Reproducibility of Growth Hormone and Cortisol Responses to the Insulin Tolerance Test and the Short ACTH Test in Normal Adults

Previous studies have demonstrated poor reproducibility of growth hormone (GH) responses to insulin tolerance testing (ITT). In order to investigate whether this is a particular feature of GH secretion we studied the reproducibility of the GH and cortisol responses to ITT simultaneously and also compared the latter with the reproducibility during short ACTH testing (SAT). Eight normal men (age 26-50) and 8 normal women (age 27-45) underwent 2 ITT and 2 SAT. In the ITT no systematic differences were observed between test 1 and 2 concerning blood glucose, GH and cortisol before and after stimulation. Similar results were obtained for cortisol during SAT. During ITT reproducibility was good for the cortisol response (coefficient of variation [CV] 10%, no sex differences) but poor for the GH response (CV 41% in men, 104% in women). Reproducibility was good for the cortisol response in SAT (CV 12%, no sex differences). The peak cortisol values during ITT (mean 585, range 448-775 nmol/l) were significantly lower than in the 2 SAT (mean 723, range 486-918 nmol/l). We conclude that the GH response during testing is more variable than the cortisol response. This could account for some of the difficulties encountered in the diagnosis of GH deficiency in adults.

Key words

Cortisol - Growth Hormone - Insulin Tolerance Test - Normal Values - Repeatability - Short ACTH Test