Test of growth hormone secretion in adults: poor reproducibility of the insulin tolerance test.

Abstract

The insulin tolerance test (ITT) is regarded as the most reliable provocative test in the diagnosis of growth hormone (GH) deficiency in adults. In the present study the test was evaluated by investigating the range of GH responses in normal adult males and females and the intra-individual reproducibility of the test. Sixteen healthy non-obese adults, eight males (median age 31.5 years) and eight females (median age 31.8 years) were tested twice with the ITT, with a minimum of 72 h between each test. The females were tested between day 3 and day 10 of their menstrual cycles. Adequate hypoglycemia was achieved in all cases with a median nadir blood glucose of 1.3 mmol/l (range 0.8-2.0). Growth hormone in serum was measured by immunoradiometric assay and low values were confirmed by a different assay. Median peak GH concentration responses to the ITT were: in males 27.9 micrograms/l, range 5.0-71.1 (test 1) and 30.5 micrograms/l, range 7.9-69.5 (test 2); and in females 9.0 micrograms/l, range 4.1-17.9 (test 1) and 8.4 micrograms/l, range 0.09-42.4 (test 2). The rise in GH concentration during the ITT was higher in males than in females. In the males, all stimulated GH values were > or = 5.0 micrograms/l. In the females, four out of 16 tests gave values below 5.0 micrograms/l and in one test the GH value was around the detection limit of the assays. There was poor reproducibility during repeated testing, with no correlation between the results of the two tests. The results did not correlate to the magnitude of the hypoglycemia.(ABSTRACT TRUNCATED AT 250 WORDS)