Impact of adiponectin and ghrelin on incident glucose intolerance and on weight change

Adiponectin and ghrelin are associated with adiposity and type 2 diabetes in several studies. We sought to prospectively determine the interaction of adiponectin and ghrelin in the development of adiposity and hyperglycaemia. Prospective observational study. 393 community-dwelling Afro-Jamaicans (mean age 47 +/- 13 years; BMI 27.3 +/- 6.3 kg/m(2); 63% women) without glucose intolerance at baseline. Anthropometry, fasting plasma glucose, 2-h plasma glucose, insulin resistance (HOMA-IR), adiponectin and ghrelin concentrations were measured at baseline and 4.1 +/- 0.9 years later. Multivariate analyses were used to explore the associations of HOMA-IR, adiponectin and ghrelin with weight change and glycaemia. The mean weight change was 2.6 +/- 5.5 kg. There were 114 incident cases of impaired glucose tolerance (IGT) and 35 cases of diabetes mellitus. Adiponectin was positively correlated with age and female sex (P-values < 0.01). After adjusting for age and sex, adiponectin and ghrelin were significantly correlated with weight at baseline and follow-up. However, they were not associated with weight change even after further adjustment for baseline weight. Adiponectin, but not ghrelin, was associated with 2-h glucose concentrations at follow-up even after adjusting for age, sex, HOMA-IR and BMI (P = 0.04). In the fully adjusted logistic regression model, adiponectin predicted incident IGT (OR 0.93; 95% CI: 0.87-0.99) and attenuated the effect of BMI on incident IGT. These longitudinal data show that adiponectin and ghrelin may not be causally involved in the development of obesity. However, adiponectin is independently associated with decreased risk of incident IGT.