Abstract
BACKGROUND: Metabolic dysfunction-associated steatotic liver disease (MASLD) is a growing global health concern, and emerging evidence suggests a potential role of vitamin D in its pathogenesis. This study aims to evaluate the impact of serum vitamin D levels on the risk of MASLD using National Health and Nutrition Examination Survey (NHANES) data. METHODS: We analyzed data from 3249 US adults based on NHANES 2021-2023 cycle. MASLD was defined by controlled attenuation parameter value ≥248 dB/m. Three multivariable logistic regression models were constructed: Model 1 (vitamin D alone), Model 2 (partially adjusted), and Model 3 (fully adjusted for demographic, metabolic, and inflammatory factors). Model performance was assessed using receiver operating characteristic analysis, calibration curves, and decision curve analysis (DCA). The mediation analysis was used to explore these associations. RESULTS: The analysis revealed that the MASLD group had significantly lower median vitamin D levels (77.35 vs. 82.10 nmol/L, P < 0.001) and a higher percentage in the lowest vitamin D quartile (27.2% vs. 22.1%, P < 0.001). The multivariable logistic regression demonstrated an inverse association between vitamin D and MASLD prevalence, with each 10 nmol/L increase correlating with a 3.5% reduction in MASLD risk [odds ratio (OR) 0.965, 95% confidence interval (CI) 0.941, 0.989, P = 0.005]. Restricted cubic splines confirmed a linear dose-response relationship (P nonlinear = 0.915). The fully adjusted model 3 showed excellent discrimination (AUC 0.825, 95% CI 0.811, 0.839), indicating excellent predictive performance for MASLD risk assessment. DCA further established that the fully adjusted model provided clinically meaningful net benefits across a wide threshold probability range (15-90%) and prevented 38-42% of unnecessary interventions for MASLD management. Mediation analysis indicated that body mass index, high-density lipoprotein cholesterol, and glycohemoglobin mediated 52.11, 19.12, and 6.95% of the association of vitamin D and MASLD. CONCLUSIONS: Our findings suggest that higher vitamin D levels were associated with lower odds of MASLD, highlighting its potential as a modifiable risk factor and a promising therapeutic target.