The Risk of Type 2 Diabetes and Coronary Artery Disease in Non-obese Patients with Non-alcoholic Fatty Liver Disease: A Cohort Study

Background: Non-alcoholic fatty liver disease (NAFLD) is not uncommon in non-obese subjects, referred to as non-obese NAFLD. It is not fully determined whether non-obese NAFLD is associated with increased risks of type 2 diabetes (T2D) and coronary artery disease (CAD) in Chinese. This study aimed to examine the association between NAFLD and risks of T2D and CAD in a non-obese Chinese population. Methods: The present cohort study included two stages. In the first cross-sectional study, 16,093 non-obese subjects with a body max index (BMI) < 25.0 kg/m 2 were enrolled from The Second Xiangya Hospital, China, from 2011 to 2014. Hepatic steatosis was evaluated by ultrasonography examination. Logistic regression analyses were used to examine the association of non-obese NAFLD with T2D and CAD at baseline. In the subsequent 5-year follow-up study, 12,649 subjects free of T2D and CAD at baseline were included, and the incidence of T2D and CAD were observed. Cox proportional hazard regression analyses were performed to determine the risk of incident T2D and CAD with NAFLD. Results: At baseline, the prevalence of NAFLD, T2D and CAD were 10.7% (1,717/16,093), 3.3% (529/16,093) and 0.7% (113/16,093), respectively. The univariate logistic regression analyses showed NAFLD associated with both T2D and CAD. Moreover, in a multivariate logistic regression model, NAFLD remained independently associated with T2D (OR: 2.7, 95% CI: 2.2–3.3, p < 0.001). However, no significant association was found between NAFLD and CAD by the multivariate logistic regression analyses (OR: 1.1, 95% CI: 0.6–1.8, p = 0.854). During a 5-year follow-up period, 177 (1.4%) patients developed T2D, and 134 (1.1%) developed CAD, respectively. In univariate Cox regression models, NAFLD associated with both T2D and CAD. Moreover, the multivariate Cox regression analysis revealed that NAFLD independently associated with an increased risk of T2D (HR: 2.3, 95% CI: 1.7–3.2, p < 0.001). However, the association between NAFLD and incident CAD was lost in the multivariate Cox regression analysis (HR = 1.5, 95% CI: 1.0–2.4, p = 0.059). Conclusions: NAFLD was an independent risk factor for T2D in non-obese subjects. However, no significant association was observed between non-obese NAFLD and incident CAD after adjusting other traditional cardiovascular risk factors, suggesting these factors might mediate the increased incidence of CAD in non-obese NAFLD patients.