Abstract
BACKGROUND: Metabolic factors are considered to influence disease progression in patients with Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD), but the impact of individual metabolic factors on the survival rate of patients with MASLD is still unclear. AIMS: This article aims to reveal how metabolic components affect the survival of patients with this disease. METHODS: A total of 3,086 participants with MASLD based on the diagnostic criteria established at the Delphi conference from NHANES III were included in this analysis. COX regression model (C-index = 0.64) was used to analyze the all-cause and attributable mortality of different number of metabolic factors. Elastic Network Regression model (C-index = 0.69), Accelerated Failure Time model and Randomized Survival Forest model (C-index = 0.63) based on machine learning were used to analyze the weight of each metabolic factor, and a Metabolism-related survival risk score formula was established and verified. RESULTS: This study found that not only the number of metabolic factors had different effects on all-cause survival in MASLD patients, but also the degree of impact of different metabolic factors on survival was quite different, among which poor glycemic control was the most important influencing factor. CONCLUSION: This study highlights the clinical value of relevant metabolic factors in predicting survival in the MASLD patient population. Related metabolic factors can be used as surrogate biomarkers for the follow-up of MASLD patients.