Evaluation of the CHG index for identifying metabolic dysfunction associated steatotic liver disease: evidence from two independent Asian non-obese populations

Jun ChenQiuyi LiangZeru ChenWang LiuXiaomi ChenHuan LiDongjie HuangShuyue ZhouRiken ChenJunqi Ren^*Xiaoling Wu^*Jinhua Liang^*

• The Second Affiliated Hospital of Guangdong Medical University, Zhanjiang, China

Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) is highly prevalent and is driven by coupled abnormalities in glucose and lipid metabolism. Single metabolic markers such as total cholesterol (TC), fasting blood glucose (FBG), or high-density lipoprotein cholesterol (HDL-C) reflect isolated pathways and often show limited predictive value. The cholesterol, high-density lipoprotein, and glucose (CHG) index integrates lipid and glycaemic information and may better capture MASLD-related metabolic risk. Methods: Two de-identified health-check cohorts from the Dryad repository were analysed: a Chinese derivation cohort from Wenzhou Medical Center (2010–2014; n=16,173) and a Japanese validation cohort from the NAGALA study (2004–2015; n=11,981). CHG was calculated from fasting TC, FBG, and HDL-C. MASLD was diagnosed by abdominal ultrasonography according to national society criteria. Associations were examined with multivariable logistic regression (CHG continuous and quartiles) and restricted cubic splines. Discrimination was compared with TC, FBG, HDL-C, and the triglyceride–glucose (TyG) index using receiver operating characteristic (ROC) curves and area under the curve (AUC). Incremental value beyond baseline models was assessed by net reclassification improvement (NRI) and integrated discrimination improvement (IDI). Results: Higher CHG was consistently associated with MASLD in both populations, showing clear dose–response gradients. Participants in the highest CHG quartile had approximately fourfold higher adjusted odds of MASLD than those in the lowest quartile. CHG yielded the highest AUC among all markers in both cohorts and improved baseline model discrimination and reclassification, with positive NRI and IDI. Conclusion: CHG is robustly and dose-dependently associated with MASLD and demonstrates superior, reproducible discriminative performance compared with traditional single markers across two independent Asian cohorts. CHG may serve as a practical, low-cost tool for MASLD screening and risk stratification at baseline health check-ups.