Abstract
OBJECTIVE: To explore the association between levels of remnant cholesterol (RC) and the incidence of non-alcoholic fatty liver disease (NAFLD) in the general population, ultimately leading to the assessment of a novel proposed model combining RC and waist circumference (R-W) as a predictor of NAFLD incidence. METHODS: A cross-sectional analysis was conducted of 4913 subjects undergoing physical examinations at Hebei General Hospital, with ultrasound findings being used to classify patients into individuals with and without NAFLD. Median RC values were used to separate subjects into those with low and high levels of RC, after which the predictive performance of RC and the novel R-W model when used to assess NAFLD risk was assessed through multifactorial logistic regression analyses and receiver operating characteristic curves. RESULTS: Subjects exhibiting higher RC levels were found to exhibit an elevated risk of NAFLD incidence with or without adjusting for confounding factors. The binary logistic regression analysis-based R-W model may thus offer utility as a tool to gauge the risk of NAFLD in the general population, as it outperformed RC or WC alone, particularly when assessing women or individuals < 45 years of age. The area under the ROC curve for this combined R-W model was 0.813 (0.802, 0.824), with respective sensitivity and specificity values of 83.86% and 63.91%, all of which were significantly higher than corresponding values for RC and WC (P < 0.001). CONCLUSION: The present supports a strong association between RC levels and NAFLD in the general populace, and the combined R-W model exhibits greater utility as a tool to predict NAFLD incidence relative to RC or WC in isolation, particularly for individuals who are female or <45 years old.