Indices of insulin action, disposal, and secretion derived from fasting samples and clamps in normal glucose-tolerant black and white children.

Abstract

To validate fasting indices of insulin sensitivity and secretion in a diverse pediatric population against gold standard estimates from euglycemic and hyperglycemic clamps. A total of 31 children (mean BMI 25.1 +/- 4.9 kg/m(2), mean age 8.7 +/- 1.4 years, 15 girls and 16 boys, 12 black and 19 white) underwent euglycemic and hyperglycemic clamps 2-6 weeks apart to derive insulin sensitivity indices (SI (Eug clamp) and SI (Hyper clamp)). Fasting samples were used to derive the homeostasis model assessment of insulin resistance index (HOMA-IR), HOMA of percent beta-cell function (HOMA-B%), quantitative insulin sensitivity check index (QUICKI), insulinogenic index, antilipolytic insulin sensitivity index (ISI-FFA), and C-peptide-to-insulin ratio. The QUICKI correlated best with SI (Eug clamp) (r = 0.69, P < 0.05) and had greater correlations to SI (Eug clamp) than did either SI (Hyper clamp) (r = 0.45, P < 0.05) or the HOMA-IR (r = -0.51, P < 0.05). Both fasting insulin and the insulinogenic index correlated well with first- and steady-phase insulin secretion (r's from 0.79 to 0.86, P < 0.05). HOMA-B% was not as highly correlated (r = 0.69-0.72, P < 0.05). Fasting C-peptide-to-insulin ratio was not significantly correlated with clamp-derived metabolic clearance rate of insulin. ISI-FFA was not correlated with the degree of free fatty acid suppression obtained from the clamps. The QUICKI, fasting insulin, and the insulinogenic index all closely correlate with corresponding clamp-derived indices of insulin sensitivity and secretion in this diverse pediatric cohort. These results, if replicated in similarly diverse populations, suggest that estimates based on fasting samples can be used to rank order insulin secretion and sensitivity in pediatric cohorts.