The NAFL Risk Score: A simple scoring model to predict 4-y risk for non-alcoholic fatty liver

Background Although several risk factors for non-alcoholic fatty liver (NAFL) have been reported, there are few clinical scores that predict its incidence in the long term. We developed and validate a scoring model for individual prediction of 4-y risk for NAFL. Methods Four-year follow-up data of 8226 initially NAFL-free subjects enrolled for an annual physical examination from Wenzhou Medical Center were analyzed. These subjects are randomly split into the training and the validation cohort. Univariate and multivariable logistic regression models were employed for model development. The selected variables were assigned an integer or half-integer risk score proportional to the estimated coefficient from the logistic model. Risk scores were tested in a validation cohort. We also compared the predictive performance of with that of the NAFLD Index by computing the area under the receiver operating characteristic curve (AUROC). Results The NAFL Risk Score was developed as 0 to 18 points comprising of BMI, TG × GGT, ALT/AST, LDL-C/HDL-C and UA in both sexes. Comparison of the observed with the estimated incidence of NAFL at both cohorts showed satisfactory precision. In addition, the NAFL Risk Score showed relatively good discriminative power (AUROC = 0.739 for males, 0.823 for females) compared with the NAFLD Index (AUROC = 0.661 for males, 0.729 for females) in these Chinese subjects. Conclusions We developed and validated the NAFL Risk Score, a new scoring model to predict 4-y risk for NAFL. The NAFL Risk Score may be clinically simple and useful for assessing individual risk for NAFL.