Plasma adiponection in nonalcoholic fatty liver is related to heoatic insulin resistence andhepatic fat cotent not to liver disease severity

lasma levels of adiponectin are decreased in patients with nonalcoholic fatty liver disease (NAFLD), but the relationship among plasma adiponectin, insulin sensitivity, and histological features is unclear. In 174NAFLD patients and 42 controls, we examined plasma adiponectin concentrations in relation to 1) lipid profile, indices of insulin resistance, and features of the metabolic syndrome (n = 174); 2) hepatic insulin resistance (clamp technique with tracer infusion) (10 patients); and 3) histological features at liver biopsy (n = 116). When the data from all subjects were combined, plasma adiponectin levels were positively associated with increased age, female gender, and plasma high-density lipoprotein levels, and negatively associated with waist circumference, body mass index, triglycerides, indices of insulin resistance, and aminotransferase levels, and also predicted the presence of the metabolic syndrome. In step-wise regression, increased age, female gender, waist circumference, triglyceride levels, and homeostasis model assessment independently associated with adiponectin (adjusted R2, 0.329). In NAFLD, adiponectin was only associated with increased age, female gender, and triglycerides (adjusted R2, 0.245). When the measured histological parameters were included in the model, plasma adiponectin levels were also inversely proportional to the percentage of hepatic fat content (adjusted R2, 0.221), whereas necroinflammation and fibrosis did not fit in the model. Adiponectin was negatively correlated with insulin-suppressed endogenous glucose production during the clamp (P = 0.011). The results demonstrate that decreased levels of circulating adiponectin in NAFLD are related to hepatic insulin sensitivity and to the amount of hepatic fat content. Hypoadiponectinemia in NAFLD is part of a metabolic disturbance characterized by ectopic fat accumulation in the central compartment. (J Clin Endocrinol Metab 90: 3498–3504, 2005)