To evaluate if plasma kisspeptin concentrations are associated with insulin secretion, as suggested by recent in vitro studies, independently of confounders. 261 nondiabetic subjects were stratified into tertiles according to kisspeptin values. Insulin secretion was assessed using indexes derived from oral glucose tolerance test (OGTT). After adjusting for age, gender, and BMI, subjects in the highest (tertile 3) kisspeptin group exhibited significantly lower values of insulinogenic index, corrected insulin response (CIR30), and Stumvoll indexes for first-phase and second-phase insulin release as compared with low (tertile 1) or intermediate (tertile 2) kisspeptin groups. Univariate correlations between kisspeptin concentration and metabolic variables showed that kisspeptin concentration was significantly and positively correlated with age, blood pressure, and 2-h post-load glucose, and inversely correlated with BMI, and waist circumference. There was an inverse relationship between kisspeptin levels and OGTT-derived indexes of glucose-stimulated insulin secretion. A multivariable regression analysis in a model including all the variables significantly correlated with kisspeptin concentration showed thar age (β = -0.338, P<0.0001), BMI (β = 0.272, P<0.0001), 2-h post-load glucose (β = -0.229, P<0.0001), and kisspeptin (β = -0.105, P = 0.03) remained associated with insulinogenic index. These factors explained 34.6% of the variance of the insulinogenic index. In conclusion, kisspeptin concentrations are associated with insulin secretion independently of important determinants of glucose homeostasis such as gender, age, adiposity, 2-h post-load glucose, and insulin sensitivity.
Read full abstract