Nomogram integrating serological markers and clinical parameters for predicting severe hepatic steatosis in patients with abnormal glucose metabolism

Abstract Background Individuals with abnormal glucose metabolism are at high risk of developing metabolic dysfunction-associated steatotic liver disease (MASLD). Severe hepatic steatosis represents a critical transition stage from simple fat accumulation to liver fibrosis and significantly increases the risk of adverse liver outcomes. This study aimed to construct a nomogram model integrating serological markers and clinical parameters for predicting the risk of severe hepatic steatosis in this population. Methods This prospective study enrolled 186 patients with abnormal glucose metabolism (including prediabetes and diabetes mellitus) who underwent FibroScan examination and serological testing between February 2023 and May 2024. According to the controlled attenuation parameter (CAP) cutoff values recommended by the manufacturer, patients were classified into severe (n = 56) and non-severe (n = 130) hepatic steatosis groups and randomly divided into training and validation cohorts at a 7:3 ratio. Least absolute shrinkage and selection operator (LASSO) regression was applied to select potential predictive variables, followed by multivariate logistic regression to identify independent predictors. Restricted cubic spline (RCS) analysis was used to assess potential nonlinear associations between independent predictors and severe hepatic steatosis, and Spearman correlation analysis was performed to evaluate their correlations with CAP values.A nomogram prediction model was constructed based on the identified independent predictors. Predictive performance was evaluated, and internal validation using 1,000 bootstrap resamples was performed to assess model stability. Results Multivariate logistic regression identified body mass index (BMI), triglycerides (TG), adiponectin (ADPN), and chemerin as independent predictors of severe hepatic steatosis. RCS analysis revealed approximately linear associations of BMI, ADPN, and chemerin with the risk of severe hepatic steatosis, whereas TG demonstrated a nonlinear relationship. Spearman correlation analysis showed that the CAP was positively correlated with BMI, TG, and chemerin and negatively correlated with ADPN. A nomogram incorporating these four variables was established. The area under the receiver operating characteristic curve (AUC) was 0.862 in the training cohort and 0.889 in the validation cohort, indicating good discriminative ability. Calibration and decision curve analysis (DCA) curves confirmed favorable calibration and clinical utility. Internal validation with 1,000 bootstrap resamples further demonstrated the stability of the nomogram. Conclusion The nomogram based on BMI, TG, ADPN, and chemerin can predict severe hepatic steatosis in patients with abnormal glucose metabolism to a certain extent, providing a novel and practical tool for individualized clinical decision-making.