| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Clinical Research Centers |
Neuroendocrine Unit (C.H., C.C., S.P., S.G.) and Combined Program in Pediatric Gastroenterology and Nutrition (C.H.), Massachusetts General Hospital, and Infectious Disease Unit (W.R.), Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts 02114
Address correspondence and requests for reprints to: Steven Grinspoon, M.D., Neuroendocrine Unit, Bulfinch 457B, Massachusetts General Hospital, Boston, Massachusetts 02114. E-mail: sgrinspoon{at}partners.org
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Nine HIV-infected women with the lipodystrophy syndrome [age, 35.7 ± 1.7 yr; body mass index (BMI), 24.7 ± 0.8 kg/m2] were compared with 14 age- and BMI-matched HIV-infected women without significant lipodystrophy (age, 37.6 ± 1.1 yr; BMI, 23.4 ± 0.6 kg/m2) and 16 age and BMI-matched healthy eumenorrheic female control subjects (age, 35.8 ± 0.9 yr; BMI, 23.1 ± 0.4 kg/m2). All of the subjects underwent a standardized physical examination and anthropometric assessment. All lipodystrophic patients reported recent onset fat redistribution, defined as increased truncal fat or neck fat or loss of facial and/or extremity fat, which was confirmed at one or more sites by physical examination in all subjects by a single investigator (C.H.). In contrast, the nonlipodystrophic subjects did not report or demonstrate clinically evident fat redistribution. Eumenorrheic subjects gave a history of regular menstrual function over the 3 months before the study. Oligomenorrheic subjects missed one or more periods in the 3 months before the study. Menstrual function was also characterized as eumenorrheic, oligomenorrheic, or amenorrheic in the years before HIV infection. Subjects with a history of diabetes mellitus, acute opportunistic infection within the prior 4 weeks, hemoglobin less than 8 g/dL (80 g/L), creatinine greater than 3.0 mg/dL (270 µmol/L), or active alcohol or substance abuse were excluded from the study. Subjects were also excluded if they had changed antiviral medications within 6 weeks of the study, had a history of previous treatment with an antidiabetic agent, or reported use of testosterone, estrogen, GH, or other steroids in the past 6 months. Subjects were not excluded based on prior history of dyslipidemia. Subjects were recruited from the multidisciplinary HIV practice at the Massachusetts General Hospital and from advertisements and were studied in the early follicular phase if eumenorrheic. All subjects were evaluated between 1998 and 1999. Patients with HIV lipodystrophy were recruited through advertisements seeking HIV-infected women with evidence of fat redistribution or were referred by their physician for evaluation of observed changes in fat distribution. 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.
Materials and Methods
Subjects were evaluated in the General Clinical Research Center at the Massachusetts General Hospital and Massachusetts Institute of Technology. Blood samples were obtained at 0800 h in all subjects after an overnight fast. Subjects with regular menstrual function were studied in the early follicular phase. Glucose was measured with a hexokinase reagent kit (Dade Dimension, Wilmington, DE). Cholesterol was measured by enzymatic hydrolysis (Dade Dimension). Serum triglycerides were measured using a lipase enzymatic method and bichromatic determination of free glycerol (Dade Dimension). High-density lipoprotein (HDL) cholesterol was measured after precipitation of low-density lipoprotein (LDL) and very low-density lipoprotein with dextran-sulfate-magnesium (Dade Dimension). LDL cholesterol was calculated indirectly. Insulin levels were measured in serum using RIA (Diagnostic Products, Los Angeles, CA). Intra-assay coefficients of variation (CV) range from 4.7–7.7%, and interassay CV were between 5.5% and 9.2%. Cross-reactivity with proinsulin at midcurve was at least 40%. Testosterone was measured by RIA (Endocrine Sciences, Inc., Calabasas Hills, CA). The free testosterone concentration was determined as the product of the percent-free testosterone, measured by equilibrium dialysis, and the total testosterone concentration (Endocrine Sciences, Inc.). The intra-assay CV of free testosterone is 6.9%, and the intra-assay CV for total testosterone is less than 8.1%. The intra-assay CV were developed using pooled sera covering the range of the assay. Estradiol was measured by microparticle enzyme immunoassay (Abbott Laboratories, Abbott Park, IL). Dehydroepiandrosterone sulfate (DHEAS) was measured by chemiluminescent assay (Immulite DHEAS04; Diagnostics Products). Gonadotropin levels were measured by microparticle enzyme immunoassay (Abbot Laboratories). Viral load was determined by RT-PCR analysis using the Roche Amplicor HIV-1 Monitor Test (Roche Molecular Systems, Branchberg, NJ; detection limit, 400 copies/mL).
Height and weight were measured for each subject. Dual-energy x-ray absorptiometry (DEXA) was performed to determine total and regional lean body and fat mass using a Hologic-4500 densitometer (Hologic, Inc., Waltham, MA). Regions of interest, including arms, legs, and trunk, were standardized (1995 Users Guide; Hologic, Inc.). The percentage of body fat was calculated by dividing the weight of fat by total body weight (10). Similarly, the percentage of truncal fat and extremity fat were determined by dividing the weight of truncal fat and extremity fat by the total amount of body fat.
Statistical analysis
Comparison of clinical variables was made by two-tailed t test. Univariate regression analysis was performed, and a multivariate regression model was constructed to predict free testosterone, including sex hormone-binding globulin (SHBG), trunk fat to extremity fat ratio, and duration protease inhibitor (PI) use (JMP Statistical Discovery Software; SAS Institute, Inc., Cary, NC). A P value of 0.05 was used to test for statistical significance. Results are mean ± SEM.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
.
Eight of nine (89%) of lipodystrophic patients demonstrated clinical
evidence of increased abdominal fat and five of nine (56%)
demonstrated peripheral fat atrophy. All patients with fat atrophy
demonstrated simultaneous evidence of increased abdominal fat.
Fifty-six percent of lipodystrophic subjects and 57% of
nonlipodystrophic subjects were eumenorrheic. Minority status was not
different between the groups. Eighty-nine percent of the lipodystrophic
group vs. 75% of the control subjects had ever received a
PI with total duration of therapy 27.3 ± 6.0 vs.
11.7 ± 3.7 months (LIPO vs. NONLIPO, P
= 0.04). All of the patients in both groups had previously received a
nucleoside reverse transcriptase inhibitor (NRTI). Current NRTI use was
89% vs. 77% (LIPO vs. NONLIPO), with total
duration of therapy 46.2 ± 4.2 vs. 44.7 ± 6.7
months (P = 0.85). Weight was stable in the 2–3 months
before the study in all three groups of patients, and weight change was
not different between the groups (0.15 ± 1.04 vs.
0.12 ± 0.68 vs. 0.15 ± 0.41 kg HIV LIPO
vs. HIV NONLIPO vs. normal controls,
P > 0.05 for all comparisons).
|
Lipodystrophic subjects were significantly hyperandrogenic by
total and free testosterone levels relative to both the control
population and nonlipodystrophic patients, despite similar age and
weight (Table 1
and Figs. 1 and 2). Five of nine lipodystrophic patients
(56%) demonstrated a free testosterone level more than 2
SD above the mean in the healthy normal control population.
In contrast, androgen levels were relatively reduced in the
nonlipodystrophic patients compared with control subjects. SHBG levels
were comparable in the lipodystrophic and control populations but were
significantly increased in the nonlipodystrophic population. The LH/FSH
ratio was significantly increased in the lipodystrophic subjects
compared with both the nonlipodystrophic and control subjects.
Estradiol levels were not different between the groups. The free
testosterone level was higher among the oligomenorrheic lipodystrophic
patients [5.3 ± 1.9 vs. 3.8 ± 0.6 pg/mL
(18 ± 7 vs. 13 ± 2 pmol/L)], but this
difference did not reach statistical significance due to the small
sample size. No significant differences in either total or free
testosterone levels were seen comparing lipodystrophic patients with
fat atrophy vs. those without fat atrophy [5.3 ± 1.4
vs. 3.8 ± 1.2 pg/mL (18 ± 5 vs.
13 ± 4 pmol/L) for free testosterone, P =
0.45].
|
|
Insulin, glucose, and lipid levels
Insulin levels were significantly increased in the lipodystrophic
patients compared with both the nonlipodystrophic and normal control
subjects. Insulin levels were also increased in the nonlipodystrophic
subjects compared with control subjects and were highly correlated with
truncal adiposity (trunk to extremity ratio) among all the HIV-infected
patients (r = 0.70, P = 0.0003; Table 3). The
correlation between percent truncal fat and fasting insulin in the
control subjects was 0.39 (P = 0.13). Fasting glucose
levels were not different in any of the study groups. Total cholesterol
and LDL levels were increased in the lipodystrophic patients compared
with the nonlipodystrophic and control subjects, but were not different
in the nonlipodystrophic subjects compared with control subjects. In
contrast, triglyceride levels were significantly increased in the
lipodystrophic subjects compared with nonlipodystrophic and control
subjects and were also significantly increased in the nonlipodystrophic
subjects compared with control subjects. Total cholesterol levels were
inversely correlated with viral load (r = -0.62,
P = 0.003). Serum triglycerides were correlated with
insulin (r = 0.53, P = 0.01) and trunk to
extremity fat ratio (r = 0.43, P = 0.04). HDL was
inversely correlated with triglyceride (r = -0.42,
P = 0.04).
|
Total fat and percent fat determined by DEXA were not different
between the groups (Table 2). The ratios of trunk fat to extremity fat
and trunk fat to total fat determined by DEXA were increased
significantly among the lipodystrophic subjects, but not the
nonlipodystrophic subjects, relative to control subjects. Similarly,
the extremity to total fat ratio was significantly reduced in the
lipodystrophic, but not nonlipodystrophic subjects, compared with
control subjects.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The pathogenesis and clinical implications of increased testosterone levels in this population are unknown. We controlled for weight as a potential confounding factor contributing to increased androgen levels in the lipodystrophic patients, investigating normal weight lipodystrophic subjects matched to the control population. In contrast to the large, case-control study of Gervasoni et al. (8), in which no difference in total testosterone levels was seen between women with and without fat redistribution, we sampled patients in the follicular phase and investigated age- and BMI-matched healthy, eumenorrheic subjects as a reference group. We also included an HIV-infected nonlipodystrophic control population to determine the influence of HIV disease, independent of lipodystrophy, on testosterone levels and controlled for minority status, which was similar in all three groups.
Hyperandrogenemia in the HIV lipodystrophy syndrome may result from either increased ovarian and/or adrenal androgen production or altered testosterone clearance. In this regard, hyperinsulinemia may result in increased ovarian androgen production or reduced SHBG. Alternatively, antiretroviral therapy might have direct effects on testosterone metabolism or indirect effects related to fat redistribution. Indeed, the lipodystrophic patients demonstrated markedly increased truncal fat and reduced extremity fat by DEXA, compared both to the healthy control population and nonlipodystrophic HIV control subjects. These body composition data illustrate the extreme degree of fat redistribution in lipodystrophic patients.
One potential mechanism for increased androgen levels in this population relates to the increased truncal adiposity in women with the HIV lipodystrophy syndrome. Free testosterone levels, as measured by equilibrium dialysis, were positively correlated with the percent truncal fat and inversely correlated with SHBG. Increased SHBG levels have been previously shown among HIV-infected patients (24, 26). Our data also suggest that SHBG levels are increased in HIV disease, but decline toward normal in the lipodystrophy syndrome in association with increasing truncal adiposity. Of note, we limited our comparison to age- and BMI-matched normal weight subjects to avoid the influence of weight per se on SHBG. SHBG levels may be further reduced in overweight lipodystrophic patients. Importantly, we chose to measure free testosterone by equilibrium dialysis. This assay method measures free testosterone independent of SHBG and, therefore, provides a true estimate of bioavailable testosterone. DHEAS levels were not significantly increased in subjects with the lipodystrophy syndrome in comparison with healthy control subjects.
Another potential explanation for relative hyperandrogenemia in the HIV-infected subjects with lipodystrophy is an effect of antiretroviral therapy. In particular, we observed a significant correlation between duration of PI use and androgen levels. The relationship was robust and remained highly significant controlling for SHBG and trunk fat to extremity fat ratio. Although a number of studies now suggest that the lipodystrophy syndrome is not due exclusively to an effect of PI therapy (27), prolonged duration of PI therapy may contribute either directly or indirectly to changes in fat redistribution. These data do not prove causality with respect to PI use and hyperandrogenemia. In fact, prolonged PI use may simply be a marker for changes in fat redistribution, which then drive the changes in androgen concentrations. However, body composition changes fell out of the model when we included duration of PI use. Furthermore, duration of NRTI therapy was not associated with testosterone levels. Alternatively, PIs may have a direct effect on androgen metabolism. For example, Eagling et al. (28) demonstrated a potential effect of PIs on CYP3A4-mediated testosterone 6 ß hydroxylation. However, there was no significant difference in androgen levels by PI treatment status (i.e. receiving or not receiving PI therapy) in our study, suggesting there is no direct effect of PI use on androgen metabolism. Our sample size is small, and additional studies are needed to investigate the direct effect of antiretroviral medications on testosterone metabolism.
Insulin and lipid levels were increased in the lipodystrophic patients compared with nonlipodystrophic patients. The pattern of dyslipidemia is characterized by a significant increase in triglyceride and an associated decrease in HDL. Gervasoni et al. (8) previously reported increased triglycerides in HIV-infected women with fat redistribution, but increased cholesterol and LDL have not previously been reported in comparison with age- and BMI-matched control subjects. In this study, we found that increasing truncal adiposity was significantly associated with triglyceride, insulin, and testosterone levels. In contrast, androgen levels were not correlated with insulin among the HIV-infected subjects.
The insulin resistance, truncal adiposity, and dyslipidemia that we demonstrated in HIV-infected women with the lipodystrophy syndrome are similar in certain respects to those found in non HIV-infected women with the polycystic ovary syndrome. We noted a high prevalence of oligomenorrhea and a significantly increased LH/FSH ratio in association with increased free testosterone levels in the lipodystrophic subjects, consistent with a PCO biochemical pattern (29). However, we did not limit our investigation to amenorrheic subjects, but instead characterized patients based on fat redistribution. It is, therefore, possible that a more extreme pattern of biochemical dysfunction might be seen among amenorrheic lipodystrophic patients.
Our data suggest that a wide spectrum of metabolic abnormalities occurs in HIV-infected women with the lipodystrophy syndrome. Amenorrhea may be seen in association with hyperandrogenemia in some lipodystrophic women, whereas hyperandrogenemia may occur in the setting of normal menstrual function in other subjects. Whether this latter group is more likely to become amenorrheic over time is an important, but unanswered, question. Ovarian imaging was not performed, and, therefore, our study does not provide information on morphological changes that might be expected in a syndrome marked by extreme truncal adiposity and insulin resistance. None of the lipodystrophic patients complained of hirsutism, but subjects were intentionally chosen to be of normal weight, and it is, therefore, possible that a more severe phenotype would be expressed in overweight women with the HIV lipodystrophy syndrome. Furthermore, only one of the lipodystrophic subjects gave a history of irregular periods before the diagnosis of HIV, suggesting that preexisting PCO was not a likely explanation for the hyperandrogenemia in this group. Additional studies in a larger population of HIV-infected women with the lipodystrophy syndrome are necessary to investigate the associated metabolic and clinical phenotypic characteristics and to distinguish potential differences based on menstrual status.
This study has a number of limitations. First, the number of study subjects is small, and these data will need to be confirmed in larger studies of HIV-infected women with the lipodystrophy syndrome. In addition, the mechanism and clinical consequences of hyperandrogenemia remain unknown. We did not measure ovarian and adrenal androgen responses to provocative stimuli and/or standardized suppression testing to further investigate ovarian and adrenal androgen metabolism. Furthermore, it remains unknown whether hyperandrogenemia is in any way causal to the lipodystrophy syndrome or rather a result of it. More likely, the lipodystrophy syndrome is associated with metabolic and anthropometric abnormalities, such as truncal adiposity and hyperinsulinemia, that result in secondary hyperandrogenemia. For example, there is a strong continuous but inverse relationship between relative truncal adiposity and SHBG among HIV-infected patients. Serial evaluation of androgen levels over time in HIV-infected women who develop the lipodystrophy syndrome should help to elucidate the timing and causes of hyperandrogenemia in this population. All patients had evidence of fat redistribution confirmed by a single investigator (C.H.) in one or more body regions and testosterone levels were not different among patients with or without lipoatrophy. However, the definition of the HIV lipodystrophy syndrome is still evolving, and it is possible that further characterization of the syndrome may identify subtypes with varying effects on androgen levels.
We have previously shown significant androgen deficiency in low weight HIV-infected patients. The nonlipodystrophic subjects chosen for study were weight-matched to the lipodystrophic and control populations and not significantly androgen deficient. Use of androgens to increase weight and lean body mass may be beneficial in HIV-infected patients with weight loss, in whom androgen levels are deficient. In contrast, androgen administration may not be necessary or prudent among women with clinical lipodystrophy and increased androgen levels. The effects of androgen administration on insulin and regional body composition in HIV-infected women are not known. Further investigation of androgen use in this population is important to determine whether there are differential effects on insulin and body composition depending on the baseline androgen levels and the testosterone dose used.
HIV-infected women with the lipodystrophy syndrome demonstrate a number of metabolic and clinical features, including hyperinsulinemia, fat redistribution, and dyslipidemia. Our data suggest that hyperandrogenemia may also be present in such patients. Additional studies are necessary to investigate the mechanisms and clinical consequences of hyperandrogenemia in women with the HIV lipodystrophy syndrome.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received May 17, 2000.
Revised July 13, 2000.
Accepted July 14, 2000.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. B Albu, S. Kenya, Q. He, M. Wainwright, E. S Berk, S. Heshka, D. P Kotler, and E. S Engelson Independent associations of insulin resistance with high whole-body intermuscular and low leg subcutaneous adipose tissue distribution in obese HIV-infected women Am. J. Clinical Nutrition, July 1, 2007; 86(1): 100 - 106. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Desruisseaux, Nagajyothi, M. E. Trujillo, H. B. Tanowitz, and P. E. Scherer Adipocyte, Adipose Tissue, and Infectious Disease Infect. Immun., March 1, 2007; 75(3): 1066 - 1078. [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Lee, J. L. Chan, E. Sourlas, V. Raptopoulos, and C. S. Mantzoros Recombinant Methionyl Human Leptin Therapy in Replacement Doses Improves Insulin Resistance and Metabolic Profile in Patients with Lipoatrophy and Metabolic Syndrome Induced by the Highly Active Antiretroviral Therapy J. Clin. Endocrinol. Metab., July 1, 2006; 91(7): 2605 - 2611. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Johnsen, S. E. Dolan, K. V. Fitch, K. M. Killilea, J. L. Shifren, and S. K. Grinspoon Absence of Polycystic Ovary Syndrome Features in Human Immunodeficiency Virus-Infected Women Despite Significant Hyperinsulinemia and Truncal Adiposity J. Clin. Endocrinol. Metab., October 1, 2005; 90(10): 5596 - 5604. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. H. Dahan, L. N. Lyle, A. Wolfsen, and R J. Chang Tumor-Level Serum Testosterone Associated With Human Immunodeficiency Virus Lipodystrophy Syndrome Obstet. Gynecol., May 1, 2004; 103(5): 1094 - 1096. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. K. S. Leow, C. L. Addy, and C. S. Mantzoros Human Immunodeficiency Virus/Highly Active Antiretroviral Therapy-Associated Metabolic Syndrome: Clinical Presentation, Pathophysiology, and Therapeutic Strategies J. Clin. Endocrinol. Metab., May 1, 2003; 88(5): 1961 - 1976. [Full Text] [PDF]
|
|
|
|
|
|
J. C. Lo and M. Schambelan Reproductive Function in Human Immunodeficiency Virus Infection J. Clin. Endocrinol. Metab., June 1, 2001; 86(6): 2338 - 2343. [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 < 0.05 compared with
HIV-infected nonlipodystrophic women.