Association between triglyceride-glucose related indices and at-risk NASH in U.S. adults with NAFLD: Results from NHANES 2017-2020

Hanchen Ma¹Wenjie Zhang^2,3Zheng Zhao¹Shuaichen Jin¹Linchuan Li^2,3Ruizhao Zong^2,3Wuyang Li^2,3Ziwei Zhang^2,3Shichang Cui^2,3Yunmiao Pan¹Jiankang Zhu^2,3*Guangyong Zhang^1,2,3*

• ¹Department of General Surgery, Shandong Provincial Qianfoshan Hospital, Cheeloo College of Medicine, Shandong University, Jinan, Shandong Province, China
• ²Department of General Surgery, The First Affiliated Hospital of Shandong First Medical University & Shandong Provincial Qianfoshan Hospital, Jinan, Shandong Province, China
• ³Laboratory of Metabolism and Gastrointestinal Tumor, The First Affiliated Hospital of Shandong First Medical University & Shandong Provincial Qianfoshan Hospital, Jinan, Shandong Province, China

IntroductionNonalcoholic fatty liver disease (NAFLD), intricately linked to insulin resistance (IR), obesity, and metabolic syndrome (MetS), affects nearly one-third of the global population. The triglyceride-glucose (TyG) index and its derivative metrics, which integrate anthropometric measures, have emerged as cost-effective tools for assessing IR severity and assessing associations with NAFLD severity. However, evidence on their association with at-risk nonalcoholic steatohepatitis (NASH), a progressive phenotype characterized by hepatic fibrosis and inflammation, remains limited. This study aimed to evaluate the relationship between TyG related indices and at-risk NASH in a nationally representative NAFLD cohort.MethodsUsing data from the 2017-2020 National Health and Nutrition Examination Survey (NHANES), adults with NAFLD diagnosed via vibration-controlled transient elastography (VCTE) were analyzed. At-risk NASH was defined using the FibroScan-AST (FAST) score. Five TyG related indices (TyG, TyG-WC, TyG-WHR, TyG-WtHR, TyG-BMI) were calculated and analyzed through weighted logistic regression, stratified analyses, restricted cubic spline (RCS), and receiver operating characteristic (ROC) analyses. Adjustments included demographics and clinical confounders.ResultsAmong participants, 10% exhibited at-risk NASH. Composite indices (TyG-WC, TyG-WHR, TyG-WtHR, TyG-BMI) each showed higher ORs than TyG alone in Model 3. TyG-WtHR exhibited the strongest association both as a continuous variable and across tertiles in Model 3, demonstrating a near-linear dose-response relationship. RCS analysis revealed no significant nonlinear relationships in composite indices, whereas the TyG index demonstrated a significant nonlinear association. Stratified analyses revealed elevated risks in males, middle-aged adults, and Non-Hispanic Black individuals. ROC analysis showed the discriminative performance of TyG-WC, while TyG-WtHR balanced sensitivity and specificity. ConclusionThis study demonstrated significant associations between TyG-related indices and at-risk NASH in NAFLD patients. These non-invasive, cost-effective biomarkers may facilitate for early risk stratification, guiding precision interventions, and optimizing clinical trial design to improve patient outcomes.