Prediction of metabolic dysfunction-associated steatotic liver disease among patients with type 2 diabetes mellitus using liver fat indices in a resource-limited clinical setting

Abstract

Metabolic dysfunction-associated steatotic liver disease (MASLD), previously termed non-alcoholic fatty liver disease (NAFLD), is the hepatic presentation of metabolic syndrome. It shares the common pathophysiology of insulin resistance with type 2 diabetic mellitus (T2DM). The aim of this study was to establish the predictive ability of non-invasive liver fat indices—Fatty Liver Index (FLI), Hepatic Steatosis Index (HSI), Visceral Adiposity Index (VAI), Triglyceride and Glucose Index (TyG), and NAFLD-Liver Fat Score (NAFLD-LFS)—to determine fatty liver among patients with T2DM. A cross-sectional analytical study was conducted on 260 newly diagnosed patients with type 2 diabetes. Anthropometric measurements were taken, while ultrasound scan was performed to determine fatty liver; its grade and blood samples were used to determine biochemical parameters. Liver fat index scores were calculated using standardized equations and analyzed using the statistical software SPSS version 23.0. The incidence of MASLD among patients with T2DM was 69.2%. Three of the liver fat indices—FLI, HSI, and NAFLD-LFS—showed strong statistically significant correlations ( p  < 0.001) with liver ultrasonographic findings while the remaining indices—VAI and TyG—showed moderately significant correlations ( p  < 0.01). The cutoff values, sensitivities, and specificities of each index were determined by receiver operating characteristic (ROC) curve analysis. FLI showed the highest diagnostic performance [area under the curve (AUC) = 0.947; 95% confidence interval (CI): 0.923–0.971] with 75% sensitivity and 100% specificity with a cutoff value of 46.12. Both HSI (AUC = 0.847; 95% CI: 0.812–0.902) and NAFLD-LFS (AUC = 0.847; 95% CI: 0.800–0.894) showed good diagnostic performance with sensitivities of 68.3% and 76.1%, specificities of 95.5% and 85%, and cutoff values of 37.02 and 0.906, respectively. A moderate diagnostic performance was observed in the remaining indices: TyG (AUC = 0.707; 95% CI: 0.642–0.772) and VAI (AUC = 0.670; 95% CI: 0.602–0.739). Based on the multivariable logistic regression analysis, both FLI [adjusted odds ratio (aOR) = 1.144, 95% CI: 1.090–1.201, p ≤ 0.001] and HSI (aOR = 1.366, 95% CI: 1.157–1.614, p ≤ 0.001) were able to independently predict the presence or absence of fatty liver. The current study showed that all five liver fat indices have the ability to predict the presence or absence of fatty liver among patients with T2DM. FLI and HSI showed the highest performance among the five indices, which can be used as screening tools for fatty liver more efficiently. The findings provide reliable alternative methods for detecting fatty liver in a resource-limited clinical setting simplifying patient management.