点击显示 收起
1 From the Translational Metabolism Unit and the Division of Diabetes, Endocrinology and Metabolism, Department of Medicine (SD, JS, RVS, KR, MM, and AB); the Department of Pediatrics and the US Department of Agriculture/Agricultural Research Service Children's Nutrition Research Center (FJ and KJE); and the Department of Radiology (JW), Baylor College of Medicine, Houston, TX; and the Endocrine Service, Ben Taub General Hospital, Houston, TX (RVS, MM, and AB)
2 SD and JS contributed equally to this work. 3 The contents of the manuscript do not necessarily reflect the views or the policies of the US Department of Agriculture, and mention of trade names, commercial products, and organizations does not imply endorsement by the US government. 4 Supported by an investigator-initiated research grant from Pharmacia Corporation and NIH grants RO1 DK59537 and HL73696 (to AB). In addition, much of the work was performed in the Baylor Children's Nutrition Research Center, which is supported by the US Department of Agriculture/Agricultural Research Service (USDA/ARS) under Cooperative Agreement no. 5862-5-01003. 5 Reprints not available. Address correspondence to A Balasubramanyam, Translational Metabolism Unit, Division of Diabetes, Endocrinology and Metabolism, Room 700B, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030-2600. E-mail: ashokb{at}bcm.tmc.edu.
ABSTRACT
Background: HIV lipodystrophy syndrome (HLS) is characterized by accelerated lipolysis, inadequate fat oxidation, increased hepatic reesterification, and a high frequency of growth hormone deficiency (GHD). The effect of growth hormone (GH) replacement on these lipid kinetic abnormalities is unknown.
Objective: We aimed to measure the effects of physiologic GH replacement on lipid kinetics in men with HLS and GHD.
Design: Seven men with HLS and GHD were studied with the use of infusions of [13C1]palmitate, [2H5]glycerol, and [2H3]leucine to quantify total and net lipolysis, palmitate and free fatty acid (FFA) oxidation, and VLDL apolipoprotein B-100 synthesis before and after 6 mo of GH replacement (maximum: 5 µg · kg–1 · d–1).
Results: GH replacement decreased the rates of total lipolysis [FFAtotal rate of appearance (
Key Words: Dyslipidemia • insulin resistance • fat redistribution • adipocyte
INTRODUCTION
HIV lipodystrophy syndrome (HLS) is characterized by variable changes in body morphology (peripheral lipoatrophy and visceral and regional adiposity) and profound metabolic abnormalities (hypertriacylglycerolemia, low HDL-cholesterol concentrations, and insulin resistance; 1) that are associated with increased cardiovascular disease risk (2–5). The urgent need for effective treatments for this heterogeneous condition is unfortunately offset by the relative ineffectiveness of conventional therapies aimed at improving the dyslipidemia and insulin resistance, perhaps because of the multiple underlying defects, which include exaggerated fasting lipolysis and inadequate oxidative disposal of the released fatty acids.
The growth hormone (GH) axis is also defective in patients with HLS. Rietschel et al (6) showed reductions in basal GH concentrations, mean overnight GH concentrations, and GH pulse amplitude in HLS patients compared with healthy controls and nonlipodystrophic HIV patients. As many as 20% of men with HLS with the phenotype of visceral obesity and peripheral fat loss fail to respond to a standardized GH stimulation test performed by using GH releasing hormone with arginine under stringent criteria (7). Several studies have assessed the effectiveness of high-dose GH therapy in correcting the abnormal fat redistribution associated with HLS (6, 8–12). In a prospective, open-label trial of 30 HLS patients receiving GH at doses of 6 mg/d or 4 mg every other day for 24 wk, visceral fat decreased significantly (11). Kotler et al (13) reported that 12 wk of GH treatment at a dose of 4 mg/d reduced visceral adipose mass and truncal fat significantly in a study of 245 HLS patients. However, the use of supraphysiologic doses of GH, especially in persons who are not necessarily GH deficient, is associated with high rates of adverse clinical effects, including glucose intolerance, arthralgias, myalgia, edema, diarrhea, and, less commonly, carpal tunnel syndrome and carcinoid tumors (8, 14). Evidence that more "physiologic" doses of GH provide clinical benefit is emerging, but the mechanisms are poorly understood. In a placebo-controlled study, Koutkia et al (15) observed improvements in lean body mass and the ratio of visceral to subcutaneous adipose tissue after treating HLS patients with GH releasing hormone to achieve physiologic serum concentrations of GH. A recent small study also showed that relatively low-dose (0.7 mg/d) GH treatment decreases the waist-to-hip ratio and increases HDL cholesterol (16).
We previously showed that accelerated total and net lipolysis with increased adipocyte reesterification and availability of free fatty acids (FFAs) for intrahepatic reesterification are key lipid metabolic abnormalities that underlie HIV lipodystrophy (17). Increased whole-body lipolysis in HLS has also been confirmed by other investigators (18, 19). Hence, the goal of the present study was to determine the effect of strictly physiologic normalization of GH on these specific lipid kinetic abnormalities in HLS patients with GH deficiency (GHD).
SUBJECTS AND METHODS
Subjects
The study was approved by the Institutional Review Board for Human Studies at Baylor College of Medicine. Seven men aged 45–50 y with HLS and GHD were recruited. HLS was defined by 3 criteria, as previously described (17): 1) fat loss in the extremities and increased abdominal girth, as observed by the patient and confirmed by his primary physician; 2) lipodystrophy score, which is based on morphologic abnormalities in each of 5 body regions, as assessed by a single investigator [A, abdominal obesity; B, "buffalo hump" (posterior cervical fat pad); C, supraclavicular fat pad; E, extremity fat loss; F, facial fat loss] and which used a 4-point intensity scale (0, no change; 1, mild change; 2, moderate change; 3, severe change; a score of 2 in 2 regions was required); and 3) fasting plasma triacylglycerol concentration >200 mg/dL (2.26 mmol/L). All HLS subjects had the mixed phenotype of HIV lipodystrophy (peripheral fat atrophy and central adiposity) as described by Saint-Marc et al (20). The men were free of diabetes mellitus, thyroid disorders, hypercortisolemia, liver or renal impairment, and hypogonadism and had had no opportunistic infections or illnesses for 5 y. All had sedentary lifestyles (exercising 2 times/wk), and none had consumed unusual diets or dietary supplements. The complete HIV lipodystrophy case definition score (21) could not be obtained because standardized computed tomography measurements of abdominal and peripheral fat were not performed.
GHD was diagnosed by the absence of a normal GH response to arginine (peak GH < 3.0 ng/mL) after a 10-h overnight fast and a fasting serum insulin-like growth factor I (IGF-I) concentration that was subnormal for age. For the stimulation test, GH and IGF-I were measured at baseline (0800), followed by an intravenous infusion of arginine (300 mL of a 10% solution) over 30 min, with GH assessment at 30, 60, 90, and 120 min. None of the subjects had previously received GH therapy or been treated with testosterone or other anabolic hormones, corticosteroids, antidiabetic agents, or drugs known to affect GH action. Lipid-lowering medications were discontinued 6 wk before the baseline stable-isotope infusion protocol. All subjects had been on a continuous, stable regimen of highly active antiretroviral drugs for 6 mo before the study, and these drugs were continued throughout the study.
Glucose tolerance and insulin sensitivity were assessed by an oral-glucose-tolerance test, which was performed after the men had fasted for 8 h after 3 d of ad libitum food intake. At 0800, the subjects consumed 75 g glucose (Glucola; Allegiance Healthcare Corp, McGaw Park, IL) orally over 3 min; blood was sampled just before glucose ingestion and at 30-min intervals for 2 h thereafter. Serum insulin and glucose concentrations were measured in each sample, and areas under the curve for glucose and insulin were calculated by the trapezoidal method. Insulin sensitivity was assessed by homeostasis model assessment (22) and the insulin sensitivity index (23). All subjects had normal glucose tolerance (as defined by American Diabetes Association criteria) before initiating growth hormone replacement.
GH treatment
Subjects self-administered recombinant human GH (Genotropin; Pharmacia, Kalamazoo, MI) by subcutaneous injection. GH treatment was initiated at a daily dose of 1 µg · kg–1 · d–1 in the evening and was titrated upward every week in 1-µg/kg increments to a maximum dose of 5 µg · kg–1 · d–1. Fasting serum IGF-1 concentrations were monitored during the dose titration protocol and were maintained within the normal range in all patients throughout the study.
Metabolic study protocol
The protocol consisted of intravenous infusions of stable isotopes to measure lipid kinetics and indirect calorimetry to measure substrate oxidation in the fasted state, before and after GH replacement. For 2 d before each metabolic study protocol, the subjects consumed a standard, balanced eucaloric diet supplied by the General Clinical Research Center kitchen that consisted of 22.5 kcal and 1 g protein · kg body wt–1 · d–1. The subjects fasted for 10 h before the start of the stable-isotope infusions.
The outcome variables were the rate of appearance (Ra) of glycerol, an index of the rate of total lipolysis; the Ra of palmitate, an index of net lipolysis; palmitate oxidation, an index of plasma fatty acid oxidation; the Ra of FFAs and FFA oxidation; fractional and absolute rates of synthesis of VLDL apolipoprotein (apo) B-100; and calculated rates of reesterification of fatty acids within the adipocyte and hepatic reesterification to triacylglycerol. The respiratory quotient and resting energy expenditure were also calculated.
At 0700, after baseline blood and breath samples were collected, a primed, constant intravenous infusion of [2H3]leucine (prime, 6 µmol/kg; infusion, 10 µmol · kg–1 · h–1) was started and maintained for 8 h. After 1 h, an intravenous prime of 5 µmol/kg of NaH13CO3 was given, followed by a 2-h primed infusion of [1-13C]sodium acetate (prime, 5 µmol/kg; infusion, 5 µmol · kg–1 · h–1). At the 5th hour of the infusion protocol, primed infusions of [1-13C]potassium palmitate (prime, 2.4 µmol/kg; infusion, 4.8 µmol · kg–1 · h–1) and [2H5]glycerol (prime, 4.5 µmol/kg; infusion, 9.0 µmol · kg–1 · h–1) were started and maintained for 3 h. Blood samples were collected hourly for the first 7 h and every 15 min during the final hour. Breath was collected every 15 min during the third and final hour of the infusion. Indirect calorimetry (Deltatrac, Sensormedics, Fullerton, CA) was performed for 30 min during the 6th hour.
Sample analyses
GH concentrations were measured by a two-site radioimmunometric assay (Corning Inc, Nichols Institute Diagnostics, San Juan Capistrano, CA). Serum IGF-1 assays were performed by chemiluminescent immunoassay (Arup Laboratories, Salt Lake City, UT). Plasma VLDL apo B-100 concentrations were measured by radial immunodiffusion by using the human apolipoprotein B "NL" Bindarid kit (The Binding Site Ltd, Birmingham, UK) after isolation from plasma by ultracentrifugation (40 000 x g for 30 min at 22 °C) (24). Plasma glucose concentrations were measured by the glucose oxidase method (YSI, Yellow Springs, OH), and plasma insulin by highly specific radioimmunoassay (Linco Research, St Charles, MO). Plasma FFA concentrations were measured by a spectrophotometric assay with reactions catalyzed by acyl-CoA synthase and acyl-CoA oxidase (Wako, Neusse, Germany).
Plasma palmitate and glycerol concentrations were determined by in vitro isotope dilution (25) with the use of [2,2-2H2]palmitate (98% 2H) and [2-13C]glycerol (99% 13C; Cambridge Isotope Laboratories, Andover, MA) as internal standards. The tracer-to-tracee ratios of plasma free palmitate were determined by negative chemical ionization gas chromatography–mass spectrometry (NCI-GC/MS) by using a Hewlett-Packard 5989B GC/MS system (Hewlett-Packard, Fullerton, CA) (26). The pentafluorobenzyl derivative was prepared and analyzed by selectively monitoring ions from mass-to-charge ratios (m/z) of 255 to 256. The plasma glycerol tracer-to-tracee ratio was measured by NCI-GC/MS on its heptafluorobutyric acid derivative, with selective monitoring of ions from m/z 680 to 685 (25). Breath 13CO2 content was determined by gas isotope ratio mass spectrometry on a Europa Tracermass Stable Isotope Analyzer (Europa Scientific, Crewe, UK).
VLDL was isolated from plasma by ultracentrifugation as described by Egusa et al (27), its apo B-l00 fraction was precipitated and isolated, and the isotopic enrichment of apo B-100–bound leucine was measured by gas chromatography–mass spectrometry as previously described (28). Briefly, the amino acids released from the protein by acid hydrolysis were purified by cation exchange chromatography and converted to the n-propyl ester, heptaflourobutyramide derivative, and the leucine isotope ratio was measured by monitoring ions at m/z 349 to 352 on a Hewlett-Packard 5988A Gas Chromatography Mass Spectrometer (Hewlett-Packard, Palo Alto, CA).
Calculations
The Ra of palmitate and the Ra of glycerol were calculated as follows:
RESULTS
HIV variables
All HLS subjects were receiving highly active antiretroviral therapy (Table 1). The RNA viral load was <400 copies/mL in 4 subjects and 400–2000 copies/mL in the others. At baseline, all men had normal plasma concentrations of thyroid-stimulating hormone, free thyroxine, testosterone, and hemoglobin and normal indexes of renal and liver function.
View this table:
TABLE 1. Baseline HIV characteristics of the male study subjects1
Growth hormone
At baseline, the subjects had a mean fasting plasma GH concentration of 0.35 ± 0.17 ng/mL. All subjects had a peak stimulated plasma GH concentration <3 ng/mL (0.81 ± 0.30 ng/mL) after arginine stimulation. The baseline IGF-1 concentration was subnormal in all subjects (
View this table:
TABLE 2. Serum biochemistry variables before and after growth hormone replacement therapy1
Lipid kinetics
GH replacement therapy in these HLS patients resulted in a significant decrease in the plasma Ra of FFAs due to total lipolysis (Ra FFAtotal: from 4.80 ± 1.24 to 3.32 ± 0.76 mmol FFA · kg fat–1 · h–1, P < 0.05; Ra glycerol: from 1.60 ± 0.41 to 1.11 ± 0.25 mmol glycerol · kg fat–1 · h–1, P < 0.05; Table 3). There was no significant change in the rate of intraadipocyte reesterification after GH therapy (from 2.93 ± 0.93 to 2.12 ± 0.54 mmol FFA · kg fat–1 · h–1; NS), but the rate of net lipolysis decreased significantly with treatment (Ra FFAnet: from 1.87 ± 0.34 to 1.20 ± 0.25 mmol FFA · kg fat–1 · h–1, P < 0.01; Ra palmitate: from 0.81 ± 0.16 to 0.56 ± 0.08 mmol palmitate · kg fat–1 · h–1, P < 0.01). The oxidation of plasma fatty acids also decreased significantly after GH replacement (from 0.28 ± 0.02 to 0.20 ± 0.02 mmol FFA · kg LBM–1 · h–1, P < 0.002); despite this, however, the rate of intrahepatic reesterification diminished after GH replacement therapy (from 0.50 ± 0.13 to 0.29 ± 0.09 mmol FFA · kg fat–1 · h–1, P < 0.03). This apparent discrepancy is resolved when these 2 parameters are expressed as a percentage of net lipolysis (76.1 ± 3.0% compared with 78.1 ± 2.6% for plasma fatty acid oxidation and 23.9 ± 3.0% compared with 21.9 ± 2.6% for intrahepatic reesterification). Expressed in this manner, the rates of plasma fatty acid oxidation and intrahepatic reesterification were identical before and after treatment, which suggests that their parallel decline after GH therapy was a direct result of decreased net lipolysis.
View this table:
TABLE 3. Lipid variables before and after growth hormone replacement therapy1
VLDL apo B-100
There were no significant differences before and after GH therapy in the rates of fractional or absolute synthesis of VLDL apo B-100 or in the plasma concentration of VLDL apo B-100 (Table 3).
Lipid and glycemic profiles
All subjects had normal glucose tolerance at baseline. There were no significant differences between pretreatment and posttreatment fasting plasma concentrations of total cholesterol, LDL cholesterol, HDL cholesterol, triacylglycerols, or insulin. There were no significant differences in glycated hemoglobin concentrations before and after GH replacement, but the fasting plasma glucose concentration increased significantly after GH treatment (5.2 ± 0.2 at baseline compared with 5.8 ± 0.3 mmol/L after 6 mo of GH replacement, P < 0.01; Table 2). Numerous indexes of insulin sensitivity based on simultaneous plasma glucose and insulin concentrations, either in the fasting state or during an oral-glucose-tolerance test [homeostasis model assessment of insulin resistance (HOMA-IR), the area under the curve of glucose divided by that of insulin, and the Insulin Sensitivity Index of Matsuda and Defronzo (22, 23)], showed no significant differences between mean values at baseline and after treatment (Table 2). However, there was a nonsignificant rise in the mean fasting plasma insulin concentration and in HOMA, largely due to the contribution of one patient who developed frank diabetes after GH treatment. This patient also had a severely elevated fasting plasma triacylglycerol concentration (6.328 mmol/L) at the end of the study, which caused a non-Gaussian skew of the triacylglycerol results. When this patient's terminal value was removed from the analysis, the mean plasma triacylglycerol concentration showed a trend toward a decline from baseline (2.46 ± 0.20 to 1.87 ± 0.27 mmol/L; P = 0.08).
Body composition
Body-composition measurements of total weight, fat mass, fat-free mass, and regional fat mass in the legs and arms (peripheral fat) or trunk and pelvis (central fat) did not change significantly in these subjects after GH replacement (Table 4). However, there was a trend toward decreased fat mass in the central compartment (319-g fat loss) and increased fat mass in the peripheral compartment (364-g fat gain) as assessed by the regional DXA measurements. Lipodystrophy score assessment (17) showed that at baseline, all patients had peripheral and facial fat loss and abdominal obesity, and 6 had abnormalities in every region (Table 5). Six of the 7 patients experienced improvements in the lipodystrophy score after treatment.
View this table:
TABLE 4. Body-composition variables before and after growth hormone replacement therapy1
View this table:
TABLE 5. Lipodystrophy scale before and after growth hormone replacement therapy1
DISCUSSION
We found that in patients with the "mixed" form of HLS and GHD, long-term, physiologic-dose GH replacement improves the underlying lipid kinetic defects through marked reductions in total and net lipolysis and the availability of fatty acids for hepatic reesterification. The results suggest that GHD contributes to the complex mechanism of adipocyte defects in HLS. The unequivocal decrease in the lipolytic rates differs from the traditional view of GH as a pro-lipolytic hormone.
GH has been used to treat several conditions, including childhood or adult GH deficiency, central obesity in persons without GHD, HIV-associated wasting, and HLS. Previous studies in HLS patients showed beneficial effects of GH treatment in regard to visceral fat (8, 11–13, 34). A supraphysiologic dose of GH (6 mg/d) for 24 wk resulted in 55% decrease in the mass of omental adipose tissue, and discontinuation of GH reversed this effect (12). These effects of supraphysiologic GH on body composition have been corroborated in patients with both HLS (6, 11, 13) and HIV-associated wasting (14, 35–37).
Although supraphysiologic GH therapy alters body fat distribution, limited data are available on its effects on abnormal lipid kinetics in HLS patients. Schwartz et al (38) reported increased total lipolysis 1 mo and 6 mo after supraphysiologic GH treatment (3 mg/d). We are not aware of prior studies investigating the effects of strictly physiologic GH replacement on lipid kinetics in HLS patients with GHD. Physiologic replacement doses of GH have been used in non-HIV-infected adults with GHD, but the kinetic data are limited. Two studies using physiologic GH (3.3 µg · kg–1 · d–1) reported increased total (38) or net (39) lipolysis in the fasted state. However, these studies evaluated lipid kinetics after very brief GH treatment (12 h or 1 wk) and did not address the critical issue of its long-term effects. Low-dose GH (2.5 and 3.3 µg · kg–1 · d–1) increased total lipolysis in adults with visceral obesity, but these subjects did not have biochemically defined GHD, and the results may not reflect the effects of long-term physiologic GH replacement on lipolytic rates in GHD patients (40). GH at a dose of 9.6 µg · kg–1 · d–1 (twice the dose used in the present study) resulted in a 60% increase in lipid oxidation in GHD subjects (41). Taken together, the results of these disparate studies stand in contrast with the present data obtained in HLS patients with GHD treated for 6 mo with a physiologic dose of GH.
Several explanations are possible for these novel results. First, all the HLS subjects in the current study had unequivocal GHD by standard criteria, whereas those in the previous studies were not necessarily GH deficient. Second, we used physiologic, replacement doses of GH, whereas the previous studies of HIV patients used markedly supraphysiologic doses. [Even studies investigating the effects of "low-dose" GH therapy in HLS patients used doses that are 2.5–30 times as high as the average replacement dose for GHD (38, 42)]. Third, the observations that lipolytic rates increase in non-HIV GHD patients treated with physiologic GH replacement were made after single infusions (12 h) or very-short-term (7 d) treatment with GH, and it was not shown that this effect persists over the long term (39, 43). The effects of GH replacement on substrate kinetics may vary widely depending on duration of treatment. Our earlier study evaluating the effects of physiologic GH replacement on protein turnover in non-HLS GHD patients at different times showed an initial decrease in protein oxidation with net protein retention in the fed state after 2 wk, but these changes were not sustained and reverted to pretreatment values after 6 mo (44). Finally, in vitro studies suggest that exposure of GH-deprived adipocytes to GH results in decreased, not increased, lipolysis (45).
The decrease in lipolysis suggests an effect of GH action at the level of the adipocyte. GH inhibits 11ß-hydroxysteroid dehydrogenase type 1 (11ß-HSD1) in adipose tissue. 11ß-HSD1 converts inactive cortisone to active cortisol; hence, its inhibition would lead to a lower ratio of cortisol to cortisone. Low-dose GH therapy decreased the plasma ratio of cortisol to cortisone in one study (46). The low dose of GH used in the present study may have caused a reduction in adipose tissue cortisol concentrations, leading to decreased lipolytic rates. In vitro studies also suggest that GH can inhibit adipocyte lipolysis through direct effects on hormone-sensitive lipase. GH rapidly inhibits both basal and catecholamine-stimulated lipolysis in GH-deprived rat adipocytes, an action mediated by blocking the activity of hormone-sensitive lipase; the mechanism was a reversal by GH of norepinephrine-stimulated phosphorylation of hormone-sensitive lipase (45). These data are consistent with our results and suggest that restoring strictly physiologic concentrations of GH may have effects on the hyperlipolytic adipocytes of HLS patients that are substantially different from those of very high doses of GH.
The other lipid kinetic changes could be simply downstream consequences of diminished fatty acid efflux. Plasma fatty acid oxidation declined in parallel with the decrease in plasma FFA release, but when expressed as a percentage of the net FFAs released, plasma fatty acid oxidation was the same before and after GH treatment. These data support the notion that lipid oxidation is largely regulated by substrate flux and that physiologic GH treatment does not alter lipid oxidation in HLS patients. There was also a decline in the calculated value of intrahepatic reesterification after treatment; despite this, there was no change in hepatic VLDL synthesis, as measured by the absolute and fractional synthetic rates of VLDL apo B-100. There are at least 2 potential reasons for this discrepancy. First, it is possible that a direct action of protease-inhibitor drugs inhibited the proteasomal degradation of apo B proteins (47); thus, apo B-100 synthesis might remain elevated even if the hepatic flux of FFAs declined. Second, the rate of apo B-100 synthesis, unlike the rate of VLDL triacylglycerol synthesis (19, 48), only indirectly measures the rate of synthesis and export of triacylglycerols by the liver; an unchanged rate of apo B-100 synthesis could still be associated with changes in the triacylglycerol content of the VLDL molecule.
Lack of a decrease in hepatic VLDL synthesis could account in part for the lack of improvement in fasting triacylglycerol concentrations after treatment. Another likely reason is that GH treatment may not affect impaired chylomicron clearance, an additional factor that contributes significantly to hypertriacylglycerolemia in HLS patients (48). A third reason could be the development of GH-mediated insulin resistance. Even with physiologic replacement doses, all the subjects experienced increased fasting glucose concentrations. One subject developed diabetes, with deterioration of fasting plasma triacylglycerol concentrations. When this subject was removed from the analysis, the group mean fasting triacylglycerol concentration showed a strong trend toward improvement compared with pretreatment levels. Hence, physiologic GH treatment may alter insulin sensitivity to varying degrees in HLS patients, and this may determine whether and to what extent the decrease in net lipolysis is matched by an improved lipid profile. A larger study is needed to assess other possible factors (eg, age, family history, degree of lipoatrophy, or visceral adiposity) that may affect plasma lipid concentrations after physiologic GH treatment.
In the present study, total fat and lean body mass did not change significantly with GH replacement. DXA showed decreased fat in the trunk and pelvic regions and increased fat in both limbs, which suggests a trend toward normalization of the fat redistribution in these patients. This observation was supported by a significant decrease in the waist-hip ratio, which suggests that GH replacement led to some decrease in visceral fat mass. The sample size was probably too small to detect significant changes in body composition, but these results are consistent with those observed in HLS patients after GH releasing hormone therapy to maintain normal plasma concentrations of IGF-I (15). The clinical lipodystrophy score also indicated improvements in some patients.
A limitation of this study is the absence of a control group receiving placebo. Nevertheless, the changes after prolonged treatment strongly suggest that physiologic GH replacement in HLS patients and in GHD has salutary effects on one of the fundamental lipid kinetic defects in HLS. In conjunction with the results of other studies using low-dose GH (49) or achieving physiologic concentrations of IGF-I (15), this suggests that physiologic GH therapy may improve some aspects of abnormal fat metabolism in HLS patients. Such therapy on its own does not alter lipid levels significantly, perhaps because fat oxidation declines pari passu with decreased lipolysis. Combining physiologic GH replacement with interventions that increase lipid oxidation may improve plasma lipid concentrations, fat redistribution, and insulin resistance in patients with HIV lipodystrophy.
ACKNOWLEDGMENTS
We thank Melanie Del Rosario for assistance in laboratory analyses; Elizabeth M Frazer, Shaji Chacko, and Daniel Donaldson for expert technical advice in mass spectrometry; and Dina Harleaux, Lynne Scott, and the nursing, pharmacy, and dietary staffs of the Baylor General Clinical Research Center for excellent care of subjects and meticulous attention to protocol.
SD was responsible for execution of the clinical protocols. JS was responsible for data collection and laboratory analysis and contributed to data analysis. RVS contributed to the study design and was primarily responsible for data analysis and interpretation and for writing the manuscript. FJ and AB designed the study and were responsible for interpretation of the data and drafting the manuscript. KR and MM recruited the subjects and were responsible for coordinating the clinical and treatment aspects of the study. KJE and JW performed the body-composition scans. None of the authors had a personal or financial conflict of interest.
REFERENCES