR = 1.02, 95% CI: 1.00–1.03, P = .0394). Adjustments for covariates did not alter the robustness of these findings. The positive BMI-psoriasis association (per 5 kg/m 2 ) remained consistent throughout subgroup analyses. In summary, we observed a significant association between higher BMI and increased psoriasis prevalence. However, the clinical significance of this association and its practical value in treatment management require further validation.