Evaluation of the growth hormone exercise test in normal and growth hormone-deficient children.

A group of 168 short but otherwise normal children (group A) and 25 children deficient in growth hormone (GH) (group B) wwere studied with an exercise stimulation test to determine the expected error of this method. In group A, 125 (74.4%) had maximum GH responses greater than 15 ng/ml, 23 (13.7%) had responses between 10 and 15 ng/ml, and 20 (11.9%) had responses less than 10 ng/ml. The mean +/- SD values were 8.4 +/- 0.4 ng/ml at 0 time, 26.3 +/- 15.0 at 20 minutes, and 10.7 +/- 8.3 at 40 minutes. The mean maximum response was 27.7 +/- 14.3 ng/ml. In group B, 22 (88%) had maximum responses less than 10 ng/ml and 3 (12%) had responses between 10 and 15 ng/ml. Patients with maximum responses less than 10 ng/ml have a high probability of being GH-deficinet, whereas patients with responses between 10 and 15 ng/ml are less likely to be GH-deficient. No patients with responses greater than 15 ng/ml were GH-deficient.