The Predictive Value Of BMI?AST And TBIL In Advanced Fibrosis Of Nonalcoholic Fatty Liver Disease

Objective: The aim of the study is to explore the predictive value of body mass index(BMI),aspartate aminotransferase(AST)and total bilirubin(TBIL)in advanced fibrosis of nonalcoholic fatty liver disease(NAFLD),and develop a model able to predict advanced fibrosis.Methods: In this study,212 NAFLD patients with complete medical records admitted to our department from April 2017 to June 2019 were collected.Patients with viral liver disease,alcoholic liver disease,autoimmune liver disease,drug-induced liver disease and severe liver and kidney dysfunction were excluded,and patients with age < 18 were excluded.The general characteristics,biochemical and imageological tests of NAFLD patients meeting the criteria were recorded.According to different Fibro Scan LSM,the subjects were divided into two groups.Those with LSM < 9.8Kpa were included in the non-advanced fibrosis group,and those with LSM? 9.8 Kpa were included in the advanced fibrosis group.SPSS24.0 statistical software was used to conduct analysis on variables that might affect the degree of hepatic fibrosis.Variables with statistical significance were included in binary logistic regression analysis to obtain risk factors for NAFLD advanced fibrosis.A model for prediction of NAFLD advanced fibrosis was established,and the predictive value of the model,BMI,AST and TBIL for NAFLD advanced ibrosis was compared according to the area under the Receiver Operating Characteristic Curve(ROC).Results: 1.Among 212 patients with NAFLD,183 were in the non-advanced fibrosis group and 29 in the advanced fibrosis group.BMI,AST,ALT and TBIL in the advanced fibrosis group were significantly higher than those in the non-advanced fibrosis group(P < 0.05).2.Multivariate analyses using binary logistic regression showed that BMI,AST and TBIL were independent risk factors for advanced fibrosis in NAFLD patients.The above variables were included in binary logistic regression.Based on the results of analysis,a logistic regression model(the following named as model)was established to predict the advanced fibrosis,the specific formula as follows:-9.127+0.183×BMI+0.042×AST+0.048×TBIL.3.ROC Curve was drawn using BMI,AST,TBIL and the model with LSM? 9.8kpa as the cut-off point,and the Area Under the Curve(AUC)was used to evaluate the predictive value of each index for NAFLD advanced fibrosis.The predictive value of the model(AUC = 0.789)for NAFLD advanced fibrosis was better than that for BMI(AUC = 0.691),AST(AUC = 0.74)and TBIL(AUC = 0.639)(P < 0.05).Conclusion: 1.The levels of BMI,AST,ALT and TBIL in the NAFLD advanced fibrosis group were significantly higher than those in the non-advanced fibrosis group.2.BMI,AST and TBIL are independent risk factors for advanced fibrosis in patients with NAFLD.3.The predictive value of the model based on the composition of BMI,AST,and TBIL for advanced fibrosis in NAFLD patients was superior to the individual diagnosis of the three.