Abstract
BACKGROUND: Non-alcoholic fatty liver disease (NAFLD) is increasingly observed in non-obese individuals. The ZJU (Zhejiang University) index has been established as a new and efficient tool for detecting NAFLD, but the relationship between the ZJU index and NAFLD within non-obese individuals still remains unclear. METHODS: A post-hoc evaluation was undertaken using data from a health assessment database by the Wenzhou Medical Center. The participants were divided into four groups based on the quartile of the ZJU Index. Cox proportional hazards regression, Kaplan-Meier analysis and tests for linear trends were used to evaluate the relationship between the ZJU index and NAFLD incidence. Subgroup analysis was conducted to test the consistency of the correlation between ZJU and NAFLD in subsgroups. Receiver operative characteristic (ROC) curve analysis was performed to evaluate the predictive performance of the ZJU index, compared with the Atherogenic index of plasma (AIP) and Remnant lipoprotein cholesterol (RLP-C) index. RESULTS: A total of 12,127 were included in this study, and 2,147 participants (17.7%) developed NAFLD in 5 years follow-up. Participants in higher ZJU quartiles tended to be female and have higher liver enzymes (including ALP, GGT, ALT, AST), GLU, TC, TG, LDL and higher NAFLD risk. Hazard Ratios (HR) and 95% confidence intervals (CI) for new-onset NAFLD in Q2, Q3, and Q4 were 3.67(2.43 to 5.55), 9.82(6.67 to 14.45), and 21.67(14.82 to 31.69) respectively in the fully adjusted model 3. With increased ZJU index, the cumulative new-onset NAFLD gradually increased. Significant linear associations were observed between the ZJU index and new-onset NAFLD (p for trend all<0.001). In the subgroup analysis, we noted a significant interaction in sex, with HRs of 3.27 (2.81, 3.80) in female and 2.41 (2.21, 2.63) in male (P for interaction<0.01). The ZJU index outperformed other indices with an area under the curve (AUC) of 0.823, followed by AIP (AUC=0.747) and RLP-C (AUC=0.668). CONCLUSION: The ZJU index emerges as a promising tool for predicting NAFLD risk in non-obese individuals, outperforming other existing parameters including AIP and RLP-C. This could potentially aid in early detection and intervention in this specific demographic.