Prevalence of non-alcoholic fatty liver disease and its association with cardiovascular disease in patients with type 1 diabetes

Abstract

Background & Aims To estimate the prevalence of non-alcoholic fatty liver disease (NAFLD) in type 1 diabetic individuals, and to evaluate whether NAFLD is associated with increased prevalence of cardiovascular disease (CVD). Methods All patients with diagnosed type 1 diabetes with available liver ultrasound data (n = 250), who regularly attended our diabetes clinic, were enrolled. Main study measures were detection of NAFLD (by patient history and liver ultrasound) and asymptomatic/symptomatic CVD (by patient history, chart review, electrocardiogram, and echo-Doppler scanning of carotid and lower limb arteries). Results The prevalence of NAFLD was 44.4%, and NAFLD was the most common cause (69.8%) of hepatic steatosis on ultrasound examination. Patients with NAFLD had a remarkably higher (p <0.001) age- and sex-adjusted prevalence of coronary (10.8% vs. 1.1%), cerebrovascular (37.3% vs. 5.5%) and peripheral (24.5% vs. 2.5%) vascular disease than their counterparts without NAFLD. In logistic regression analysis, NAFLD was associated with prevalent CVD (as composite endpoint), independently of age, sex, diabetes duration, hemoglobin A1c, smoking history, systolic blood pressure, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, and medication use (adjusted odds ratio 7.36, 95% confidence intervals 1.60–34.3, p <0.01). Conclusions Our findings suggest that NAFLD is very common in type 1 diabetic subjects and is associated, independently of several confounding factors, with a higher prevalence of CVD. Future prospective studies are needed to evaluate whether NAFLD predicts incident CVD events in type 1 diabetes. To estimate the prevalence of non-alcoholic fatty liver disease (NAFLD) in type 1 diabetic individuals, and to evaluate whether NAFLD is associated with increased prevalence of cardiovascular disease (CVD). All patients with diagnosed type 1 diabetes with available liver ultrasound data (n = 250), who regularly attended our diabetes clinic, were enrolled. Main study measures were detection of NAFLD (by patient history and liver ultrasound) and asymptomatic/symptomatic CVD (by patient history, chart review, electrocardiogram, and echo-Doppler scanning of carotid and lower limb arteries). The prevalence of NAFLD was 44.4%, and NAFLD was the most common cause (69.8%) of hepatic steatosis on ultrasound examination. Patients with NAFLD had a remarkably higher (p <0.001) age- and sex-adjusted prevalence of coronary (10.8% vs. 1.1%), cerebrovascular (37.3% vs. 5.5%) and peripheral (24.5% vs. 2.5%) vascular disease than their counterparts without NAFLD. In logistic regression analysis, NAFLD was associated with prevalent CVD (as composite endpoint), independently of age, sex, diabetes duration, hemoglobin A1c, smoking history, systolic blood pressure, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, and medication use (adjusted odds ratio 7.36, 95% confidence intervals 1.60–34.3, p <0.01). Our findings suggest that NAFLD is very common in type 1 diabetic subjects and is associated, independently of several confounding factors, with a higher prevalence of CVD. Future prospective studies are needed to evaluate whether NAFLD predicts incident CVD events in type 1 diabetes.