What is the effect of rosiglitazone treatment on insulin secretory function in insulin-resistant individuals? It depends on how you measure it.

The goal of this study was to compare methods used to quantify the effect of rosiglitazone (RSG) on insulin secretory function, particularly estimates based on changes in fasting plasma glucose and insulin concentration vs daylong insulin responses to meals. To do this, we compared these measures of insulin secretion before and 3 months after RSG treatment in insulin-resistant individuals, subdivided into nondiabetic subjects (n = 29) and patients with type 2 diabetes mellitus (2DM) (n = 22). Insulin resistance was quantified by determining the steady-state plasma glucose concentration during the insulin suppression test and insulin secretory function by homeostasis model assessment of β-cell function (HOMA-β) and the total integrated daylong plasma insulin responses to mixed meals (insulin area under the curve). Baseline fasting and daylong plasma glucose concentrations were higher (P < .001) in patients with 2DM, associated with lower HOMA-β values (P < .001). However, neither fasting nor daylong insulin concentrations after mixed meals differed in the 2 groups. Insulin sensitivity improved (P < .001) after RSG administration, with decreases of 31% ± 23% and 21% ± 14% in steady-state plasma glucose concentration in nondiabetic and diabetic subjects, respectively. Although fasting and daylong plasma glucose and insulin concentrations fell (P < .001) in both groups of RSG-treated individuals, HOMA-β decreased in nondiabetic subjects and did not change in those with 2DM. In conclusion, RSG administration improved insulin sensitivity in both groups, associated with lower fasting and daylong glucose concentrations. Fasting and daylong insulin concentrations were also lower in both groups of RSG-treated subjects, but the values of HOMA-β indicated either a decrease (nondiabetics) or no change (diabetics) in insulin secretory function. These results suggest that measurements of HOMA-β may not provide a complete view of insulin secretory function, either when comparing diabetic with nondiabetic individuals or when assessing the response to RSG treatment in insulin-resistant individuals.