Tryptophan derivatives as non‐invasive diagnostic indicators for obesity‐related MASLD in children and adolescents

Abstract Aims Metabolic dysfunction‐associated steatotic liver disease (MASLD) represents a major health burden in children and adolescents with obesity. This study aimed to characterize metabolic alterations associated with obesity‐related MASLD and to evaluate the potential of tryptophan (TRP)‐derived metabolites as non‐invasive biomarkers for early diagnosis. Materials and Methods A total of 30 normal‐weight individuals and 80 patients diagnosed with obesity were included in the discovery cohort. Circulating metabolites were quantified and compared among normal‐weight individuals, patients with obesity, and those with obesity‐related MASLD to identify metabolic changes associated with disease status. In vitro experiments using mouse hepatocytes and liver organoids were conducted to assess the effects of TRP‐derived metabolites on intracellular lipid accumulation. Finally, a non‐invasive diagnostic model was developed using machine learning techniques and validated in an independent cohort of 112 individuals. Results Circulating TRP metabolism was markedly elevated in children and adolescents with obesity‐related MASLD compared with both normal‐weight controls and individuals with simple obesity. TRP‐derived metabolites significantly promoted intracellular lipid accumulation in mouse hepatocytes and liver organoids via the induction of oxidative stress. The non‐invasive diagnostic model based on TRP‐derived metabolites demonstrated robust performance in differentiating obesity‐related MASLD from simple obesity in both the discovery and validation cohorts. Discussion These findings highlight the pivotal role of TRP metabolism in the pathogenesis of MASLD in children and adolescents with obesity. Elevated TRP‐derived metabolites may contribute to hepatic lipid accumulation through oxidative stress and serve as promising non‐invasive biomarkers for the early diagnosis of obesity‐related MASLD. The machine learning–based diagnostic model offers a practical and less invasive alternative to current diagnostic approaches such as liver biopsy.