X/00/$03.00/0 Vol. 85, No. 4 The Journal of Clinical Endocrinology & Metabolism Copyright 2000 by The Endocrine Society

Transcription

1 X/00/$03.00/0 Vol. 85, No. 4 The Journal of Clinical Endocrinology & Metabolism Printed in U.S.A. Copyright 2000 by The Endocrine Society Diagnosis of Growth Hormone (GH) Deficiency in Adults with Hypothalamic-Pituitary Disorders: Comparison of Test Results Using Pyridostigmine Plus GH-Releasing Hormone (GHRH), Clonidine Plus GHRH, and Insulin- Induced Hypoglycemia as GH Secretagogues* HANS C. HOECK, PETER VESTERGAARD, POUL E. JAKOBSEN, JANNIK FALHOF, AND PETER LAURBERG Department of Endocrinology and Medicine, Aalborg Hospital, DK-9000 Aalborg, Denmark ABSTRACT The insulin tolerance test (ITT) is widely accepted as the method of choice to evaluate GH secretion capacity in adults with hypothalamic-pituitary disorders. However, the test is not suitable in the elderly or in patients with cardiovascular disease or seizure disorders. In recent years alternatives to the ITT have been introduced. The purpose of the present study was to investigate the diagnostic outcome with the ITT, the pyridostigmine plus GHRH (PD GHRH) test, the clonidine plus GHRH (CLO GHRH) test, and insulin-like growth factor I (IGF-I) in an unselected group of patients with hypothalamicpituitary disease. An evaluation of the reproducibility of the different stimulation tests was included in the study. Based on repeated testing with the various GH stimulation tests in healthy adult males and females, the lower limits of normality for the ITT, the PD GHRH test, and the CLO GHRH test were 3.92, 12.8, and 19.0 g/l, respectively. A consecutive group of 26 unselected patients with hypothalamic-pituitary disorders, 13 males and 13 females (median age, 44 ys), were tested twice with all stimulation tests, except that only 10 patients were tested once with the CLO GHRH test due to sideeffects related to clonidine. The peak GH responses between test 1 and test 2 correlated significantly in both the ITT and the PD GHRH test (P 0.02), and no significant difference was observed in the median peak response to repeated testing. In addition, no sex difference was SEVERAL STUDIES in groups of adult patients with hypothalamic-pituitary disorders and severe GH deficiency have demonstrated clear benefits from substitution therapy with recombinant human GH (1 4). However, the clinical characteristics of the GH deficiency syndrome can be vague in the individual patient (5). A single measurement of GH in serum is useless for diagnosing GH deficiency, and evaluation includes provocative testing (6, 7). The unspecific and variable responses in healthy adults to GH stimulation tests are well recognized (8 11), but other approaches to measure GH secretion capacity in adults have been disappointing (12 15). Received September 29, Revision received December 15, Accepted December 23, Address all correspondence and requests for reprints to: Hans C. Hoeck, M.D., Ph.D., Department of Endocrinology and Medicine, Aalborg Hospital, Reberbansgade, DK-9000 Aalborg, Denmark. hans.chr.hoeck@dadlnet.dk. * This work was supported by grants from Stiftstidendes Julelotteri, Det Obelske Familiefond, Kristen Tøfting og Dagmar Tøftings Fond, and Nordjyllands Lægekredsforenings Forskningsfond. observed. The coefficients of variation (CV) were 96% (ITT) and 45% (PD GHRH), but in the majority of patients low values were repeatedly low. The peak GH response was significantly higher during the PD GHRH test than during the ITT (P 0.008). In the 10 patients tested with the PD GHRH and CLO GHRH tests there was no significant difference in the peak GH response (P 0.398). When the test specific cut-off values were used, no significant difference in diagnostic outcome was observed between the various tests (P 0.3). In contrast, the diagnosis obtained with IGF-I differed significantly from all GH stimulation tests (P 0.03). Twenty (77%) and 22 (85%) patients were diagnosed to be GH deficient with the ITT and the PD GHRH test, respectively. Of the 14 patients with multiple pituitary failure ( 2 hormones affected), GH deficiency was present in more than 90% regardless of the type of stimulation test used. The IGF-I levels were only subnormal in 42% of the patients and did not correlate with the peak GH responses in any of the stimulation tests (P 0.05). Except for 1 patient all patients with subnormal IGF-I were GH deficient in all stimulation tests. It is concluded that in patients with hypothalamic-pituitary disease and a normal IGF-I level 2 stimulation tests should be performed to establish a diagnosis of GH deficiency. In patients with a subnormal IGF-I value a single GH stimulation test should be sufficient to confirm the presence of GH deficiency. (J Clin Endocrinol Metab 85: , 2000) The insulin tolerance test (ITT) has been shown to be useful in pituitary patients with multiple hormone deficiencies (12, 13, 16). However, the ITT may be unpleasant to the patient and is contraindicated in the presence of cardiovascular disease or seizure disorder. Alternative stimulation tests have been used in pediatric endocrinology for decades (17, 18), but some have proven to be less useful in adults (19, 20). In recent years new GH stimulation tests have been introduced (10, 13, 19, 21, 22). Some of the tests have fewer side-effects and are more potent stimulators of GH secretion in healthy adults compared to the ITT. However, uncertainty remains to what extent the same diagnosis is obtained with the different GH stimulation tests in the same patient. This was investigated in the present study with three different GH stimulation tests in an unselected group of patients with hypothalamic-pituitary disease. The diagnosis of GH deficiency was compared when using either a fixed cut-off value or a test-specific cut-off value calculated from test results in healthy adults (11). In addition, the reproducibility of the tests was investigated. 1467

2 1468 HOECK ET AL. JCE&M 2000 Vol 85 No 4 Subjects and Methods Twenty-six patients, 13 males and 13 females, with hypothalamicpituitary disorders and variable degrees of hypopituitarism (all receiving stable substitution therapy where appropriate, including substitution with gonadal hormones) were tested twice with the ITT and the pyridostigmine plus GHRH (PD GHRH) test. In addition, 10 of the patients (5 males and 5 females) were tested once with the clonidine plus GHRH (CLO GHRH) test. There was a minimum of 72 h between the tests. The patients were recruited consecutively at their routine visits in the department and finally by asking only females so as to obtain an equal number of males and females. None of the patients had received GH substitution therapy in childhood. The individual characteristics of the patients are listed in Table 1. Based on repeated test results in healthy adults with the ITT [8 males and 8 females; age, yr; body mass index (BMI), kg/m 2 ], the CLO GHRH test (same subjects as in the ITT) and the PD GHRH test (15 males and 16 females; age, yr; BMI, ), a 95% confidence range for normal GH response was calculated for each test (11, 23). As the healthy subjects were tested twice with all tests, the mean peak GH response of test 1 and test 2 in each subject was used for the calculation. All values were logarithmically transformed before calculation. The calculated lower 95% confidence range was used as the cut-off value for a normal stimulated GH response. In the ITT the cut-off value was 3.92 g/l, in the CLO GHRH test it was 19.0 g/l, and in the PD GHRH test it was 12.8 g/l. The patients and controls were not matched for age and BMI, as we investigated healthy adults to establish our own cut-off values for the different tests. Hence, the differences in these parameters occurred by chance. In both the ITT and PD GHRH test there was no difference in the males (P 0.05), but the female patients were older and had a higher BMI than the controls (P 0.05). The clinical studies were mainly performed in an ambulatory setting, i.e. the participants were asked not to eat after 2000 h the previous evening and to avoid major physical activity and shower bath before arrival. The participants were asked to be transported by car or bus. All subjects were tested in the supine position. An indwelling heparin-locked cannula was inserted in an antecubital vein. The subjects then rested for 30 min before initiating one of the test procedures. TABLE 1. Characteristics of patients Patient no. Sex (M/F) Age (yr) BMI (kg/m 2 ) The ITT The test procedure was the same in all patients, except that some subjects were tested under in-house conditions depending on the preference of the patient. This did not affect the results, as no significant difference was observed between the peak GH responses of the 2 tests in the 14 patients in whom 1 of the tests was performed in an ambulatory setting (P 0.79). Blood samples were drawn for the determination of glucose and GH before insulin (Insulin Actrapid Human, Novo Nordic, Copenhagen, Denmark; 0.15 IU/kg BW) was injected iv. Blood glucose was measured with Reflolux (Roche, Mannheim, Germany) with short intervals. When blood glucose had declined to 2.2 mmol/l or less, blood samples were collected for measurements of GH and glucose (glucose oxidase method). Thereafter, blood samples were drawn 30 and 60 min later for determination of GH and glucose. In all subjects a decline in blood glucose of 2.2 mmol/l or less (median, 1.7 mmol/l) was observed during the ITT with the standard dose of insulin. Five minutes after adequate hypoglycemia had been observed and blood samples had been taken, 50 g glucose in an aqueous solution were given orally over a period of 10 min to prevent late hypoglycemic episodes. In none of the patients were serious side-effects, necessitating the infusion of glucose, reported. The CLO GHRH test Before starting the test, blood samples for measurements of GH were collected, and 30 min later clonidine (Catapresan, Boehringer Ingelheim GmbH, Ingelheim, Germany; 300 g) was given orally with a glass of water. Blood samples for measurement of GH were collected 15, 30, 45, and 60 min later. Then GHRH-(1 29)-NH 2 (Groliberin, Pharmacia & Upjohn, Inc., Stockholm, Sweden; 1 g/kg BW) was injected iv, and blood samples for measurement of GH were drawn at 15, 30, 45, 60, 75, and 90 min after the injection of GHRH. The test was performed in 10 patients and only once, as the adverse effects to clonidine (extreme drowsiness, dryness of the eyes and mouth) were so unpleasant that further studies with this test was omitted. Cause of hypopituitarism Tumor size Treatment Hormone deficiency 1 M Gonadotrope adenoma Macroadenoma S/ G/A 2 M Suprasellar lymphoma S/X G/A/T 3 M Prolactinoma Macroadenoma / G/A/T 4 M Cranial trauma / G/A/T 5 M Null cell adenoma Macroadenoma S/X G/A 6 M Craniopharyngioma Macroadenoma S/ A/T 7 M Null cell adenoma Macroadenoma S/ 8 M Null cell adenoma Macroadenoma S/ A 9 M Null cell adenoma Macroadenoma S/ 10 M Gonadotrope adenoma Macroadenoma S/ G/A 11 M Medulloblastoma /X 12 M Prolactinoma Macroadenoma S/ G 13 M Medulloblastoma S/X 14 F Cushing s disease Microadenoma S/ 15 F Cushing s disease Microadenoma /X 16 F Chiasmatic glioma /X G/A 17 F Null cell adenoma Macroadenoma S/X G/A/T 18 F Null cell adenoma Macroadenoma S/X A 19 F Prolactinoma Macroadenoma S/ 20 F Suprasellar cyst S/ A/V 21 F Cushing s disease Microadenoma S/X 22 F Null cell adenoma Macroadenoma S/ G/A 23 F Cushing s disease Macroadenoma S/X G 24 F Craniopharyngioma Macroadenoma S/ G/A 25 F Cushing s disease Microadenoma /X G/T 26 F Prolactinoma Microadenoma S/ G/A/T S, Surgery; X, radiation therapy; G, gonadal insufficiency; A, adrenal insufficiency; T, thyroid insufficiency; V, vasopressin insufficiency.

3 DIAGNOSIS OF GH DEFICIENCY IN ADULTS 1469 The PD GHRH test Blood samples for measurement of GH were drawn. Thirty minutes later pyridostigmine (Mestinon, Roche; 120 mg) was given orally with a glass of water. Blood samples for measurement of GH were collected 15, 30, 45, and 60 min later. Then GHRH-(1 29)-NH 2 (Groliberin, Pharmacia & Upjohn, Inc.) or GHRH-(1 44) [Somatrel hgrf-(1 44), Ferring Pharmaceuticals Ltd., Copenhagen, Denmark; 1 g/kg BW] was injected iv, and blood samples for measurement of GH were drawn at 15, 30, 45, 60, 75, and 90 min after the injection of GHRH. In general, the PD GHRH test was well tolerated in all patients. Only mild gastrointestinal discomfort was reported in a few subjects. Glucose and GH (immunoradiometric assay from CIS-Bio International, Gif-sur-Yvette, France) were measured as previously described (23). Insulin-like growth factor I (IGF-I) was measured on a fasting blood sample using a noncompetitive immunoradiometric assay (Diagnostics Systems Laboratories, Inc., Webster, TX) as previously described (11). Informed written consent was obtained from all participants, and the study protocols were approved by the scientific ethical committee of Viborg and Nordjylland County. The statistical evaluation was carried out using nonparametric tests, including Spearman s rank correlation coefficient, Wilcoxon s ranking test for unpaired data, Fisher s exact test, and the 2 test. P 0.05 was considered to be statistically significant. The coefficient of variation was calculated from CV [( (t1 t2) 2 )/(2 n)]/mean, where t1 is the peak GH response in test 1, and t2 is the peak response in test 2. Results The individual peak GH responses to all tests and the serum IGF-I values are shown in Table 2. In both the ITT and PD GHRH tests there were no significant differences in the peak GH responses between tests 1 and 2 (P 0.07). In addition, no sex difference was observed in either of the tests (P 0.10). A significant correlation between the peak GH response in tests 1 and 2 was observed in both the ITT and the PD GHRH test (P 0.02). However, although the peak GH responses in the majority of patients were repeatedly low, the peak GH responses to repeated testing were highly variable in both the ITT and the PD GHRH test, with coefficients of variation of 96% and 45%, respectively. The mean peak response of test 1 and test 2 was significantly higher during the PD GHRH test than during the ITT (P 0.008). In contrast, there was no significant difference between the CLO GHRH test response and the PD GHRH response in the 10 patients evaluated by both tests (P 0.398). To evaluate the diagnostic outcome of using the different tests the test results were compared to the test-specific cut-off values obtained in healthy controls (11). This is shown in Fig. 1, in which the corresponding IGF-I values are presented as well. No significant difference in diagnostic outcome was found among the various stimulation tests (P 0.3). In contrast, the diagnosis obtained with IGF-I differed significantly from those obtained from the ITT and the PD GHRH test (P 0.006). This difference was also present when analyzing the 10 patients undertaking the CLO GHRH test separately (P 0.03). In the ITT 20 of the 26 patients (77%) had a peak GH response below 3.92 g/l in both tests. In the patients with affection of 1 or no other pituitary hormone a peak GH TABLE 2. Insulin-like growth factor I levels and peak GH responses in 26 patients after repeated stimulation with the insulin tolerance test and pyridostigmine and clonidine in combination with GHRH Patient no. Sex (M/F) IGF-I ITT 1 b ITT 2 c PD GHRH 1 (nmol/l) a mmol/l g/l mmol/l g/l ( g/l) PD GHRH 2 ( g/l) CLO GHRH ( g/l) 1 M d M d d M d M d M M d M d M d d M d M d M d M d d M F d d F d F d d F d d F d F d d F d F d d F d d F d d F d F d F d IGF-I, Insulin-like growth factor I; ITT 1, insulin tolerance test 1; ITT 2, insulin tolerance test 2; PD GHRH 1, pyridostigmine in combination with GHRH (test 1); PD GHRH 2, pyridostigmine in combination with GHRH (test 2); CLO GHRH, clonidine in combination with GHRH. a The values with bold type are below the normal range for age and sex. b Lowest detected blood glucose and corresponding peak GH response in test 1. c Lowest detected blood glucose and corresponding peak GH response in test 2. d Patients tested in an in-house setting.

4 1470 HOECK ET AL. JCE&M 2000 Vol 85 No 4 FIG. 1. The diagnosis of GH deficiency using test-specific peak GH values. Comparisons of GH tests and IGF-I. The serum IGF-I and peak GH responses from the mean of tests 1 and 2 in 26 patients [13 males (F) and 13 females (E)] during the ITT, the CLO GHRH test, and the PD GHRH test. A, Results in patients with only one or no other pituitary hormone affected; B, data from patients with multiple hormone deficiencies. The IGF-I values were logarithmically transformed and are shown as a percentage of the lower limit of the age-adjusted reference value. The peak GH values were logarithmically transformed and are presented as a percentage of the calculated lower 95% confidence limit from each test in the control subjects (11). The 95% confidence ranges are indicated by the bars and were calculated from the mean of the peak GH response in test 1 and test 2 in healthy adults after logarithmic transformation (11). response below 3.92 g/l was present in 50%. This increased to 93% in the patients with multiple hormone deficiencies ( 2 hormones affected). In the PD GHRH test (using 12.8 g/l as the cut-off value) the figures were 85%, 67%, and 100%, respectively. The peak GH responses during the CLO GHRH test were all below the calculated cut-off limit of 19.0 g/l for this test. The IGF-I level was subnormal in 42% of the patients (Table 2 and Fig. 1). No correlation was observed between the IGF-I levels and the stimulated peak GH responses in any of the tests (P 0.05). However, with a single exception, all subjects with a subnormal IGF-I had subnormal test responses with all GH stimulation tests (Fig. 1). In some of the patients the IGF-I levels were within the lower end of the

5 DIAGNOSIS OF GH DEFICIENCY IN ADULTS 1471 age-adjusted reference interval, although the patients had a repeatedly blunted stimulated GH response. No statistically significant difference was found between the IGF-I values in patients with no other hormone deficiency and those with multiple hormone defects. Discussion This study is the first to evaluate repeated testing with the ITT and the PD GHRH test in a larger unselected group of patients with hypothalamic-pituitary disease. The stimulated peak GH responses were variable in all tests, but clear differences were observed. The range of peak GH responses was wider during the PD GHRH test compared to the ITT, although the variability was lower. This observation could support the hypothesis that somatostatin tone has a major impact on the magnitude of the stimulated GH response (24). In general, the results obtained with the different tests indicate that GH secretion after injury to the hypothalamicpituitary region is not just either present or absent, but ranges from normal down to values around the detection limit of most GH assays. However, despite a high variability to repeated testing, low GH values were reproducible in both the ITT and the PD GHRH test, indicating the presence of prominent injury to the GH regulatory system in these patients. It is well recognized that adiposity impairs the magnitude of the GH response to provocative testing (25, 26). Sixteen of the 26 patients had a BMI above 25, which may bias the results, as the control group mainly consisted of lean subjects (11). Unfortunately, the CLO GHRH test was associated with unpleasant side-effects, limiting the use of this test in a clinical setting. The studies were performed with two different preparations of GHRH. Initially, GHRH-(1 29)-NH 2 was used. Based on previous studies the biological activity of human GHRH resides in the first 29 amino acids (27). The manufacturer ceased the production of GHRH-(1 29)-NH 2 during the study period, and we changed to GHRH-(1 44). As no significant differences have been observed between the two preparations on a weight basis, it is unlikely that this has introduced any significant bias to the results of the study (28, 29). Similar to the observations in healthy adults, significant differences in the peak GH response to the different tests was observed (11). The PD GHRH tests also provoked higher GH responses compared to the ITT. This difference was present when analyzing the 12 patients with deficiency of two or more pituitary hormones separately (test 1, P 0.011; test 2, P 0.001). Andersen et al. observed a similar significant difference in the peak GH response during the ITT and the PD GHRH test (10). However, Ghigo et al. found almost identical results with the same tests in pituitary patients (13). The controversy is readily explained by differences in the characteristics and the number of the patients studied. In contrast to the study by Andersen et al. and the present study, the majority of the 11 patients in the study by Ghigo et al. had deficiencies of 3 pituitary hormones (10, 13). In the present study the degree of impaired GH secretion correlated with the degree of hypopituitarism in the PD GHRH test. This observation is identical to findings with the ITT in a large group of pituitary patients (16). A cut-off limit of 3 g/l has been suggested for the ITT (7, 30), and a recent study observed no differences in the stimulated GH response in pituitary patients during the ITT and PD GHRH test (13). The present data demonstrate major diagnostic differences as a consequence of using a fixed cut-off value of 3 g/l in the three tests. In contrast, no systematic diagnostic differences between the tests employing the test-specific cut-off values calculated from repeated testing in healthy adults were observed. Three patients had a low normal GH response in the ITT, but were diagnosed GH deficient in the PD GHRH test. All of them had IGF-I levels in the normal range. In addition, one patient (patient 13) had a marked difference in the GH response, with a repeatedly subnormal response to the ITT and a low IGF-I, but a normal response to the PD GHRH test. This patient had been treated with radiotherapy for a hypothalamic medulloblastoma. This might indicate that either the hypothalamus or the connection between the hypothalamus and the pituitary had been damaged and that the patient s pituitary was intact. A direct pituitary stimulus using PD GHRH could then release GH. In contrast, hypoglycemia may exert its GH-releasing effects at the hypothalamic level, and no GH release would occur. In accordance with the observations by others (10, 12) the IGF-I levels did not correlate with the peak GH responses in any of the tests (ITT, P 0.05; PD GHRH, P 0.1). Nearly all patients with subnormal IGF-I had a blunted peak GH response regardless of the type of test used, whereas some patients with normal IGF-I had very low responses to all tests. No statistically significant difference was found between the IGF-I values in patients with no other hormone deficiency and multiple hormone defects. This is in sharp contrast to the observations with the GH stimulation tests, where the amount of GH secreted correlated with the degree of hypopituitarism in all tests. These findings may suggest that the GH stimulation tests are better estimators of GH reserve, compared to IGF-I, in patients with hypothalamicpituitary disorders. The results of the present study have practical clinical implications in the evaluation of patients with hypothalamicpituitary disease. In patients with subnormal IGF-I a single GH stimulation test seems sufficient to confirm the diagnosis. In patients with normal IGF-I it is suggested to perform two stimulation tests. Provided that test-specific cut-off values are used, the diagnostic value of the PD GHRH test is equal to that of the ITT. Acknowledgments The GHRH used in the studies was kindly provided by Pharmacia & Upjohn, Inc. (Groliberin), and Ferring Pharmaceuticals Ltd. A/S (Somatrel). We are grateful to Mrs. Ester Ditzel, Mrs. Anni Nielsen, Mrs. Marianne Køhler, and Mrs. Thea Kragh for their skillful assistance in conducting the experiments and in the analysis of the hormonal samples. References 1. Salomon F, Cuneo RC, Hesp R, Sonksen PH The effects of treatment with recombinant human growth hormone on body composition and metabolism in adults with growth hormone deficiency. N Engl J Med. 321:

6 1472 HOECK ET AL. JCE&M 2000 Vol 85 No 4 2. Jorgensen JO, Pedersen SA, Thuesen L, et al Beneficial effects of growth hormone treatment in GH-deficient adults. Lancet. 1: Janssen YJ, Doornbos J, Roelfsema F Changes in muscle volume, strength, and bioenergetics during recombinant human growth hormone (GH) therapy in adults with GH deficiency. J Clin Endocrinol Metab. 84: ter Maaten JC, de Boer H, Kamp O, Stuurman L, van der Veen EA Long-term effects of growth hormone (GH) replacement in men with childhood-onset GH deficiency. J Clin Endocrinol Metab. 84: Shalet SM, Toogood A, Rahim A, Brennan BM The diagnosis of growth hormone deficiency in children and adults. Endocr Rev. 19: Rose SR, Ross JL, Uriarte M, Barnes KM, Cassorla FG, Cutler GBJ The advantage of measuring stimulated as compared with spontaneous growth hormone levels in the diagnosis of growth hormone deficiency. N Engl J Med. 319: Anonymous Consensus guidelines for the diagnosis and treatment of adults with growth hormone deficiency: summary statement of the Growth Hormone Research Society Workshop on Adult Growth Hormone Deficiency. J Clin Endocrinol Metab. 83: Greenwood FC, Landon J, Stamp TC The plasma sugar, free fatty acid, cortisol, and growth hormone response to insulin. I. In control subjects. J Clin Invest. 45: Devesa J, Diaz MJ, Tresguerres JA, Arce V, Lima L Evidence that 2 -adrenergic pathways play a major role in growth hormone (GH) neuroregulation: 2-adrenergic agonism counteracts the inhibitory effect of muscarinic cholinergic receptor blockade on the GH response to GH-releasing hormone, while 2-adrenergic blockade diminishes the potentiating effect of increased cholinergic tone on such stimulation in normal men. J Clin Endocrinol Metab. 73: Andersen M, Hansen TB, Støving RK, et al The pyridostigmine-growthhormone-releasing-hormone test in adults. The reference interval and a comparison with the insulin tolerance test. Endocrinol Metab. 3: Hoeck HC, Jakobsen PE, Vestergaard P, Falhof J, Laurberg P Differences in reproducibility and peak growth hormone responses to repeated testing with various stimulators in healthy adults. Growth Horm IGF Res. 9: Hoffman DM, O Sullivan AJ, Baxter RC, Ho KK Diagnosis of growthhormone deficiency in adults. Lancet. 343: Ghigo E, Aimaretti G, Gianotti L, Bellone J, Arvat E, Camanni F New approach to the diagnosis of growth hormone deficiency in adults. Eur J Endocrinol. 134: Svensson J, Johannsson G, Bengtsson BA Insulin-like growth factor-i in growth hormone-deficient adults: relationship to population-based normal values, body composition and insulin tolerance test. Clin Endocrinol (Oxf). 46: Gill MS, Toogood AA, O Neill PA, Thorner MO, Shalet SM, Clayton PE Urinary growth hormone (GH), insulin-like growth factor I (IGF-I), and IGF-binding protein-3 measurements in the diagnosis of adult GH deficiency. J Clin Endocrinol Metab. 83: Toogood AA, Beardwell CG, Shalet SM The severity of growth hormone deficiency in adults with pituitary disease is related to the degree of hypopituitarism. Clin Endocrinol (Oxf). 41: Frasier SD A preview of growth hormone stimulation tests in children. Pediatrics. 53: Fraser NC, Seth J, Brown NS Clonidine is a better test for growth hormone deficiency than insulin hypoglycaemia. Arch Dis Child. 58: Rahim A, Toogood AA, Shalet SM The assessment of growth hormone status in normal young adult males using a variety of provocative agents. Clin Endocrinol (Oxf). 45: Vahl N, Jorgensen JO, Jurik AG, Christiansen JS Abdominal adiposity and physical fitness are major determinants of the age associated decline in stimulated GH secretion in healthy adults. J Clin Endocrinol Metab. 81: Aimaretti G, Corneli G, Razzore P, et al Comparison between insulininduced hypoglycemia and growth hormone (GH)-releasing hormone plus arginine as provocative tests for the diagnosis of GH deficiency in adults. J Clin Endocrinol Metab. 83: Micic D, Popovic V, Kendereski A, Peino R, Dieguez C, Casanueva FF The sequential administration of growth hormone-releasing hormone followed 120 minutes later by hexarelin, as an effective test to assess the pituitary GH reserve in man. Clin Endocrinol (Oxf). 45: Hoeck HC, Vestergaard P, Jakobsen PE, Laurberg P Test of growth hormone secretion in adults: poor reproducibility of the insulin tolerance test. Eur J Endocrinol. 133: Casanueva FF, Popovic V, Leal Cerro A, Alvarez CV, Zugaza JL, Dieguez C The physiology of growth hormone secretion. In: Melmed S, ed. Molecular and clinical advances in pituitary disorders. Los Angeles: Endocrine Research; Williams T, Berelowitz M, Joffe SN, et al Impaired growth hormone responses to growth hormone-releasing factor in obesity. A pituitary defect reversed with weight reduction. N Engl J Med. 311: Castro RC, Vieira JG, Chacra AR, Besser GM, Grossman AB, Lengyel AM Pyridostigmine enhances, but does not normalise, the GH response to GH-releasing hormone in obese subjects. Acta Endocrinol (Copenh). 122: Rivier J, Spiess J, Thorner M, Vale W Characterization of a growth hormone-releasing factor from a human pancreatic islet tumour. Nature. 300: Losa M, Schopohl J, Muller OA, von Werder K Stimulation of growth hormone secretion with human growth hormone releasing factors (GRF1 44, GRF1 40, GRF1 29) in normal subjects. Klin Wochenschr. 62: Laron Z Usefulness of the growth hormone-releasing hormone test regardless of which fragment is used (GHRH 1 44, 1 40 or 1 29). Isr J Med Sci. 27: Thorner MO, Bengtsson BA, Ho KY, et al The diagnosis of growth hormone deficiency (GHD) in adults [Letter]. J Clin Endocrinol Metab. 80: