Effect of Growth Hormone on Body Composition and Visceral ...

0021-972X/07/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 92(11):4265– 4270 Copyright © 2007 by The Endocrine Society doi: 10.1210/jc.2007-0786

Effect of Growth Hormone on Body Composition and Visceral Adiposity in Middle-Aged Men with Visceral Obesity Magdalena Pasarica, Jeffrey J. Zachwieja, Lilian DeJonge, Stephen Redman, and Steven R. Smith Pennington Biomedical Research Center (M.P., L.D., S.R., S.R.S.), Baton Rouge, Louisiana 70808; and Gatorade Sports Science Institute (J.J.Z.), Barrington, Illinois 60010 Rationale: GH replacement in GH-deficient adults results in an improvement in metabolic status. GH might also decrease visceral adiposity in obese adults that are not GH deficient. Objective: Our objective was to determine the effects of supraphysiological GH therapy on the metabolic syndrome and visceral adiposity in men with low blood levels of IGF-I and the durability of these effects after stopping GH therapy. Design: The study was a double-blind, placebo-controlled 6-month intervention trial followed by a blinded follow-up period of 6 months. Subjects: Thirty nondiabetic middle-aged men with central adiposity (body mass index ⬎ 27 kg/m2; waist circumference ⬎ 102 cm) participated.

baseline and placebo) and 8.8% reduction in visceral adiposity. GH increased resting energy expenditure by 172.5 ⫾ 41.6 kcal/24 h after 6 months of therapy. Fasting insulin, glucose, and the quantitative insulin sensitivity check index for insulin resistance increased during GH therapy. The effects of GH on fatness and visceral adiposity disappeared shortly after GH withdrawal, but weight remained increased over baseline and when compared with the placebo group (P ⬍ 0.05). Conclusion: These data suggest that GH therapy is associated with small but statistically significant decreases in visceral adiposity and an increase in lean mass and body weight. In viscerally obese subjects, supraphysiological GH administration is not an effective treatment; however, additional studies are needed to evaluate the effects of low-dose, physiological GH treatment. (J Clin Endocrinol Metab 92: 4265– 4270, 2007)

Results: After 6 months of GH therapy, we observed an increase in weight and lean body mass (2.5 ⫾ 0.6 kg, P ⬍ 0.05 compared with

O

BESITY IS RELATED to several metabolic disturbances such as insulin resistance, impaired insulin secretion, non-insulin-dependent diabetes mellitus, hypertension, dyslipidemia, and cardiovascular disease (1, 2). The metabolic risks associated with obesity are closely correlated with a central (abdominal), rather than a gluteal-femoral fat pattern. These complications of obesity have been attributed to increases in visceral adipose tissue (VAT) with an associated rise in portal vein free fatty acid levels (3). GH is a lipolytic hormone that decreases VAT, increases energy expenditure, and increases lean mass in individuals with GH deficiency. For example, Snel et al. (4) found changes in VAT that averaged 38%. GH therapy in GHdeficient adults also improves total and low-density lipoprotein cholesterol. Obesity is associated with a decrease in IGF-I and deficits in GH secretion. This is particularly apparent in individuals with visceral adiposity (5, 6). The mechanism by which obesity is associated with decreased GH secretion is unclear, although increases

First Published Online September 4, 2007 Abbreviations: BMI, Body mass index; CT, computed tomography; DEXA, dual-energy x-ray absorptiometry; DSAT, deep sc adipose tissue; REE, resting energy expenditure; SAT, sc abdominal adipose tissue; SSAT, superficial SAT; TAT, total abdominal adipose tissue; VAT, visceral abdominal adipose tissue. JCEM is published monthly by The Endocrine Society (http://www. endo-society.org), the foremost professional society serving the endocrine community.

in free fatty acids, a hallmark of visceral adiposity, are known to decrease GH secretion. Also, Frystyk et al. (7) demonstrated increased free IGF-I in obesity, whereas total IGF-I is consistently lower in obese men and women. Free IGF-I might serve to inhibit GH secretion through a negative feedback loop. Given the defects in GH secretion seen in visceral adiposity, Svensson et al. (8) tested the effectiveness of GH therapy in men with visceral obesity. They observed a striking reduction in VAT (⫺18%) along with an increase in lean mass (⫹2%). Furthermore, they demonstrated that GH decreases low-density lipoprotein cholesterol, without a change in high-density lipoprotein cholesterol. Importantly, GH therapy was associated with a transient increase in fasting insulin in men with visceral obesity, but long term (9-month) therapy increased insulin sensitivity compared with baseline, as measured by the euglycemic, hyperinsulinemic clamp (9). Based on these observations, we proposed the hypothesis that GH administration to achieve pharmacological levels would reduce visceral adiposity in nondiabetic middle-aged men with central adiposity. Second, we proposed that GH administration would exhibit a favorable change in fat oxidation, energy expenditure, fasting insulin, and blood pressure. Last, we followed these subjects for 6 months after GH therapy was withdrawn to determine the durability of the effects of GH on body composition.

4265

4266

J Clin Endocrinol Metab, November 2007, 92(11):4265– 4270

Subjects and Methods

Pasarica et al. • GH Therapy and Visceral Adiposity

Treatment was allocated based on baseline BMI and waist circumference by the statistician.

Population Thirty male subjects completed a comprehensive medical evaluation before participation in this study. All subjects were free of significant neurological, metabolic, endocrine, cardiac, respiratory, or gastrointestinal disease as evidenced by a normal physical examination and laboratory assessment. The characteristics of the population are presented in Table 1. Participants were selected on the following criteria: male, aged 40 –70 yr, central obesity defined as waist circumference greater than 102 cm and body mass index (BMI) more than 27 and less than 35 kg/m2, no weight loss in last 12 months, total IGF-I level less than 241 ng/ml (⬃25th percentile for the assay), and body habitus that permitted accurate computed tomography (CT) scan acquisition and analysis. Subjects were excluded if they had diabetes, known coronary heart disease, exercised more than 3 h/wk, or were unwilling or unable to abstain from alcohol for 72 h before the measurements of energy expenditure, fasting blood work, etc.

GH therapy

Informed consent

Body composition was measured by dual-energy x-ray absorptiometry (DEXA) using a Hologic QDR 2000 (Hologic, Waltham, MA). Lean body mass is calculated as body mass ⫺ (body fat ⫹ bone mineral content) and is presented in kilograms and as body fatness as a percentage of total body mass. Coefficients of variation for the measurement of lean body mass, fat mass, and percent fat are 0.8, 1.6, and 1.7%, respectively. CT images were acquired for each volunteer using a GE high-speed CT scanner (General Electric Medical System, Milwaukee, WI) as previously described (10). After removal of all metal clothing objects, the volunteer was placed in the supine position in the CT scanner with the arms over the head. CT images were acquired at the level of the interspace between the fourth and fifth lumbar vertebrae with a slice thickness of 10 mm at 140 kV and 340 mA. Seven additional slices were obtained at fixed 10-mm intervals above and below the L4 –5 level. Images were analyzed using an off-scanner image analysis program (AnalyzePC; Analyze Direct, Lenexa, KS). For each subject, an x-ray attenuation histogram was created for both adipose tissue and skeletal muscle (psoas). This histogram was then used to determine the attenuation value for adipose tissue for each individual scan as described by Kvist et al. (11). The midpoint between the mean attenuation value for adipose tissue and the mean attenuation value for skeletal muscle was used as the upper boundary for classifying pixels as adipose tissue or other soft tissue. Houndsfeld units for the upper and lower boundary averaged ⫺34 and ⫺190, respectively. Superficial sc adipose tissue (SSAT) and deep sc adipose tissue (DSAT) were separated by tracing the fascia superficialis with a mouse-driven cursor and adjusted by the reader. All images were measured by a single reader. Total adipose tissue (TAT) was defined as the sum of adipose tissue pixels inside a line tracing the skin. VAT was segmented by drawing a line that begins at the linea alba, bisects the rectus abdominus, the internal oblique, and the iliacus, and continues laterally around the peritoneum surrounding the vertebral body to join at the midline anterior to the vertebral body. All pixels inside this line that met criteria for adipose tissue x-ray attenuation were counted as intraabdominal adipose tissue. All pixels outside this line are classified as SAT (TAT ⫺ VAT ⫽ SAT). Adipose tissue pixels between the fascia superficialis and the skin were defined as SSAT. DSAT was defined as the total SAT pixels minus the superficial adipose tissue (SAT ⫺ SSAT ⫽ DSAT). The sum of pixels for each region were multiplied by the pixel size in square millimeters and divided by 100 to convert to areal measurements (square millimeters) in square centimeters. From the eight cross-sectional areas, VAT volume (liters) was calculated using the equations of Kvist et al. (11) and converted to kilograms of VAT mass using the conversion factor 0.9193 kg/liter adipose tissue. Body weight was measured on a calibrated scale. Heart rate and blood pressures were measured by certified research associates using standard Pennington protocols.

This experiment was performed under the authorization of U.S. Food and Drug Administration Investigational New Drug permit 58324 and with the approval of the Pennington Biomedical Research Center Institutional Review Board. All subjects signed a written informed consent document.

Study design Eligible subjects were enrolled into the 50-wk, randomized, placebocontrolled, double-blind study. Subjects were instructed by the physician at baseline to consume a healthy diet but were not given specific instructions to limit calories or change diet composition. Similarly, patients were instructed to keep their physical activity at baseline levels. Patients were informed of the nature of the treatment regimen and study design. Weight, waist circumference, and blood pressure were checked and adverse events queried at each clinic visit (weekly for wk 1– 4, every 2 wk until completion of the GH treatment phase, and every month thereafter). After a 12-h fasting period, blood draws and indirect calorimetry were performed.

TABLE 1. Characteristics of the study population Randomized to:

GH

Placebo

Age (yr) Height (cm) Weight (kg) BMI (kg/m2) Body fat (kg) Lean (kg) Fat (%) TAT (kg) VAT (kg) SAT (kg) SSAT (kg) DSAT (kg) Waist (cm) Waist/hip ratio Glucose (mg/dl) Insulin (␮U/ml) Triglycerides (mg/dl) Systolic BP (mm Hg) Diastolic BP (mm Hg) Heart rate (bpm) REE (kcal/24 h) REE adjusted (kcal/24 h䡠kg LBM) Respiratory quotient QUICKI (no units)

47.87 ⫾ 8.02 180.13 ⫾ 5.51 103.35 ⫾ 10.22 32.9 ⫾ 2.8 35.40 ⫾ 8.23 62.19 ⫾ 5.61 35.93 ⫾ 5.93 16.53 ⫾ 3.97 6.75 ⫾ 1.58 9.78 ⫾ 3.08 4.04 ⫾ 1.35 5.75 ⫾ 1.97 109.7 ⫾ 8.7 1.01 ⫾ 0.09 105.3 ⫾ 11.36 14.05 ⫾ 7.45 167 ⫾ 73.49 129.2 ⫾ 11.11 78.93 ⫾ 11.83 67.4 ⫾ 21.24 1921 ⫾ 841 30.9 ⫾ 1.9

50.33 ⫾ 6.98 178.12 ⫾ 5.31 105.77 ⫾ 13.05 32.3 ⫾ 2.6 35.2 ⫾ 8.55 64.106 ⫾ 7.06 35.11 ⫾ 5.34 16.34 ⫾ 4.06 6.42 ⫾ 1.86 9.92 ⫾ 2.65 4.03 ⫾ 1.00 5.89 ⫾ .179 110.3 ⫾ 8.9 0.98 ⫾ 0.06 104.73 ⫾ 10.68 20.45 ⫾ 21.21 163 ⫾ 74.75 139.2 ⫾ 11.41 82.4⫾10.06 66.73 ⫾ 16.19 1950 ⫾ 1355 30.3 ⫾ 2.2

0.83 ⫾ 0.18 0.32 ⫾ 0.02

0.84 ⫾ 0.17 0.32 ⫾ 0.03

Cell content results are shown as mean ⫾ SD. BP, Blood pressure; LBM, lean body mass.

GH therapy (recombinant GH, Neutropin; Genentech, South San Francisco, CA) was administered for a total of 24 wk followed by a 6-month washout period). GH dosage (0.95 mg/d) was calculated to achieve pharmacological (supraphysiological) levels of GH. Dosage was based on total body weight in kilograms: wk 1, 2.5 ␮g/kg䡠d every evening at bedtime; wk 2, 5.0 ␮g/kg䡠d every evening at bedtime; wk 3, 7.5 ␮g/kg䡠d every evening at bedtime; wk 4 –26, 10 ␮g/kg䡠d every evening at bedtime. Volunteers were instructed in self-administration of GH. Prefilled syringes were provided at a weekly or biweekly clinic, visit and used syringes were returned and counted.

Body composition

Analytical laboratory methods Blood samples were obtained via antecubital venipuncture. Glucose and triglycerides were analyzed on a Beckman Synchron CX7 (Brea, CA). Insulin was measured using a microparticle enzyme immunoassay on the Abbott IMX (Abbott Park, IL). IGF-I was measured using the Nichols



total IGF-I RIA assay (San Juan Capistrano, CA). Nichols RIA for IGF-I has an intraassay variance of 2.4% and interassay variance of 5.2%. These assays were conducted simultaneously.

Statistical analysis All analyses were performed using SAS, version 6.12 (SAS, Inc., Cary, NC). Data from one subject were excluded from analysis after wk 20 because he began fasting for religious reasons. The quantitative insulin sensitivity check index (QUICKI) was calculated from the fasting insulin and glucose values: QUICKI ⫽ 1/(log insulin ⫹ log glucose) as described by Katz et al. (12). All endpoints were analyzed as change from baseline in a mixed model (SAS Proc-Mixed). Treatment effects were determined with the entire dataset and at each time point (wk 8 or 12, 24, and 50). A P value ⬍ 0.05 was considered significant for all analyses.

Results

Summary data for the population before treatment are shown in Table 1. Twenty-eight subjects completed the protocol. Two subjects withdrew due to drug-related side effects Growth Hormone Placebo

after the 12-wk time point. One subject experienced profound malaise and peripheral edema that did not respond to dose reduction. A second volunteer withdrew due to symptoms of carpal tunnel syndrome and edema that resolved upon discontinuation of the drug. Blood levels of IGF-I increased in all subjects receiving GH, from 157 ⫾ 11 to 399 ⫾ 18 and 374 ⫾ 23 ng/ml at the 8- and 24-wk time points (P ⬍ 0.05 for time and treatment effects). IGF-I baseline levels are in the normal age-related reference range (61–285 ng/ml), whereas after treatment, IGF-I levels represented 3 ⫾ 1.2 and 2.5 ⫾ 1.6 sd score above the expected mean, as intended. IGF-I levels remained constant in the control patients: 157 ⫾ 11 vs. 172 ⫾ 15 and 172 ⫾ 19 ng/ml at the 8- and 24-wk time points, respectively. GH therapy resulted in an increase in body weight that was statistically significant from both baseline values and the placebo-treated subjects at multiple time points (Fig. 1). Body weight at month 12 was greater than baseline and was significantly greater in GH-treated patients compared with placebo-treated patients. During GH therapy, the increase in body weight (⫹2 kg) was due to an increase in lean body mass (Fig. 2). Importantly, lean body mass returned to baseline levels after a 6-month washout period. The decrease in

Washout

Treatment Period 4

2

#

#

# #

#

GH treatment

#

2.0 1.5

#

#

1

*

0

#

* * * *

*

* *

*

-1 -2

2

4

8

12

16

20

24

Weeks

7 8 9 10 11 12

% body fat Change from Baseline

Change from Baseline Body Weight (kg)

Dose escalation

washout

# #

#

3

4267

Months

1.0 0.5 0 -0.5 -1.0 -1.5 -2.0 -2.5 -3.0 -3.5

*

*

# # 0

12

24

50

week

4

#

Dose escalation

3

1

#

0 -1 4

8

4.0 3.5

2

2

Growth Hormone Placebo

#

12

16

20

24

7 8 9 10 11 12

-2

Lean Body mass (kg) Change from Baseline

5 Change from Baseline Waist circumference (cm)

Washout

Treatment Period

#

3.0 #

2.5 2.0

*

1.5

*

1.0 0.5 0 -0.5

Weeks

Months

FIG. 1. Body weight and waist circumference. Values are presented as least-squares means ⫾ SEM. #, Significant change from baseline (P ⬍ 0.05); *, significant difference from placebo at each time point (P ⬍ 0.05). GH or placebo treatment was applied until wk 24, at which time all intervention was stopped. This was followed by a 6-month follow-up (washout) period.

-1.0

0

12

24

50

week

FIG. 2. Body fatness (percent) and lean mass (kilograms). Values are presented as least-squares means ⫾ SEM. #, Significant change from baseline (P ⬍ 0.05); *, significant difference from placebo at each time point (P ⬍ 0.05).

4268



0.6

0.5 VAT mass

#

0.4

0.4

0.2

# kg

0 kg

#

0.3

0.2

-0.2

#

0.1

*

0

*

*

-0.1

-0.4

-0.2

-0.6 -0.8

DSAT mass

#

# 0

14

24

#

-0.3 -0.4

50

0

14

24

50

week

week

FIG. 3. Abdominal fat compartments. Values are presented as least-squares means ⫾ SEM. #, Significant change from baseline (P ⬍ 0.05); *, significant difference from placebo at each time point (P ⬍ 0.05). 0.8

1.5

SAT mass

0.6 #

0.4

0

kg

kg

0.2

*

-0.2

TAT mass

1.0

#

0.5

#

0

*

*

-0.5

-0.4

-0.8

-1.0

#

-0.6 0

14

24

50

-1.5

week

body fat (kilograms) during GH therapy at the end of GH therapy, as measured by DEXA, was not significant. The decrease in body fat (percent) was statistically significant at the 12- and 24-wk time points. During GH therapy, CT-measured VAT decreased by 0.5 ⫾ 0.13 and 0.6 ⫾ 0.13 kg at the 12- and 24-wk time points, respectively (P ⬍ 0.05) (Fig. 3). Small changes in SAT were observed at the 12-wk (P ⬍ 0.05) but not the 24-wk time points (P ⬎ 0.05). Waist circumference did not change during GH therapy (Fig. 1). At the final 1-yr observation, waist circumference was greater in the GH-treated subjects (P ⬍ 0.05 for change from baseline). Fasting glucose and insulin increased during GH therapy (P ⬍ 0.05, Table 2). Insulin sensitivity estimated by the QUICKI formula demonstrated a decrease in insulin sensitivity (P ⬍ 0.05). The effects on metabolic variables are shown in Table 2. After 8 wk, GH therapy increased resting metabolic rate by 221 ⫾ 36 kcal/24 h (P ⬍ 0.05) with no change in the respiratory quotient. After adjustment for the increases in lean body mass, the change in resting energy expenditure (REE) was significant from baseline but was no longer significantly different from the placebo group (P ⫽ 0.10). For all endpoints except weight and waist circumference, treatment effects were no longer apparent at wk 50 (data not shown for metabolic variables).

#

# 0

14

24

50

week

Discussion

The aim of this study was to test the hypothesis that GH therapy in non-GH-deficient adult men with visceral adiposity would result in a reduction in visceral adiposity without adverse effects on the metabolic syndrome. Parts of this hypothesis were borne out. However, the magnitude of the effect on adiposity was much less than anticipated from studies that examined replacement therapy in GH-deficient adults. GH therapy decreased visceral adiposity by only 0.6 ⫾ 0.13 kg. The direction of this effect was expected from earlier studies in GH-deficient individuals (4) and non-GH-deficient obese men (9). Other interventions that reduce body weight in non-GH-deficient obese men and women such as caloric deficit, very-low-calorie diet, or exercise produce greater reductions in visceral adiposity ranging from 9 – 49% (reviewed in Ref. 13). When compared with these other interventions and the cost of GH therapy, these results do not support the use of GH as a therapy in this population of visceral obese middle-aged men to decrease VAT. As expected, GH increased lean mass. The size of the effect (⬃3 kg) is significant and resulted in an increased REE. This is consistent with previous reports. It is also possible that some of the apparent increases in lean body mass are due to the known effects of GH on total body water. Respiratory quotient was not affected by GH therapy. The



TABLE 2. Treatment effects on metabolic variables during GH therapy wk 8 or 12

wk 24

Placebo Systolic BP (mm Hg) ⫺3.8 ⫾ 3.0 ⫺6.9 ⫾ 3.0a Diastolic BP (mm Hg) 0.13 ⫾ 2.6 ⫺2.5 ⫾ 2.6 Resting metabolic rate (kcal/24 h) 48.7 ⫾ 36.5b 32.1 ⫾ 38.1b REE adjusted (kcal/24 h䡠kg LBM) 0.6 ⫾ 0.5 0.6 ⫾ 0.5 Respiratory quotient (no units) ⫺0.020 ⫾ 0.01 ⫺0.004 ⫾ 0.02 Triglyceride (mg/dl) 35.2 ⫾ 24.2 8.4 ⫾ 17.93 Insulin (␮g/ml) ⫺7.0 ⫾ 4.2b ⫺6.9 ⫾ 5.4b b Glucose (mg/dl) ⫺6.8 ⫾ 2.6 ⫺7.3 ⫾ 2.2a,b b QUICKI (no units) 0.01 ⫾ 0.01 0.01 ⫾ 0.01b GH therapy Systolic BP (mm Hg) 2.69 ⫾ 3.06 ⫺0.1 ⫾ 3.2 Diastolic BP (mm Hg) ⫺0.47 ⫾ 2.70 0.56 ⫾ 2.8 Resting metabolic rate (kcal/24 h) 221.2 ⫾ 35.6a,b 172.5 ⫾ 41.5a,b REE adjusted kcal 1.8 ⫾ 0.5a 1.4 ⫾ 0.5a,c (kcal/24 h䡠kg LBM) Respiratory quotient (no units) ⫺0.02 ⫾ 0.01 0.017 ⫾ 0.017 Triglyceride (mg/dl) 26.8 ⫾ 24.2 10.74 ⫾ 18.8 Insulin (␮g/ml) 5.7 ⫾ 4.19b 13.0 ⫾ 5.7a,b a,b Glucose (mg/dl) 2.0 ⫾ 2.6 5.8 ⫾ 2.4b QUICKI (no units) ⫺0.02 ⫾ 0.01a,b ⫺0.02 ⫾ 0.01a,b Values are presented as least-square means ⫾ SE; n ⫽ 30 at wk 12 and n ⫽ 28 at wk 24 due to two dropouts in the GH arm. Body fat and blood pressure (BP) were measured at wk 12 and 24. All other measures were performed at wk 8 and 24. LBM, Lean body mass. a Significant differences from baseline. b Significant treatment effects at each time point. c P ⫽ 0.1.

increase in REE was due primarily to an increase in lean mass as demonstrated by the nonsignificant effect when REE is adjusted for lean body mass. The increase in metabolic rate was not accompanied by a decrease in body weight or total fat mass. In fact, body weight increased slightly. This suggests that there was a compensatory increase in food intake, a decrease in physical activity, or a decrease in nonexercise thermogenesis (i.e. fidgeting) (14). No data are available in this study to separate these possibilities. Overall abdominal adipose tissue (TAT) decreased during GH therapy, a finding in apparent conflict with the observed changes in waist circumference. This discrepancy is explained by two observations. First, although waist circumference is highly correlated with TAT and VAT, waist circumference also measures the mass of the viscera. CT measures of adipose tissue do not measure the viscera and are generally recognized as the gold standard for the measurement of regional adiposity. Second, it is generally recognized that waist circumference is a highly variable measure that can miss small changes in VAT. Indeed, the magnitude of the change in TAT and waist circumference illustrates this point. In addition to studying body composition, we determined the effect of GH supplementation on insulin sensitivity. However, this is a secondary endpoint of the study, so GH doses were not chosen to increase insulin sensitivity. The effect of GH therapy on insulin sensitivity seems to be dependent on length of treatment and the dose used (15). Yuen et al. (15, 16) used the homeostasis model assessment index and oral glucose tolerance test to show that lower than standard levels of GH determine a significant increase in insulin sensitivity in only a short time period (7 or 14 d). A similar

4269

dose of GH administrated for 6 months in obese patients showed a decrease in insulin sensitivity, as determined by hyperinsulinemic euglycemic clamp (17) or no change as determined by homeostasis model assessment index (18). Although none of the patients became overtly diabetic in our study, glucose and insulin levels rose as previously reported for this dose of GH. GH treatment also resulted in an increase in insulin resistance as estimated by the QUICKI calculation. In the study of Johannsson et al. (9), GH therapy at equivalent doses was associated with a transient increase in fasting insulin in men with visceral obesity, but long-term (9 months) therapy increased insulin sensitivity compared with baseline. The difference may be attributable to a greater reduction in VAT in their study. This may offset the negative effects of GH on insulin sensitivity. No significant changes in blood pressure were seen during GH therapy. Cessation of GH treatment on body composition seems to be dose dependent. Three months after cessation of low-dose GH treatment, Albert et al. (18) reported no change in body composition. In our study, most of the metabolic and body composition effects were short lived, and the treatment effects were no longer apparent after 6 months withdrawal. At the end of the protocol, body fatness, VAT, SAT, DSAT, SSAT, and TAT were equivalent to the placebo group. Body weight, however, never returned to baseline levels and was 2.5 kg higher in the GHtreated group 6 months after GH therapy was withdrawn. We believe that this finding has practical as well as statistical significance. Given the lesser precision of DEXA and CT when compared with the measurement of body weight, we are unable to parse out the observed increase in body weight. In other words, a change in body fat was not detected by DEXA. Waist circumference increased in both groups after withdrawal of therapy. The significant treatment effect on waist circumference at wk 50 is small, and a type II error cannot be ruled out. There were no treatment differences in CT-measured TAT or VAT at wk 50, suggesting that if the difference is indeed real; it most likely represents a combined increase in the lean body mass and adipose tissue. There are several reasons that might explain the differences between the magnitude change in VAT in our study (⬃9%) and that of Johannson et al. (9) (18%). First, the population was geographically and possibly genetically different from our population. Second, the Johannson protocol treated patients for 9 months. Our protocol treated patients for a 6-month period. We chose 6 months because their timecourse data demonstrated a plateau in VAT change at 6 months. Also, there are differences in the age and BMI between the studies. Our average weight was slightly higher (103.7 vs. 101.7 kg), and age was younger (47.9 vs. 58.3 yr). Last, the level of habitual physical activity was not stated in the Johannson paper, and we did not strictly control physical activity. As such, differences in habitual physical activity may account for the observed differences. In summary, GH therapy results in an increase in lean body mass and body weight and a small reduction in visceral adiposity in middle-aged men with visceral obesity. These changes are also associated with a decrease in insulin sensitivity as estimated by the QUICKI. Body weight increased during GH therapy and was higher in the GH group at the

4270


end of the 6-month washout period when compared with baseline and the placebo-treated group. The magnitude of the reduction in visceral adiposity with GH therapy is small, averaging 0.6 kg. If a state of chronic IGF-I excess is maintained as part of the weight reduction therapy, other unwanted side effects may occur including cartilage hypertrophy and cardiomyopathy. When compared with diet or exercise interventions, and considering the associated side effects and high cost, the use of GH therapy as a therapy in non-GH-deficient middle-aged men with visceral adiposity cannot be recommended. Acknowledgments We acknowledge the participation and dedication of the volunteers. Received April 10, 2007. Accepted August 29, 2007. Address all correspondence and requests for reprints to: Steven R. Smith, M.D., 6400 Perkins Road, Pennington Biomedical Research Center, Baton Rouge, Louisiana 70808. E-mail: [email protected]. This work was supported by the Genentech Foundation for Growth and Development (S.R.S.) as well as Genentech, Inc., which provided the GH, the U.S. Department of Agriculture (Grant 96034323-3031 to S.R.S. and L.D.), and the Clinical Nutrition Research Unit (Grant 1P30 DK072476). This trial has been submitted to clinicaltrials.gov. Study ID is NCT00453557. Disclosure Statement: The authors have nothing to disclose.

References 1. Vague J 1996 Sexual differentiation. A determinant factor of the forms of obesity. 1947. Obes Res 4:201–203 2. Vague J 1996 The degree of masculine differentiation of obesities: a factor determining predisposition to diabetes, atherosclerosis, gout, and uric calculous disease. 1956. Obes Res 4:204 –212 3. Bjorntorp P 1990 “Portal” adipose tissue as a generator of risk factors for cardiovascular disease and diabetes. Arteriosclerosis 10:493– 496 4. Snel YE, Brummer RJ, Doerga ME, Zelissen PM, Bakker CJ, Hendriks MJ, Koppeschaar HP 1995 Adipose tissue assessed by magnetic resonance imaging in growth hormone-deficient adults: the effect of growth hormone replacement and a comparison with control subjects. Am J Clin Nutr 61:1290 –1294


5. Veldhuis JD 1998 Neuroendocrine control of pulsatile growth hormone release in the human: relationship with gender. Growth Horm IGF Res 8(Suppl B):49 –59 6. Vahl N, Jorgensen JO, Skjaerbaek C, Veldhuis JD, Orskov H, Christiansen JS 1997 Abdominal adiposity rather than age and sex predicts mass and regularity of GH secretion in healthy adults. Am J Physiol 272:E1108 –1116 7. Frystyk J, Vestbo E, Skjaerbaek C, Mogensen CE, Orskov H 1995 Free insulin-like growth factors in human obesity. Metabolism 44:37– 44 8. Svensson J, Bengtsson BA, Taskinen MR, Wiklund O, Johannsson G 2000 A nine-month, placebo-controlled study of the effects of growth hormone treatment on lipoproteins and LDL size in abdominally obese men. Growth Horm IGF Res 10:118 –126 9. Johannsson G, Marin P, Lonn L, Ottosson M, Stenlof K, Bjorntorp P, Sjostrom L, Bengtsson BA 1997 Growth hormone treatment of abdominally obese men reduces abdominal fat mass, improves glucose and lipoprotein metabolism, and reduces diastolic blood pressure. J Clin Endocrinol Metab 82:727– 734 10. Smith SR, Lovejoy JC, Greenway F, Ryan D, Dejonge L, De La Bretonne J, Volafova J, Bray GA 2001 Contributions of total body fat, abdominal subcutaneous adipose tissue compartments, and visceral adipose tissue to the metabolic complications of obesity. Metabolism 50:425– 435 11. Kvist H, Sjostrom L, Tylen U 1986 Adipose tissue volume determinations in women by computed tomography: technical considerations. Int J Obes 10: 53– 67 12. Katz A, Nambi SS, Mather K, Baron AD, Follmann DA, Sullivan G, Quon MJ 2000 Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab 85:2402–2410 13. Smith SR, Zachwieja JJ 1999 Visceral adipose tissue: a critical review of intervention strategies. Int J Obes Relat Metab Disord 23:329 –335 14. Levine JA, Eberhardt NL, Jensen MD 1999 Role of nonexercise activity thermogenesis in resistance to fat gain in humans. Science 283:212–214 15. Yuen K, Ong K, Husbands S, Chatelain P, Fryklund L, Gluckman P, Ranke M, Cook D, Rosenfeld R, Wass J, Dunger D 2002 The effects of short-term administration of two low doses versus the standard GH replacement dose on insulin sensitivity and fasting glucose levels in young healthy adults. J Clin Endocrinol Metab 87:1989 –1995 16. Yuen K, Wareham N, Frystyk J, Hennings S, Mitchell J, Fryklund L, Dunger D 2004 Short-term low-dose growth hormone administration in subjects with impaired glucose tolerance and the metabolic syndrome: effects on ␤-cell function and post-load glucose tolerance. Eur J Endocrinol 151:39 – 45 17. Franco C, Brandberg J, Lonn L, Andersson B, Bengtsson BA, Johannsson G 2005 Growth hormone treatment reduces abdominal visceral fat in postmenopausal women with abdominal obesity: a 12-month placebo-controlled trial. J Clin Endocrinol Metab 90:1466 –1474 18. Albert SG, Mooradian AD 2004 Low-dose recombinant human growth hormone as adjuvant therapy to lifestyle modifications in the management of obesity. J Clin Endocrinol Metab 89:695–701

JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.