Abstract
BACKGROUND: While the association between vitamin D and non-alcoholic fatty liver disease (NAFLD) is well-established in obese populations, its role in lean NAFLD remains controversial. OBJECTIVE: This study aimed to investigate the potential nonlinear dose-response relationship between serum 25-hydroxyvitamin D [25(OH)D] concentrations and lean NAFLD by leveraging large-scale data from the National Health and Nutrition Examination Survey (NHANES). METHODS: We conducted a cross-sectional analysis of 4,201 adults with BMI < 25 kg/m(2). Lean NAFLD was defined using the US Fatty Liver Index (USFLI). The association between serum vitamin D and lean NAFLD was assessed employing multivariable logistic regression models, alongside more advanced techniques including restricted cubic splines (RCS) and generalized additive models (GAM) to capture potential nonlinearity. RESULTS: Initial linear models revealed no significant association. However, subsequent nonlinear analyses uncovered a pronounced 'L-shaped' curve relationship (P-nonlinear < 0.05). A segmented regression approach identified an inflection point at 60.3 nmol/L. Notably, at serum vitamin D levels below this threshold, concentrations were significantly inversely associated with lean NAFLD risk (Odds Ratio [OR] = 0.71, 95% Confidence Interval [CI]: 0.57-0.89). Conversely, above this inflection point, the inverse association attenuated and plateaued. The robustness of this nonlinear association was corroborated by the GAM analysis. CONCLUSION: This large-scale study is the first to identify a significant 'L-shaped' nonlinear association between serum vitamin D and lean NAFLD, wherein the association is most potent at concentrations below approximately 60 nmol/L. This finding highlights an epidemiological inflection point and underscores the need for nonlinear approaches in future research.