Fasting and OGTT-derived measures of insulin resistance as compared with the euglycemic-hyperinsulinemic clamp in nondiabetic Finnish offspring of type 2 diabetic individuals.

Abstract

It remains to be established which surrogate marker is the most predictive of insulin resistance. We evaluated the incremental value of the Matsuda insulin sensitivity index (ISI) to homeostasis model assessment of insulin resistance (HOMA-IR) for detecting insulin resistance. This was a cross-sectional analysis of the EUGENE2 Kuopio cohort. Participants included 266 nondiabetic Finnish offspring of type 2 diabetic individuals (aged 24-50 years). Insulin resistance (the lowest whole-body glucose uptake quartile) was measured by the euglycemic-hyperinsulinemic clamp. We used the area under the receiver operating characteristic curve, an insensitive method to model improvement. Reclassification was assessed by the net reclassification improvement and integrated discrimination improvement. We generated four strata using, as cut points, the 0.05, 0.25, and 0.70 probabilities of having insulin resistance. These were observed probabilities at body mass index of 20, 27, and 35 kg/m(2), respectively. The area under the receiver operating characteristic curve of the Matsuda ISI based on 5 time points (0, 30, 60, 90, and 120 minutes) did not differ statistically from that of HOMA-IR (0.897 and 0.875, P = .080). However, the Matsuda ISI added incremental value to HOMA-IR for the detection of insulin-resistant individuals (net reclassification improvement, 26.2%, P < .001; integrated discrimination improvement, 6.3%, P < .001). A modified Matsuda ISI based on 4 time points (0, 30, 60, and 120 minutes) yielded similar results. The Matsuda ISI has additional value for the detection of insulin resistance beyond the ability of HOMA-IR. The Matsuda ISI reclassifies a net of 26% of individuals more appropriately.