RESUMO
BACKGROUND: According to research, the fatty liver index (FLI) is associated with diabetes. However, few studies have been conducted to investigate the relationship between FLI and diabetes risk from various perspectives. This study comprehensively investigated the relationship between FLI and incident diabetes in a large Japanese population. METHODS: This retrospective cohort study included 14,280 participants from Murakami Memorial Hospital in Japan from 2004 to 2015. The independent and dependent variables are FLI and risk of type 2 diabetes mellitus (T2DM), respectively. To examine the link between FLI and incident T2DM, Cox proportional-hazards regression was employed. In addition, we performed a number of sensitivity studies to guarantee the validity of the results. Moreover, we conducted subgroup analyses. RESULTS: After adjusting covariates, the results showed that FLI was positively associated with the risk of T2DM (HR = 1.019, 95%CI: 1.012, 1.025). Additionally, the sensitivity analysis showed how reliable the outcomes were. And a stronger association between FLI and incident T2DM was observed in the regular exercisers (HR = 1.036, 95%CI: 1.019-1.053, P < 0.0001) and the population without ethanol consumption (HR = 1.028, 95%CI: 1.017-1.039, P < 0.0001). Besides, receiver operating characteristic (ROC) curve analysis showed that FLI was better than waist circumference, triglycerides, body mass index, and gamma-glutamyl transferase in predicting incident T2DM. CONCLUSION: FLI is positively associated with incident T2DM.