Phenotype clustering of metabolic dysfunction‐associated fatty liver disease illustrates indispensable role of uric acid: A data‐driven association and cluster analysis

Abstract The classification of metabolic dysfunction‐associated fatty liver disease (MAFLD), derived from non‐alcoholic fatty liver disease (NAFLD), is a consensus‐driven proposed nomenclature that is defined inclusively as fatty liver disease (FLD) concomitantly with the presence of obesity, metabolic dysregulation, and/or type 2 diabetes mellitus (T2DM). However, questions persist regarding the pertinent factors and dynamic changes of MAFLD under the current definition. A data‐driven cluster analysis was performed on 145 patients meeting both MAFLD and NAFLD criteria at a tertiary hospital in Taiwan. Initially, high‐correlation factors associated with the severity of hepatic steatosis were identified. Subsequently, cluster analysis was employed to delineate distinct phenotypes among MAFLD patients. Finally, the factors were subjected to further analysis based on the categorical classification of MAFLD, steatosis staging, and the resulting clusters. The top five correlation factors—Controlled Attenuation Parameter (CAP), Body Mass Index (BMI), Alanine Aminotransferase (ALT), Triglycerides (TG), Insulin, and Serum Uric Acid (UA)—are significantly associated with the severity of hepatic steatosis. Subsequent clustering unveiled substantial disparities among three cohorts across 12 of 16 biochemical parameters, encompassing CAP, BMI, Aspartate Aminotransferase (AST), ALT, TG, TG‐HDL ratio, insulin, Homeostatic Model Assessment of Insulin Resistance (HOMA‐IR), UA, age, Creatinine (Cr), and liver stiffness. In addition, three distinct MAFLD clusters emerged: Cluster I ( n = 19; characterized by mild hepatic steatosis, mild obesity, and age‐related), Cluster II ( n = 88; featuring moderate hepatic steatosis), and Cluster III ( n = 38; exhibiting severe hepatic steatosis, severe obesity, and insulin resistance‐related markers). In our study, five metabolic parameters, BMI, ALT, TG, insulin, and UA, prove indicative of hepatic steatosis in MAFLD patients. The indispensable role of serum UA in the development and severity of MAFLD was highlighted. This data‐driven cluster analysis also offers an innovative classification tool and personalized medical recommendations for distinct MAFLD phenotyping.