Abstract
PURPOSE: The aim of this study was to comprehensively compare the predictive effect of 10 body fat indexes on MAFLD in different sex, age and BMI subgroups. PATIENTS AND METHODS: A total of 5403 physical examination data were included and divided into the MAFLD group (N=2632) and non-MAFLD group (N=2771). The differences and correlation of 10 promising indicators between the two groups were compared, including fatty liver index (FLI), hepatic steatosis index (HSI), lipid accumulation product (LAP), visceral fat index (VAI), cardiometabolic index (CMI), body adiposity index (BAI), and triglyceride-glucose index (TyG), waist circumference index (WC), body mass index (BMI), waist to height ratio (WHtR). Logistic regression was used to analyze the risk of MAFLD under different adjustment conditions. The operating characteristic curve of different genders, BMI levels and age subgroups was plotted. RESULTS: Male gender, smoking, alcohol drinking, and higher age are risk factors for MAFLD. In addition to BAI, the other 9 indicators had a high correlation with MAFLD, the area under the curve (AUC) value was >0.7, and the prediction effect was better in females, BMI<24 kg/m2, age <35 years subgroup, among which FLI (AUC: 0.912, 95% CI: 0.905-0.920), LAP (AUC: 0.894, 95% CI: 0.8866-0.903), and HSI (AUC: 0.881, 95% CI: 0.872-0.890) have better prediction effects. CONCLUSION: Our study confirmed the accuracy of body fat-related indexes in predicting MAFLD in people of different sexes, ages, and BMI levels. Among them, FLI, LAP and HSI have high predictive value and can be utilized as simple and cost-effective tools for screening MAFLD in clinical settings.