| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Massachusetts General Hospital Program in Nutritional Metabolism (P.K., B.C., S.G.) and Neuroendocrine Unit, Harvard Medical School, Boston, Massachusetts 02114; and General Clinical Research Center (J.B.), Massachusetts Institute of Technology, Cambridge, Massachusetts 02139
Address all correspondence and requests for reprints to: Steven Grinspoon, M.D., Program in Nutritional Metabolism, 55 Fruit Street, LON 207, Massachusetts General Hospital, Boston, Massachusetts 02114. E-mail: sgrinspoon{at}partners.org.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The diagnosis of GH deficiency (GHD) in adults can be difficult because of the lack of a specific cutoff to establish the condition. Furthermore, GHD may be partial. Most often, the diagnosis of true GHD in the adult is made using the results from standardized stimulation testing in the context of a pituitary mass or prior insult, such as radiation therapy. However, increasingly, it is becoming recognized that nutritional status can affect GH secretion, and reduced GH secretion has been shown in obesity (5) and in relationship to excess visceral fat in HIV-infected patients (4). Whether such patients have true GHD or functional reduction in GH secretion and whether they might benefit from GH replacement remain controversial.
In this study, we investigated GH response rates to standardized GHRH-arginine stimulation testing among HIV patients with (LIPO) and without (NONLIPO)lipodystrophy, in comparison with C subjects of similar age and body mass index (BMI).
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Inclusion and exclusion criteria. Seventy-two male subjects [HIV-infected (n = 56) and healthy C subjects (n = 16) similar in age and BMI] were enrolled in the study between October 2001 and December 2003. HIV-infected subjects were characterized at study entry as LIPO (n = 39) and NONLIPO (n = 17), based on evidence of fat redistribution as previously described. All HIV-infected subjects were required to be on a stable antiretroviral regimen for at least 6 wk before entrance into the study. Subjects were categorized as LIPO based on a waist to hip ratio (WHR) of at least 0.90 and evidence of fat redistribution, including increased fat under the chin, at the back of the neck, or in the abdominal, chest, or breast areas or decreased fat in the arms, legs, or face. A score for fat change (between 0 and 2) was graded for each site, and subjects with an increased WHR and lipodystrophy score of at least 2.0 were characterized as LIPO. HIV-infected subjects with a lipodystrophy score of less than 2.0 (without significant fat redistribution) were characterized as NONLIPO patients. Similar criteria were used in prior studies from our group to identify LIPO patients (4). Subjects with diabetes mellitus, BMI less than 20 kg/m2, hemoglobin less than 9 g/dl, and use of GH, GHRH, oral, or parenteral glucocorticoids, megesterol acetate, or antidiabetic agents within the 3 months before study initiation were excluded. Subjects with symptoms of pituitary disease, known history of pituitary disease, or history of radiation were excluded.
Written informed consent was obtained from each subject before testing, in accordance with the Committee on the Use of Humans as Experimental Subjects of the Massachusetts Institute of Technology and the Subcommittee on Human Studies at the Massachusetts General Hospital. Subjects underwent GHRH and arginine stimulation testing in the morning after a 12-h fast.
Screening outpatient visit. After a 12-h overnight fast, subjects reported to the General Clinical Research Center for a screening visit, at which time eligibility was determined based on a fasting blood glucose level less than 126 mg/dl.
GH assessment
Subjects received standardized stimulation testing with GHRH-arginine [GHRH 1–29 (Geref, Serono, Inc., Norwell, MA) 1-µg/kg iv bolus along with simultaneous administration of arginine hydrochloride (0.5 g/kg) (maximum dose, 30 g) given iv over 30 min]. GH levels were collected at –15, 0, 15, 30, 45, 60, 90, and 120 min after GHRH administration.
Body composition analysis
Height, weight, and BMI were determined in the fasting state during the screen visit. Cross-sectional abdominal computed tomography (CT) scanning was performed to assess the distribution of sc and visceral abdominal fat. A lateral scout image was obtained to identify the level of the L4 pedicle, which served as a landmark for the single-slice image. Scan parameters for each image were standardized (144 cm table height, 80 kV, 70 mA, 2 sec, 1-cm slice thickness). Fat attenuation coefficients were at –50 Hansfield units as described by Borkan et al. Abdominal visceral (VAT) and sc fat areas (SAT) were then determined (6). Total fat and extremity fat were determined by dual-energy x-ray absorptiometry (DXA) (Hologic 4500, Hologic Inc., Waltham, MA) as previously described (7).
Biochemical and immunological function
Fasting glucose, insulin, IGF-I, IGF binding protein (IGFBP)-1, IGFBP-3, free fatty acids (FFAs), triglyceride, testosterone, TSH, CD4+ cell count, and viral load were determined after a 12-h fast before any stimulation testing. A 75-g oral glucose tolerance test was performed for glucose and insulin on a separate occasion after a 12-h fast.
Laboratory methods
GH was measured by two-site RIA with an intraassay coefficient of variation (CV) of 4.4% (Corning, Inc. Nichols Institute Diagnostics, San Juan Capistrano, CA). The interassay CV was 6.6%. The sensitivity of the assay was determined to be 0.05 ng/ml. IGF-I was measured by two-site RIA with an intraassay CV of 4.9% (Diagnostic Systems Laboratories, Inc., Webster, TX). The interassay CV was 5.1%. The sensitivity of the assay was determined to be 2.6 ng/ml. IGFBP-1 was measured by two-site RIA with sensitivity of 0.33 ng/ml and an intraassay CV of 4.2% (Diagnostic Systems Laboratories, Inc.). IGFBP-3 was measured by two-site RIA with sensitivity of 0.5 ng/ml and an intraassay CV of 2.9% (Diagnostic Systems Laboratories, Inc.) Serum TSH levels were run using a two-site RIA (DiaSorin, Inc., Stillwater, MN; intraassay CV average, 3.1–3.3%; sensitivity, 0.01 µIU/ml based on serial dilutions).
Nonesterified fatty acid concentrations were measured by an in vitro enzymatic colorimetric assay kit (Wako Chemicals USA, Inc., Richmond, VA). The intraassay CV for fatty acids ranged from 1.1–2.7%. The published normal range for fatty acids is 0.1–0.6 mmol/liter. Insulin concentrations were measured in serum by RIA (Diagnostic Products, Los Angeles). The intraassay and interassay CVs ranged from 4.7–7.7% and from 5.5–9.2%, respectively. Glucose and triglyceride concentrations were measured by standard techniques. Testosterone was measured by RIA (Diagnostic Products). The intraassay CV ranged from 5–11%, and the interassay CV ranged from 5.9–12%. The sensitivity for the assay was 4 ng/dl.
The CD4 count was determined by flow cytometry (Becton Dickinson Immunocytochemistry Systems, San Jose, CA), and the HIV viral load was determined by ultrasensitive assay (Amplicor HIV-1 Monitor Assay; Roche Molecular Systems, Indianapolis, IN), with limits of detection of 50–75,000 copies/ml.
Statistical analysis
Demographic, body composition, and biochemical indices were compared among the three groups by overall ANOVA for continuous variables and likelihood ratio for dichotomized variables. In stepwise regression modeling, peak GH response and GH area under the curve (AUC) were the dependent variables; and age, BMI, VAT, SAT, insulin, triglycerides, and FFA concentrations were tested for entry into the model as dependent variables at P = 0.10. Among the HIV-infected subjects, group (LIPO or NONLIPO) was also tested in stepwise regression modeling. Final P values and parameter estimates are shown for each model. All statistical analyses were made using SAS JMP Statistical Database Software (version 4; SAS Institute, Inc., Cary, NC). Statistical significance was defined as a two-tailed 
-value of P 
0.05. Results are mean ± SEM unless otherwise indicated.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Thirty-nine male HIV-infected LIPO patients and 17 NONLIPO patients were compared with sixteen healthy, HIV-negative male C subjects. The clinical characteristics of the study patients are shown in Table 1
. The mean ages and BMI’s were similar among the three groups (Table 1). The LIPO subjects demonstrated higher WHR, visceral abdominal fat by CT, VAT:SAT ratio and trunk:lower extremity fat ratio (by DXA) than either the NONLIPO or C group. In contrast, lower extremity fat was reduced in the LIPO group (Table 1).
|
). Testosterone and TSH levels were not different among the groups and were normal in all subjects. FFA concentrations were higher in the LIPO group, but this difference did not reach statistical significance. IGF-I (P < 0.05, LIPO vs. NONLIPO) and IGFBP-3 were lowest in the LIPO group (P < 0.05, LIPO vs. NONLIPO and LIPO vs. C). CD4 count, viral load, and antiretroviral use were similar between the HIV groups (Table 1).
The rate of failure in response to the combined GHRH-arginine test was 18% in the LIPO group vs. 5.9% in the NONLIPO group and 0% in the C group, using a stringent cutoff of 3.3 ng/ml (P < 0.05 for LIPO vs. C). Using less stringent cutoffs, the failure rates in the LIPO group rose to 30.8% at 5.0 ng/ml and 38.5% at 7.5 ng/ml, with corresponding increases in the C group to 12.5% (P < 0.05 vs. NONLIPO at 5.0 ng/ml and P < 0.05 vs. NONLIPO and C at 7.5 ng/ml) (Fig. 1). At 9.0 ng/ml, 43.6% of the LIPO group vs. 17.7% and 25% of the NONLIPO and C groups, respectively, failed the stimulation testing (P = not significant).
|
). Among the subjects in the LIPO group, visceral fat area (parameter estimate = –0.09, P = 0.008), basal FFA level (parameter estimate = –18.7, P = 0.04), and basal insulin level (parameter estimate = –0.33, P = 0.007) were significant inverse predictors of peak GH response to GHRH and arginine (Table 2). Among subjects in the C group, only sc abdominal fat predicted peak GH response to combined GHRH-arginine in regression modeling (parameter estimate = –0.24, P = 0.009) (Table 2). Generally, similar results were seen in modeling for GH AUC (Table 2).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The optimal testing paradigm and cutoff for inadequate GH response are controversial. Insulin testing to evoke hypoglycemia is traditionally regarded as the gold standard test to determine GHD but was not appropriate in our subjects because of the potential risk associated with this test. The insulin tolerance test (ITT) is useful in patients with hypothalamic and/or pituitary disease and evaluates the entire hypothalamic-somatotroph axis. However, the test is contraindicated in patients with hypoglycemia, seizures, and ischemic heart disease. It is unknown whether the GHRH-arginine is equivalent to the ITT in patients with HIV. Because the GHRH-arginine test has less risk and has been shown to be reliable in other populations to detect GHD (8), it was chosen for our study.
The GHRH-arginine test compares well with the insulin test in diagnosing GHD (8) and has been considered a reasonable alternative test. The GH response to GHRH-arginine is independent of age, and there is less inter- and intraindividual variability than with other stimulation tests. The combined GHRH-arginine test is considered by many to be the best diagnostic alternative to the ITT (9). The lowest limit of GH response to ITT in normal subjects has been reported as 5 ng/ml by some (but not other) authors (10, 11). The peak GH cutpoint of 9 ng/ml for the GHRH-arginine tests represents the first centile limit of the normal GH response in a large population of normal subjects (11). Furthermore, Aimaretti et al. (12) showed that 92.5% of hypopituitary adults had a GH peak below this limit.
However, because the optimal cutoff for inadequate GH response is unknown, we chose a highly stringent cutoff of 3.3 ng/ml, because this value has been previously shown to demonstrate 95% specificity, with respect to the ITT test, for diagnosing GHD (8) and to minimize the misclassification of subjects as GH deficient when this is not the case. A caveat in this regard is that the 3.3 ng/ml cutoff was previously determined in comparison with the ITT in patients with known pituitary disease, in contrast to the patients described in this study, with nutritional disturbances in GH response to GHRH-arginine. Nonetheless, there were significant differences between the LIPO and C groups using this cutoff, with 0% of the C subjects failing the test compared with 18% of the subjects in the LIPO group. With increasing cutoffs, as have been proposed by others in the literature (11, 12), e.g. to 9.0 ng/ml, rates of failure to GHRH-arginine increased as anticipated (up to 44%), but a relatively large percentage of C subjects also failed the test (25%). Significant differences were not seen between the groups at the 9.0 ng/ml cutoff. Further studies are necessary to determine the best threshold to define reduced GH response rates to the GHRH-arginine test in the HIV population; but our data, using a strict threshold of 3.3 ng/ml, demonstrate that approximately one in five such LIPO subjects will fail to stimulate adequately.
The mean serum IGF-I was significantly lower among subjects in the LIPO group. However, IGF-I was below the normal range in only 8% of the LIPO subjects compared with 0% the NONLIPO subjects or C subjects (P = not significant). IGF-I was not correlated with peak response to GHRH-arginine stimulation testing in any of the groups. IGF-I is a relatively poor test to diagnose GHD in adults and has been shown to be in the normal range, even among patients who fail ITT testing (13, 14, 15). In contrast, IGF-I might be more useful for the diagnosis of GHD in childhood-onset or young-adult-onset GHD patients (16). However, because different IGF-I assays may yield different results, and serum IGF-I concentrations may be decreased by a variety of causes, caution is generally used in applying low IGF-I diagnostic cutpoints to the diagnosis of adult GHD in clinical practice (17). Furthermore, in contrast to GH resistance in undernutrition, prior studies have shown that GH secretion and IGF-I can be dissociated in simple obesity and have suggested an increased sensitivity to GH in overnutrition and obesity (18). Relatively increased sensitivity to GH may also occur in HIV patients with visceral adiposity and normal weight, although this has not been formally investigated in the HIV population.
IGFBP-3 was also significantly lower, but in the normal range, in the HIV lipodystrophy group. IGFBP-3 is N-glycosylated and is the major IGF binding moiety in plasma, given that more than 75 percent of circulating IGF-I is bound to this protein. IGFBP-3 concentrations are increased by GH administration, and IGFBP-3 functions as the major GH-responsive carrier of IGF-I (19). Of the six IGFBPs, IGFBP-3 is the major serum carrier protein for IGF peptides and the most GH dependent (20). IGFBP-3 levels are less nutritionally dependent, and the normal range varies only modestly with age. In our study, IGFBP-3 levels were significantly lower in the LIPO group in comparison with the other C groups. In contrast to some reports among non-HIV-infected patients (21, 22), IGFBP-3 concentrations did not correlate with peak GH response to GHRH-arginine in the LIPO subjects. Although lower IGF-I and IGFBP-3 levels support the observation of reduced GH secretion in the LIPO group, neither test is specific enough to use for diagnostic purposes in this population.
GH secretion is suppressed in obesity, with 24-h GH levels in obese men reduced by 75% in comparison with age-matched normal-weight C subjects (5). In addition, decreased responsiveness to stimulation tests such as GHRH, arginine, and combined GHRH-arginine has been demonstrated in non-HIV-infected subjects with obesity (23). Veldhuis et al. (24) demonstrated that the degree of body fatness was associated inversely with indices of GH secretion. In contrast, in our study, visceral fat, not BMI, was the major determinant of GH responsiveness in LIPO subjects. Similar results were seen among the entire HIV group, controlling for group (e.g. LIPO or NONLIPO). In contrast, only sc abdominal fat predicted peak GH response to GHRH and arginine in the C group.
The rates of failure in response to GH stimulation testing in the LIPO group were not a simple function of obesity, because response rates differed in comparison to a C group of healthy subjects similar in BMI. Furthermore, the inadequate response was not due to HIV per se, because lower failure rates were seen in age- and BMI-matched NONLIPO subjects. Importantly, our results can only be generalized to the patient population we recruited, with excess visceral adiposity and simultaneous loss of sc and extremity fat. It is less likely that any differences in GH response rates would be seen among a population with more pure lipoatrophy, because total body fat and extremity fat did not correlate with peak GH response among patients in HIV patients.
What are the mechanisms of reduced GH in the LIPO population? In prior studies, we have shown a strong inverse association between overnight GH secretion rates and visceral adiposity in frequent sampling. We now extend this finding to show a strong inverse relationship between visceral adiposity and GH response to GHRH-arginine. In addition, elevated FFA and insulin levels were significant independent variables in the regression models for peak GH response to GHRH-arginine in LIPO subjects.
We have also shown decreased ghrelin as well as an inhibitory effect of increased FFA on GH response to GHRH alone (25). We now show that peak GH response to GHRH-arginine is also inversely associated with FFA concentration. FFA remained negatively associated with peak GH response to GHRH-arginine in stepwise regression analysis controlling for age, BMI, and visceral fat, suggesting a potential inhibitory role with respect to GH secretion in the LIPO population. In healthy, obese subjects, Cordido et al. (26) demonstrated improved GH secretion in response to acipimox and reduced FFA. We also recently showed improved GH response to GHRH alone using acipimox, suggesting that FFA may play an inhibitory role with respect to GH secretion in the HIV lipodystrophy population (25).
Are HIV patients who fail the GHRH-arginine test truly or functionally GH deficient? Subjects in this study did not have any risk factors for, or symptoms of, pituitary dysfunction, and testosterone levels were normal in all patients. Although we did not perform pituitary imaging, it is unlikely that pituitary disease per se would explain our findings of inadequate GH response rates in up to nearly 20% of LIPO patients. More likely, our subjects have a functional GHD related, in part, to excess visceral fat. Whether this represents true GHD is not clear, because recent studies suggest that GH levels can change significantly over time, with reduction in weight and changes in body composition in otherwise healthy subjects (27, 28). This study was limited to men, to control for the influence of gender on GH secretion. Further studies determining GH response in LIPO women are needed.
A number of prior studies have investigated GH in HIV patients, both with the wasting syndrome (29) and with LIPO subjects (30, 31). Most prior studies in LIPO patients have investigated relatively large, supraphysiological doses of GH. Reported side effects with supraphysiological GH dosing in the HIV population include impaired glucose tolerance and joint aches. Recent studies suggest that such effects may be minimized with lower dosing (32). In contrast, Johannsson et al. (33) reported on the use of low-dose GH in non-HIV-infected patients with abdominal obesity, suggesting that GH administration might improve insulin resistance and reduce visceral adiposity in this population. Similarly, physiological GH dosing may result in improved insulin resistance and abdominal fat but has not been investigated in the HIV population. Furthermore, prior studies have not targeted a specific subpopulation of HIV patients with abnormal GH responses to GH stimulation testing.
In conclusion, our study suggests that approximately 20% of HIV-infected patients with lipodystrophy and increased visceral adiposity will have inadequate GH stimulation to GHRH-arginine using a stringent cutoff of 3.3 ng/ml and may therefore be functionally GH deficient. Such patients may benefit from GH administration in terms of improved body fat and metabolic parameters, and further studies are necessary to test the efficacy of low-dose GH in this population. Use of a less stringent cutoff may increase the number of patients failing the GH stimulation test but decrease the specificity of the test. Further studies are needed to optimally identify HIV patients with decreased GH secretion and to assess whether such patients might benefit from physiologic GH replacement.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online October 13, 2004
Abbreviations: AUC, Area under the curve; BMI, body mass index; C, control (group); CT, computed tomography; CV, coefficient of variation; DXA, dual-energy x-ray absorptiometry; FFA, free fatty acid; GHD, GH deficiency; IGFBP, IGF binding protein; ITT, insulin tolerance test; LIPO, HIV patients with lipodystrophy (i.e. with fat redistribution); NONLIPO, HIV patients without lipodystrophy; SAT, sc adipose tissue; VAT, visceral adipose tissue; WHR, waist to hip ratio.
Received July 15, 2004.
Accepted September 23, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. D'Amico, J. Shi, R. V Sekhar, F. Jahoor, K. J Ellis, K. Rehman, J. Willis, M. Maldonado, and A. Balasubramanyam Physiologic growth hormone replacement improves fasting lipid kinetics in patients with HIV lipodystrophy syndrome Am. J. Clinical Nutrition, July 1, 2006; 84(1): 204 - 211. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Koutkia, B. Canavan, J. Breu, and S. Grinspoon Effects of Growth Hormone-Releasing Hormone on Bone Turnover in Human Immunodeficiency Virus-Infected Men with Fat Accumulation J. Clin. Endocrinol. Metab., April 1, 2005; 90(4): 2154 - 2160. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
, P