Early Identification of Metabolic Syndrome among Young Adults Using Hepatic Biomarkers: Insights from Conventional to Machine Learning Models

Introduction: Metabolic Syndrome (MetS) is progressively prevalent among young adults in India, until now early diagnosis remains a challenge. While liver enzymes are commonly used in clinical screening, composite indices like the Fatty Liver Index (FLI) and Hepatic Steatosis Index (HIS) may offer improved diagnostic accuracy. Methods: A cross-sectional study was conducted on 366 college-going adults (aged 18–25) in India. Liver enzymes (ALT, AST, GGT) and hepatic indices (FLI, HIS) were evaluated for their ability to predict MetS, defined using IDF criteria. Diagnostic accuracy was assessed using ROC analysis, logistic regression, and machine learning models. Results: FLI and HIS were significantly elevated in MetS individuals. FLI showed the highest AUC (0.93), outperforming ALT (0.63) and AST (0.60). Machine learning models classified FLI and HIS as the most important predictors. Predictive performance remained consistent across genders. Conclusion: FLI and HIS are superior to traditional liver enzymes for early, non-invasive detection of MetS through hepatic biomarkers among young adults.