Abstract
The association between estimated glucose disposal rate (eGDR) and metabolic-associated fatty liver disease (MAFLD) is currently lacking. This study aimed to investigate the relationship between eGDR-related indices and the prevalence of MAFLD. The study analyzed data from the 2017 to 2018 cycle of the National Health and Nutrition Examination Survey. Multivariate logistic regression models were employed to examine the relationship between eGDR-related indices and MAFLD. Receiver operating characteristic curves were used to compare the diagnostic validity of eGDR-related indices and other insulin resistance indicies. A total of 2273 participants were included. In the fully adjusted multivariate logistic regression model, eGDR-waist-hip ratio (OR, 0.65; 95% CI, 0.58-0.72; P <.001), eGDR-waist circumference (WC) (OR, 0.62; 95% CI, 0.56-0.67; P <.001), eGDR-body mass index (BMI) (OR, 0.62; 95% CI, 0.56-0.67; P <.001), and lnGDR (OR, 0.76; 95% CI, 0.66-0.88; P <.01) were negatively associated with MAFLD. Mediation analysis indicated that android: gynoid ratio played a role in this association, explaining 28.8%, 17.2%, 18.6%, and 33.6% of the associations, respectively (all P for mediation <.001). Receiver operating characteristic curves found that eGDR-WC (areas under curves, 0.794; 95% CI, 0.776-0.813) and eGDR-BMI (areas under curves, 0.791; 95% CI, 0.772-0.810) had greater diagnostic validity for MAFLD. The present study suggests a potential inverse association between eGDR and MAFLD. In addition, eGDR-WC and eGDR-BMI could potentially serve as useful screening metrics for identifying populations at early risk of MAFLD.