Comparison between insulin-induced hypoglycemia and growth hormone (GH)-releasing hormone plus arginine as provocative tests for the diagnosis of GH deficiency in adults

There is now wide consensus that, within an appropriate clinical context, GH deficiency (GHD) in adults must be shown biochemically by provocative testing of GH secretion and that appropriate cut-off limits have to be defined for each provocative test. Insulin-induced hypoglycemia (ITT) is indicated as the test of choice, and severe GHD, to be treated with recombinant human GH replacement, is de fined by a GH peak response to ITT of less than 3 mu g/L. GKRH + arginine (GHRH+ARG) is one of the most promising tests in alternative to ITT. In fact, it has been reported as a potent, reproducible, and age-independent test and that it is able to distinguish between GHD and normal adults. The aim of the present study was to compare the GH response to ITT and GHRH+ARG in a large group of hypopituitary adults (n = 40; 29 male and 11 female; age: 36.4 +/- 2.1 yr). The third centile limit of the peak GH response to ITT has been reported as 5 mu g/L, whereas in our lab, that to GHRH+ARG is 16.5 mu g/L. In hypopituitary adults, the mean peak GH response to ITT (1.5 +/- 0.2 mu g/L, range: 0.1-8.5 mu g/L)) was lower (P < 0.001) than that to GHRH+ARG (3.0 +/- 0.4 mu g/L, range 0.1-12.0 mu g/L), though there was positive correlation (r = 0.61, P < 0.001) between the GH responses to the 2 tests. The peak GH response to GHRH+ARG, but not that to ITT, was positively (though weakly) associated with insulin-like growth factor-I levels (r = 0.35, P < 0.03). Childhood and adult onset GKD patients, as well as patients with single and multiple pituitary insufficiencies, had similar peak GH responses to ITT or GHRH+ARG. Analyzing individual GH responses, 4/40 (10%) of the hypopituitary patients had GH peaks higher than 5 mu g/L after ITT; moreover, 3 other patients (7%) had GH peaks, after ITT, higher than 3 mu g/L. On the other hand, after GHRH+ARG, all patients had GH peaks lower than 16.5 mu g/L, whereas 21/40 (52.5%) had GK peaks higher than 3 mu g/L. Because 3 mu g/L is the arbitrary cut-off for ITT, the third centile limit of which is 5 mu g/L, we arbitrarily considered 9 mu g/L as the cut-off point for GHRH+ARG. It is noteworthy that 37/40 (92.5%) patients had a GH peak, after GHRH+ARG, below this limit. In conclusion, our present results confirm that the ITT test is a reliable provocative test for the diagnosis of adult GKD, whereas they show that the GHRH+ARG test is, at least, as sensitive as the ITT test (provided that appropriate cut-off limits are considered). Note that even the arbitrary cut-off point below which severe GHD is demonstrated has to be appropriate to the potency of the test.