Related Experiment Video
Updated: Sep 17, 2025
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
TyG-BMI as a superior predictor of MAFLD and pre-MAFLD in Chinese adults: a cross-sectional study
Xinru Huang1, Yanyun Hu1, Lining Dong1
1Department of Endocrinology and Metabolism, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, 200080, China.
Background:
Metabolic-associated fatty liver disease (MAFLD), a highly prevalent global liver disorder, requires early diagnosis and treatment to delay progression, yet lacks simple diagnostic tools. We aimed to evaluate the validity of the triglyceride-glucose index (TyG) and its related indexes in predicting MAFLD and pre-MAFLD among Chinese individuals.
Methods:
This cross-sectional study enrolled 9,627 Chinese participants (mean age 46.53 ± 11.87 years; 62.1% males) categorized into three groups based on MAFLD diagnostic criteria. Baseline characteristics were analyzed using one-way ANOVA and Kruskal-Wallis tests. To evaluate independent associations, we performed logistic regression analyses of anthropometric markers and TyG-related indexes with MAFLD status. Furthermore, the dose-response relationship between the TyG and related indexes and MAFLD risk was examined using restricted cubic splines (RCS). Finally, receiver operating characteristic (ROC) curve analysis quantified the predictive performance of each index for MAFLD and pre-MAFLD.
Results:
Mean blood pressure, anthropometric indicators, and median TyG index and related parameters were significantly higher in the MAFLD group compared to the non-MAFLD group (p <.001). The pre-MAFLD group exhibited intermediate metabolic values between the non-MAFLD and MAFLD groups. Logistic regression analyses revealed anthropometric indicators, TyG index and related indicators were independent risk factors for MAFLD, adjusted for age, gender, blood pressure, blood glucose, and liver function markers. RCS analyses showed a nonlinear positive association between TyG index and related indexes and MAFLD risk (p for nonlinear <.0001). TyG-BMI demonstrated the highest area under the curve (AUC) for both MAFLD (0.874, 95%CI 0.867-0.881) and pre-MAFLD (0.823, 95%CI 0.810-0.836), with optimal cutoffs of 217.41 and 202.54, respectively.
Conclusions:
TyG-related indices, particularly TyG-BMI, showed superior discriminative ability for MAFLD and pre-MAFLD compared to traditional anthropometric measures in this Yangtze River Delta population, demonstrating an S-shaped dose-response relationship with MAFLD risk. These findings support TyG-BMI as a potential first-line screening tool.
