Relationship between leptin, insulin, body composition and liver steatosis in non-diabetic moderate drinkers with normal transaminase levels.

ABSTRACT

OBJECTIVE: Obesity and insulin resistance play a major role in the development of liver steatosis (LS), but also relative leptin resistance has been reported to correlate with LS in humans. Our objective was to investigate the relationship between serum leptin, insulin, obesity and LS in non-diabetic males (n = 74) and postmenopausal females (n = 50) with normal transaminase levels and low-to-moderate alcohol intake. METHODS: A medical history to retrieve information about health status, current medications, alcohol consumption and history of viral or toxic hepatitis; a physical examination including height, weight, waist circumference and blood pressure; a fasting blood draw for the determination of glucose, insulin, leptin, lipid profile, transaminases and uric acid; an oral glucose tolerance test to exclude type 2 diabetes; a dual-energy X-ray absorptiometry scan to assess fat mass (FM) and lean body mass (LBM), and an echography of the liver to assess LS. RESULTS: Fasting leptin and insulin were highly correlated with FM in men (R = 0.767 and R = 0.495 respectively, P < 0.001) and women (R = 0.713 and R = 0.526 respectively, P < 0.001). After correction for FM, leptin showed a significant negative correlation with LBM in men (R = -0.240, P = 0.039), but not in women (R = -0.214, P = 0.132). The positive relationship observed between leptin, insulin and LS persisted after adjustment of leptin and insulin for body composition only in men (R = 0.415, P < 0.001 and R = 0.339, P = 0.003 respectively for leptin and insulin vs LS). Adjusted means (95% confidence intervals) of leptin increased significantly across categories of LS in men even when insulin was considered in the model (absent = 7.1 ng/ml (5.6-8.5), mild = 8.2 ng/ml (7.2-9.2), moderate/severe = 12.1 ng/ml (10.3-14.0); P < 0.001), whereas no significant relationship was observed between insulin and LS after leptin was accounted for. CONCLUSION: Serum concentrations of leptin and insulin are positively correlated in men independently of body composition, but not in postmenopausal women. In men, the steatogenic effect of hyperinsulinemia/insulin resistance in the context of low-to-moderate alcohol consumption appears to be mediated by high concentrations of serum leptin, whereas body fat alone could identify postmenopausal women at high risk for LS.