Exploratory mediation analysis of body roundness index in the nonlinear association between CHG index and metabolic dysfunction-associated steatotic liver disease among non-diabetic adults.

Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most prevalent chronic liver disease worldwide, highlighting an urgent need for non-invasive early biomarkers. The cholesterol, high-density lipoprotein, and glucose index (CHG) is a novel composite marker reflecting disturbances in glucose and lipid metabolism. However, the independent association between CHG and MASLD, and the potential statistical mediating effect of visceral adiposity, as represented by the body roundness index (BRI), remains unclear, especially in non-diabetic individuals. Methods: This cross-sectional study included 13,682 non-diabetic Japanese adults. Multivariable logistic regression, generalized additive models, and two-piecewise linear regression were used to explore the association, nonlinear relationship, and threshold effect between CHG and MASLD. Receiver operating characteristic (ROC) curve analysis was performed to evaluate CHG's discriminatory performance. Exploratory mediation analysis was performed to statistically decompose the association between CHG and MASLD via BRI. Results: The prevalence of MASLD was 15.25%. After full adjustment, each 1-unit increase in CHG was independently associated with a higher risk of MASLD (OR = 4.46, 95% CI: 3.00-6.64). A significant dose-response trend was observed across CHG quartiles (Q4 vs. Q1: OR = 4.09, 95% CI: 2.83-5.91). A nonlinear threshold effect was identified at CHG = 5.52. CHG showed excellent discriminatory performance (AUC = 0.84), with an optimal cutoff at 5.23 (sensitivity = 0.81, specificity = 0.73), and retained robust discriminatory capacity in non-obese and young individuals. BRI showed a significant statistical mediating effect in the CHG-MASLD association, accounting for 46.15% of the total observed association. Subgroup mediation analyses revealed that this mediating effect was more pronounced in non-obese and normotriglyceridemic individuals. Conclusion: In non-diabetic Japanese adults, CHG is independently and nonlinearly associated with MASLD risk, with BRI as a potential statistical mediator. This mediating effect exhibits heterogeneity and is more pronounced in non-obese individuals. The causal direction of this pathway remains to be verified in prospective studies. CHG shows potential as a low-cost indicator for early MASLD risk stratification, especially in non-obese and younger populations who are often missed by conventional screening.