The diagnosis of severe growth hormone deficiency in elderly patients with hypothalamic-pituitary disease.

Abstract

OBJECTIVE: Adults over the age of 60 years with organic disease of the hypothalamic-pituitary axis may be deficient in growth hormone (GH) to a degree that is distinct from the age-related decline in GH secretion and sufficient to cause perturbations of body composition, serum lipid profile and bone metabolism. In order to determine the best method for detecting GH deficiency in this age group we have compared spontaneous GH secretion, a provocative test of GH secretion, the arginine stimulation test (AST), and basal estimates of circulating insulin-like growth factors (IGF) and IGF-binding proteins (IGFBP). DESIGN: Twenty-four patients (16 male) with organic hypothalamic-pituitary disease and 24 controls (17 male) were studied. The groups were matched for BMI but the patients were slightly younger than the controls, 66.0 (61.0-85.7) vs. 70.6 (60.8-87.5) years (P = 0.04). All subjects underwent a 24-h GH profile (20-minute sampling), measurement of serum IGF-I, IGF-II, IGFBP3, IGFBP2 and growth hormone binding protein (GHBP) and, after an overnight fast, an AST (intravenous arginine 20 g/m2 over 30 minutes). GH concentrations were measured using an ultrasensitive chemiluminescence assay (sensitivity 0.002 microgram/l). Normative data for serum IGF-I, IGF-II, IGFBP3 and IGFBP2 were obtained from 125 subjects aged 60-87 years. RESULTS: None of the parameters studied was able to distinguish between all the GH deficient patients and the healthy controls. The median (range) area under the GH profile (AUCGH) and peak GH response to arginine were lower in the patients than in the controls, 310.05 (18.90-2193.36) vs. 2518.20 (526.76-12024.97) min mU/l (P < 0.00001), 1.07 (0.08-17.90) vs. 23.06 (4.60-109.98) mU/l (P < 0.00001), respectively. There was a significant relationship between AUCGH and peak GH response to arginine in the patients (r = 0.89, P < 0.0001) and in the controls (r = 0.56, P = 0.005). Serum IGF-I, IGFBP2, and IGFBP3 levels were significantly lower in the patients compared with the normal range, 102 (14-162) vs. 142 (59-298) micrograms/l (P < 0.0001), 415 (122-1868) vs. 640 (140-2585) micrograms/l (P = 0.0007) and 2.29 (0.81-3.75) vs. 2.59 (1.00-3.52) mg/l (P = 0.009), respectively. The degree of overlap between the two groups, however, was too great to make these measurements useful diagnostically. Serum IGF-II and GHBP concentrations in the patients were not significantly different from the normal range. The patients were divided into groups determined by the number of anterior pituitary hormone deficits present. There was a significant downward trend in the peak GH response to arginine with increasing severity of hypopituitarism (J = -3.04, P = 0.0012). Ninety per cent of patients with two or three additional pituitary deficiencies had a peak GH response less than 2.0 mU/l. CONCLUSIONS: Of the indices studied the arginine stimulation test is more effective than GH markers, such as IGF-I or IGFBP3, or measurement of spontaneous GH secretion for diagnosing GH deficiency in adults over the age of 60 years. By relating the peak GH response to the degree of hypopituitarism, a GH response less than 2.0 mU/l is suggestive of severe GH deficiency in this age group under the appropriate clinical circumstances.