GH stimulation tests: evaluation of GH responses to heat test versus insulin-tolerance test

Objective: Heat exposure has been shown to stimulate GH release, but the sp ecificity and the reproducibility have not been determined, and the test ha s not been compared with validated GH stimulation tests in adulthood. We th erefore tested the specificity and the reproducibility of the heat exposure test in healthy subjects and compared the results with those obtained with the insulin-tolerance test (ITT). Design: Ten healthy non-obese men, aged 31.3 +/- 4.80 years, underwent four GH stimulation tests in random order: two ITTs and two heat exposure tests . In the heat test, subjects were placed in a hot bath with water temperatu re at 40.3 +/- 0.11 degrees C for 45 min, resulting in an identical (P = 0. 477) significant increase in tympanic temperature of 1.26 +/- 0.05 and 1.41 +/- 0.07 degrees C in the two tests. Results: Peak GH response to the heat exposure test was less than the peak GH response to ITT (5.25 +/- 1.72 vs 15.5 +/- 3.17 mu g/l, P= 0.006). Furth ermore the specificity (arbitrary cut-off level= 3 mu g/l) of the heat test was lower than of the ITT (8/17 vs 18/20, P = 0.006). The coefficient of V ariation did not differ between the two tests (heat test 0.31, ITT 0.36, P= 0.77). Peak GH values in the individual tests were highly correlated (heat , r= 0.908, P = 0.002; ITT, r= 0.815, P = 0.004). Reproducible increments i n the circulating levels of stress hormones were observed during ITT, but t hese hormones remained largely unchanged during heat exposure. Conclusions: The heat exposure test is not a reliable GH stimulation test c ompared with the ITT in adults. This study documents that the ITT has a hig h specificity and reproducibility in the diagnosis of GH deficiency in adul thood. We propose that the heat exposure test is not used in the diagnosis of this condition in adulthood.