What question did this study set out to answer?

The aim was to develop and validate a non-invasive model to identify individuals with at-risk metabolic dysfunction-associated steatohepatitis (MASH) in type 2 diabetes.

June 7, 2026

2217-P: DiabetesMASH Index for Noninvasive Identification of At-Risk Metabolic Dysfunction–Associated Steatohepatitis in Type 2 Diabetes

Key Points

The aim was to develop and validate a non-invasive model to identify individuals with at-risk metabolic dysfunction-associated steatohepatitis (MASH) in type 2 diabetes.
A multicenter derivation cohort of 525 individuals undergoing liver biopsy was used for model development.
Validation was conducted using transient elastography-based cohorts from China (3,711), Singapore (422), and the US NHANES (1,674).
At-risk MASH was defined by liver biopsy in the derivation cohort and the FibroScan-aspartate aminotransferase (FAST) score in validation cohorts.
The DiabetesMASH Index (DMI) showed AUROC of 0.79 in the derivation cohort and 0.91, 0.88, and 0.90 in validation cohorts, respectively (95% CI range from 0.75 to 0.94).
Sensitivity was 85% with a negative predictive value of 92% at a rule-out threshold of 0.184, while specificity was 90% with a positive predictive value of 57% at a rule-in threshold of 0.452.
The indeterminate zone was 35.8% in the derivation cohort and ranged from 19.1% to 24.8% in validation cohorts.

Abstract

Introduction and Objective: Metabolic dysfunction-associated steatohepatitis (MASH) is highly prevalent among individuals with type 2 diabetes (T2D) and significantly increases the risk of liver-related events. Individuals with at-risk MASH (MASH+F2-4) are the main target for emerging therapies. We aimed to develop and validate a non-invasive model to identify at-risk MASH in adults with T2D. Methods: A multicenter derivation cohort undergoing liver biopsy (n=525) was used for model development. External validation was conducted in three transient elastography-based cohorts: a Chinese cohort (n=3,711), a Singapore cohort (n=422), and a US NHANES cohort (n=1,674). At-risk MASH was defined histologically in the derivation cohort and by the FibroScan-aspartate aminotransferase (FAST) score in validation cohorts. Performance was compared with the Fibrosis-4 index (FIB-4), non-alcoholic fatty liver disease fibrosis score (NFS), and AST/alanine aminotransferase (ALT) ratio (AAR). Results: The DiabetesMASH Index (DMI), which incorporates age, waist circumference, ALT, and hemoglobin A1c, showed satisfactory discrimination in the derivation cohort (AUROC 0.79; 95% CI 0.75 - 0.83) and validation cohorts (AUROC 0.91 95% CI 0.89 - 0.94, 0.88 0.80 - 0.96, and 0.90 0.86 - 0.94), outperforming the FIB-4, NFS, and AAR. Calibration analyses showed agreement between predicted and observed risks. Rule-out and rule-in thresholds were 0.184 (sensitivity 0.85; negative predictive value NPV 0.92) and 0.452 (specificity 0.90; PPV 0.57). The indeterminate zone accounted for 35.8% of patients in the derivation cohort and 19.9%, 24.8%, and 19.1% in validation cohorts. Conclusion: DMI is a validated, affordable, and accessible non-invasive tool for identifying at-risk MASH in T2D. Disclosure Y. Chen: None. B. Dong: None. C. Liu: None. Z. Li: None. M. Zheng: None. K. Chen: None. G. Goh: None. B. Ye: None. X. Liang: None. A. Hu: None. Y. Li: None. Q. Zeng: None. Y. Sun: None. W. Gou: None. H. Bian: None. J. Lv: None. J. Shen: None. X. Liang: None. X. Zhou: None. M. Huang: None. Q. Zhang: None. Y. Wang: None. X. Xiao: None. Y. Wang: None. Y. Qi: None. Y. Zhou: None. J. He: None. Z. Dai: None. X. Qi: None. Funding The Key Research and Development Program of Jiangsu Province (BE2023767a), the Health Research Program of Anhui (AHWJ2023A30169), and the Natural Science Foundation of Anhui Province (2508085QH314)

Bookmark

Cite This Study

CHEN et al. (Fri,) studied this question.

synapsesocial.com/papers/6a250be87def13d035e1be7f https://doi.org/https://doi.org/10.2337/db26-2217-p

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

Bookmark