Interactive Journal of Medical Research (Aug 2024)
Establishment and Evaluation of a Noninvasive Metabolism-Related Fatty Liver Screening and Dynamic Monitoring Model: Cross-Sectional Study
Abstract
BackgroundMetabolically associated fatty liver disease (MAFLD) insidiously affects people's health, and many models have been proposed for the evaluation of liver fibrosis. However, there is still a lack of noninvasive and sensitive models to screen MAFLD in high-risk populations. ObjectiveThe purpose of this study was to explore a new method for early screening of the public and establish a home-based tool for regular self-assessment and monitoring of MAFLD. MethodsIn this cross-sectional study, there were 1758 eligible participants in the training set and 200 eligible participants in the testing set. Routine blood, blood biochemistry, and FibroScan tests were performed, and body composition was analyzed using a body composition instrument. Additionally, we recorded multiple factors including disease-related risk factors, the Forns index score, the hepatic steatosis index (HSI), the triglyceride glucose index, total body water (TBW), body fat mass (BFM), visceral fat area, waist-height ratio (WHtR), and basal metabolic rate. Binary logistic regression analysis was performed to explore the potential anthropometric indicators that have a predictive ability to screen for MAFLD. A new model, named the MAFLD Screening Index (MFSI), was established using binary logistic regression analysis, and BFM, WHtR, and TBW were included. A simple rating table, named the MAFLD Rating Table (MRT), was also established using these indicators. ResultsThe performance of the HSI (area under the curve [AUC]=0.873, specificity=76.8%, sensitivity=81.4%), WHtR (AUC=0.866, specificity=79.8%, sensitivity=80.8%), and BFM (AUC=0.842, specificity=76.9%, sensitivity=76.2%) in discriminating between the MAFLD group and non-fatty liver group was evaluated (P<.001). The AUC of the combined model including WHtR, HSI, and BFM values was 0.900 (specificity=81.8%, sensitivity=85.6%; P<.001). The MFSI was established based on better performance at screening MAFLD patients in the training set (AUC=0.896, specificity=83.8%, sensitivity=82.1%) and was confirmed in the testing set (AUC=0.917, specificity=89.8%, sensitivity=84.4%; P<.001). ConclusionsThe novel MFSI model was built using WHtR, BFM, and TBW to screen for early MAFLD. These body parameters can be easily obtained using a body fat scale at home, and the mobile device software can record specific values and perform calculations. MFSI had better performance than other models for early MAFLD screening. The new model showed strong power and stability and shows promise in the area of MAFLD detection and self-assessment. The MRT was a practical tool to assess disease alterations in real time.