Supplemental Material for: Detecting Early Liver Fibrosis in Obesity: A Head-to-Head Comparison of Five Clinical Scores in Metabolic Dysfunction-Associated Steatotic Liver Disease

We performed a cross-sectional study of 95 individuals undergoing bariatric surgery with intraoperative liver biopsy as reference. FIB-4, NFS, APRI, BARD, and Hepamet were calculated using pre-operative clinical and laboratory data. Fibrosis was staged by NASH CRN criteria.

Diagnostic performance was evaluated using AUROC analysis and diagnostic measures with published and optimized (Youden index) cut-offs. The interaction between MetS and score performance was tested using logistic regression.</p><p dir="ltr">Results: Fibrosis was present in 10.5% of participants. Using published cut-offs, FIB-4 (AUROC 0.752) and Hepamet (AUROC 0.746) showed fair discrimination, while NFS and APRI performed poorly and BARD showed no meaningful discrimination.

Optimized cut-offs improved performance of FIB-4 (AUROC 0.789) and Hepamet (AUROC 0.781), both yielding high negative predictive values (>95%). BARD remained suboptimal. MetS did not significantly modify the association between any score and fibrosis.</p><p dir="ltr">Conclusions: In obese patients at risk of MASLD, FIB-4 and Hepamet provided the most reliable detection of biopsy-proven fibrosis, including early disease.

Population-specific cut-offs improved accuracy, while MetS did not affect score performance. These findings support FIB-4 and Hepamet as practical tools for early fibrosis risk stratification in obese MASLD.</p>