Skip Navigation

Reduced Testosterone Levels in Human Immunodeficiency Virus-Infected Women with Weight Loss and Low Weight

  1. Jeannie S. Huang,
  2. Stephanie J. Wilkie,
  3. Sara Dolan,
  4. James H. Gallafent,
  5. Negar Aliabadi,
  6. Meghan P. Sullivan, and
  7. Steven Grinspoon
  1. Neuroendocrine Unit and Program in Nutritional Metabolism, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts
  1. Dr. Steven Grinspoon, Program in Nutritional Metabolism, Massachusetts General Hospital, 55 Fruit St., LON, Ste. 207, Boston, MA 02114 (sgrinspoon{at}partners.org).
 
Next Section

Abstract

Risk factors for hypoandrogenemia among low-weight, human immunodeficiency virus (HIV)–infected patients are not known. Testosterone levels of 69 HIV-infected women with low weight and weight loss were compared with levels for 25 healthy, age- and body mass index—matched control subjects. HIV-infected subjects were of low weight, with a mean (± standard deviation) weight loss of -17.6% ± 9.7% from preillness maximum, and 42% of HIV-infected subjects had a body mass index of <20 kg/m2. Forty-nine percent of the HIV-infected population versus 8% of the control population exhibited low free testosterone levels (P <.001). Among HIV-infected women, when we controlled for chronic hepatitis status, age, and time of blood sampling, weight loss of >10% of maximum weight was a significant predictor of low free testosterone levels. Free testosterone levels did not differ by drug class or antiretroviral regimen. In conclusion, decreased androgen levels are common among HIV-infected women reporting significant weight loss, independent of exposure to antiretroviral medications.

In the United States, women represent an increasing percentage of patients with AIDS [1], yet little is known regarding the sex-specific and hormonal factors that may contribute to HIV-related morbidity in this population. HIV-infected women studied before the generalized introduction of HAART have reduced androgen levels [2–4]. Recent data suggest that reduced androgen levels in this population may result from altered androgen metabolism [5]. Reduced androgen levels may further contribute to weight loss, poor sense of well-being, fatigue, and decreased functional status in this population. Prior studies have not investigated androgen levels in relationship to weight loss or antiretroviral use in HIV-infected women in the era of HAART. In this study, we investigated the relationship between reduced androgen levels and antiretroviral regimen among HIV-infected women screened for low weight or weight loss.

Previous SectionNext Section

Methods

From 1997 through 2001, we investigated hormone levels among 69 HIV-infected women (age, 18–50 years) who had low weight and/or weight loss and who had been screened for participation in an interventional study of testosterone administration for women with the AIDS wasting syndrome. Subjects were recruited from community advertisements seeking patients with low weight and/or weight loss and from contact with physicians in the multidisciplinary HIV practice at the Massachusetts General Hospital (Boston) and other community clinics. Subjects were excluded from testing if they had used growth hormone, systemic corticosteroids, megestrol acetate, estrogens, androgens, or any other hormonal preparation that could affect androgen levels <3 months before the study or if they had changed antiviral regimens <6 weeks before the study. Women were characterized as eumenorrheic (normal menstrual function), oligomenorrheic (<3 but ≥1 menstrual period during the 3 months before the study), posthysterectomy, or amenorrheic (no menstrual periods in the 3 months immediately before the study). Subjects were asked to recall preillness, non—pregnancy-related maximum weight. Height was measured by stadiometer, and weight was measured to the nearest 0.1 kg with the subject undressed; body mass index (BMI) and percentage of ideal body weight were determined from these measurements. Percentage of weight loss from maximum was calculated by subtracting measured weight from the reported preillness maximum weight and dividing by preillness maximum weight. Data on use of antiretroviral medications was obtained from medical history and was categorized by current use of protease inhibitors, nucleoside reverse-transcriptase inhibitors, or nonnucleoside reverse-transcriptase inhibitors.

Twenty-five healthy HIV-uninfected women were also recruited for comparison of testosterone levels. Control subjects were age- and BMI-matched to the HIV-infected patients. Subjects were healthy, not pregnant or seeking pregnancy, and without history of use of oral contraceptives or other hormonal therapy during the 3 months before the study. Written informed consent was obtained from each subject before testing, in accordance with the Subcommittee on Human Studies at the Massachusetts General Hospital.

Subject visits were scheduled independent of the menstrual cycle and time of day. Thus, serum samples were obtained randomly and were analyzed with respect to time of day and day of menstrual cycle in subanalyses. All samples obtained from the same patient were tested in duplicate. The free testosterone concentration was determined as the product of the percentage of free testosterone, measured by equilibrium dialysis, and the total testosterone concentration (Endocrine Sciences). The intraassay coefficient of variation of free testosterone is 6.9% and that of total testosterone is <8.1%. The intraassay coefficients of variation were developed with use of pooled sera covering the range of the assay. The normal range for total testosterone concentration in adult women is 10–55 ng/dL (0.4–1.9 nM); for free testosterone, the normal range in adult women is 1.1–6.3 pg/mL (3.8–21.8 pM) (Endocrine Sciences). The normal ranges for total testosterone (n = 215) and free testosterone (n = 141) were determined for healthy subjects (age, 18–46 years) tested during the course of the day and also without regard to timing of menstrual cycle (Endocrine Science). The interassay coefficient of variation for total testosterone is 8%–15%, and the interassay coefficient of variation for free testosterone is 8.9%–11.9%. The sensitivity of total testosterone determination is 3 ng/dL (0.1 nM). The sensitivity of the determination of percentage of free testosterone by this method is 0.1%. Sex hormone binding globulin (SHBG) levels were measured by immunoradiometric assay with an intraassay coefficient of variation of <4% and an interassay coefficient of variation of 7.8%–10.6% (Endocrine Sciences). The normal ranges for this assay are 40–120 nM for premenopausal women and 28–112 nM for postmenopausal women.

Hormone levels (absolute and categorized as less than the normal range or not) were compared among HIV-infected patients by menstrual status (eumenorrheic or not) and by dichotomized weight (weight loss of >10% or <10% from preillness maximum) and antiretroviral regimen grouping (current use of a protease inhibitor, nucleoside reverse-transcriptase inhibitor, and nonnucleoside reverse-transcriptase inhibitor, or HAART). HAART was defined as the use of 2 nucleoside reverse-transcriptase inhibitors and a nonnucleoside reverse-transcriptase inhibitor or protease inhibitor.

Student's t test was used to compare normally distributed continuous variables, nonparametric testing with use of the Wilcoxon rank sum test was used to compare nonnormally distributed continuous variables, and the χ2 test was used for categorical variables. Univariate and multivariate regression analyses were used to determine the relationship between hormone levels, weight, duration of HIV infection, and weight loss. Statistics were computed by use of JMP software (SAS Institute). Results are presented as mean ± SD, unless otherwise indicated. P <.05 indicates statistical significance.

Previous SectionNext Section

Results

Clinical and demographic data for the entire cohort are shown in table 1. The HIV-infected and healthy control groups were well matched with regard to age, BMI, and racial background. Mean times of sampling were also similar between groups, and comparable proportions of subjects in both groups had blood samples obtained in the morning (12% vs. 19% had blood samples obtained in the morning for HIV-uninfected vs. HIV-infected subjects; P =.55).

Table 1
View larger version:
Table 1

Clinical and demographic data for study subjects.

Demographic characteristics of the HIV-infected population. HIV-infected subjects had low weights, with a BMI of 21.0 ± 3.0 kg/m2. Subjects lost a mean of -17.6% ± 9.7% from preillness maximum, and more HIV-infected subjects than control subjects had a BMI of <20 kg/m2. Seventy-five percent of HIV-infected subjects were receiving antiretroviral therapy, and 49% were receiving HAART as defined above. Fifty-seven percent of HIV-infected subjects were eumenorrheic (37% during the early follicular phase, 45% at periovulation, and 18% during the late luteal phase). Three HIV-infected subjects had a history of thyroid disease. In terms of reported viral coinfection, 32% of HIV-infected study participants were coinfected with hepatitis C virus, and 13% were coinfected with hepatitis B virus.

Hormone levels. Total (P <.01) and free (P =.002) testosterone levels were significantly reduced among HIV-infected patients compared with recruited control subjects (table 2), with serum free testosterone levels that were less than the normal range in 49% of HIV-infected subjects but in only 8% of the healthy control subjects (table 2; P <.001). In contrast, only 26% of HIV-infected subjects had total testosterone levels that were less than the normal range. SHBG levels were increased in the HIV-infected group compared with HIV-uninfected group, and they were greater than the normal range in a greater proportion of HIV-infected subjects than healthy control subjects (54% vs. 44%), but this did not reach statistical significance. In a subanalysis involving HIV-infected subjects, SHBG levels were elevated among HIV-infected subjects who were coinfected with hepatitis B and/or hepatitis C virus compared with HIV-infected subjects who were not coinfected (median, 175 nM [interquartile range, 129–236 nM] vs. 110 nM [interquartile range, 90–177 nM]; P =.03).

Table 2
View larger version:
Table 2

Hormonal data for study subjects.

Relationship of clinical variables to free testosterone levels in all subjects. Clinical variables were entered into a logistic regression model to predict low free testosterone levels in the entire cohort. When we controlled for the dichotomous variables of presence of HIV disease, morning blood sampling, and eumenorrhea, a change in weight of >10% of maximum weight remained a significant predictor of low free testosterone level (table 3). Compared with subjects who did not have significant weight loss, subjects with change in weight of >10% of maximum weight had an OR of having hypoandrogenemia of 3.5.

Table 3
View larger version:
Table 3

Findings of a multivariate model of predictors of low free testosterone levels among 84 women.

Relationship of clinical variables to free testosterone levels in HIV-infected subjects. Free testosterone levels were compared by disease and menstrual status and by various weight and medication variables in the HIV-infected cohort (table 4). Significant differences in free testosterone levels were seen in the comparison by percentage of weight loss. Fifty-eight percent of the patients with weight loss of >10% versus 24% of subjects with weight loss of <10% of body weight had a free testosterone level that was less than the normal range for the assay (P =.02). Free testosterone levels were significantly higher in the eumenorrheic subjects (n = 39). No differences in free testosterone levels were seen in comparisons by drug use category or by disease status (AIDS vs. non-AIDS). Although free testosterone levels were not significantly different according to status of coinfection with hepatitis B and/or C virus, an increased proportion of coinfected patients had low free testosterone levels (P <.05).

Table 4
View larger version:
Table 4

Relationship of clinical variables to low free testosterone levels in HIV-infected subjects.

Among the HIV-infected patients, free testosterone levels correlated significantly with age (r = -0.41; P <.001), duration of HIV infection (r = -0.44; P <.001), and percentage of change in weight (r = 0.36; P =.002) but not weight itself (in comparisons with BMI, historical low weight, and percentage of ideal body weight). In a multivariate model involving HIV-infected subjects (table 5), when we controlled for age, time of blood sampling, menstrual status, and chronic hepatitis status, having >10% change in weight from a preillness maximum weight was a significant predictor of serum free testosterone level (r2 =.39; P <.001 for whole model).

Table 5
View larger version:
Table 5

Findings of a multivariate model of predictors of serum free testosterone levels among 64 HIV-infected women.

Among the 39 eumenorrheic HIV-infected subjects, significantly greater proportions of patients demonstrated low free testosterone levels in the early follicular phase (during days 1–7 [n = 14], 71.4% had levels below the normal range) than in the other phases of the menstrual cycle (during days 8–22 [n = 17], 29.4% had levels below the normal range [P =.03]; during days 23–30 [n = 7], 14.3% had levels below the normal range [P =.02]). Although free testosterone levels tended to decrease during the course of the day, when we controlled for day of menstrual cycle and weight loss of >10%, the time at which serum samples were obtained did not remain a significant predictor of low free testosterone level in a multivariate regression model that included eumenorrheic HIV-infected subjects (r2 =.38; P <.001 for whole model).

Previous SectionNext Section

Discussion

In this study, we investigated testosterone levels in relationship to weight, menstrual status, and treatment parameters among HIV-infected women. Our data demonstrate severely reduced testosterone levels in HIV-infected women with weight loss of >10% of preillness maximum weight. In contrast, other weight parameters, including historical low weight and percentage of ideal body weight, as well as use of antiretroviral medication, did not contribute significantly to testosterone levels in this population. Menstrual status did appear to correlate with serum free testosterone levels but did not remain a significant predictor in multivariate modeling.

Prior studies have demonstrated reduced serum testosterone levels in HIV-infected women [2–5], but the mechanism of androgen deficiency in this population is not known. We have previously reported reduced adrenal androgen and increased cortisol responses to adrenocorticotropic hormone stimulation testing in women with AIDS wasting [5], suggesting shunting of adrenal steroid metabolism toward cortisol, which may contribute to reduced testosterone levels. This stress response may indeed be maladaptive if loss of testosterone itself further contributes to low weight and decreased functional status.

In the present study, weight loss was a significant predictor of low testosterone levels in both HIV-infected and HIV-uninfected women (table 3 and table 5). Nevertheless, HIV-infected women did have a greater risk for hypoandrogenemia than did HIV-uninfected, weight-matched control subjects (table 2), and HIV disease was associated with an increased risk of low testosterone levels. In addition, HIV-infected women coinfected with hepatitis B or C virus had an increased prevalence of hypoandrogenemia compared with persons who did not have coinfection (table 4). These data may reflect the additional contribution of chronic disease to hypoandrogenemia [6, 7]. However, we did not demonstrate any difference in free testosterone levels between HIV-infected women with and without AIDS. Similarly, there was no correlation between either CD4 cell count or HIV load and free testosterone level.

Weight loss remains common among HIV-infected patients, even in the era of HAART [8]. However, low testosterone levels may have also contributed to weight loss, and the association between weight loss and decreased androgen levels, in terms of cause and effect, remains unknown among HIV-infected women. In a prior pilot study, physiological testosterone administration was shown to increase weight in HIV-infected women with wasting and relative androgen deficiency [9], suggesting that low testosterone levels may contribute to weight loss in this population. Similarly, dehydroepiandrosterone replacement has been shown to significantly increase body cell mass in HIV-infected persons [10].

Amenorrhea and/or irregular menstrual cycles have been found with increased prevalence among women with HIV infection [11, 12]. Such changes in ovarian function may reflect an alteration in androgen metabolism that leads to hypoandrogenemia. However, ovarian function has been shown to be relatively intact by human chorionic gonadotropin stimulation in prior studies [5], and disturbances of the menstrual cycle did not remain a significant predictor when we accounted for other factors in multivariate analysis.

In the present study, testosterone levels were determined in a relatively large group of HIV-infected women being screened for a study of AIDS wasting. These women did not have any prior medical history of other conditions or use of medications, such as megestrol acetate and estrogen, that might affect androgen status. Subjects were characterized in terms of menstrual function and antiretroviral medication use, as well as by simple parameters of weight and weight history. Given these relatively simple historical parameters, we would anticipate our results to be generalizable and reproducible in the population of HIV-infected women with weight loss. Furthermore, subjects were tested across the menstrual cycle, with 44% of eumenorrheic subjects studied during the period of the anticipated midcycle increase in free testosterone levels demonstrated by Sinha-Hikim et al. [13]. Among the eumenorrheic subjects studied during the early follicular phase, 71% demonstrated reduced serum levels of free testosterone. Therefore, an even greater degree of androgen deficiency would have been seen if testing were limited to the early follicular phase.

We investigated the relationship of testosterone with weight and weight loss by use of an established normal range determined from a large number of healthy volunteers. The specific cutoffs that constitute androgen deficiency in women remain unknown, but androgen levels were significantly reduced in HIV-infected women compared with age- and BMI-matched healthy control subjects.

In this study, weight loss was primarily documented by comparison between current body weight and self-reported historical maximum body weight. Such data may be thus be limited by reporting bias. Nevertheless, historical maximum weight is readily accessible and reflects clinically available information that can be used to identify persons at risk for hypoandrogenemia. In addition, several studies have demonstrated that reported body weight tends to be underestimated rather than overestimated [14, 15], particularly in women [16]. Therefore, the correlation between free testosterone levels and changes in weight may be even more robust when studied with use of actual measurements rather than reported maximum weights.

The use of total testosterone levels to screen for hypoandrogenemia in subjects with altered SHBG levels is not reliable. Consistent with data from prior studies [17], SHBG levels were increased among HIV-infected women compared with healthy control subjects in the present study. As demonstrated in our cohort and in other studies, this increase in SHBG levels may be secondary to chronic hepatitis and coinfection with hepatitis B and/or C viruses. In addition, SHBG levels may also be altered in women undergoing the menopause transition [18]. Therefore, given the difficult interpretation of total testosterone levels in women with abnormal SHBG levels, as found in HIV-infected women with chronic hepatitis and early menopause, our data suggest the need to assess gonadal status by an SHBG-independent equilibrium dialysis method. Indeed, the prevalence of low total testosterone was much lower than the prevalence of low free testosterone, probably because of the increase in SHBG levels.

Our study population was sufficiently large to assess the relationship between the time that blood samples were obtained and androgen levels. Samples were randomly obtained from subjects throughout the day, and our data suggest that the free testosterone level decreased as expected with respect to the time of blood sampling in our population. In this study, we demonstrated that HIV-infected women of low weight or with significant weight loss have an increased likelihood of hypoandrogenemia in comparison with a similarly sampled control population. Furthermore, weight loss of >10% remained a significant predictor of low free testosterone levels on multivariate regression analysis when we controlled for time of blood sampling, chronic hepatitis status, eumenorrheic categorization, and age.

In contrast to androgen deficiency in men, androgen deficiency in women has not been well characterized. In this population screening study, we did not perform detailed assessment of quality of life or functional status. Thus, it remains unclear whether there are functional consequences of androgen deficiency in this population. In men, a number of studies have now shown a significant benefit of androgen replacement in terms of weight, muscle mass, strength, and quality of life [19–22]. Future studies of testosterone replacement are needed to determine whether there are similar benefits of physiological testosterone supplementation in HIV-infected women.

In this study, we demonstrated the relationship between weight loss and androgen levels in HIV-infected women. A large percentage of HIV-infected women with significant weight loss have reduced androgen levels, even in the era of HAART. Determination of the functional consequences of androgen deficiency and the role of physiological androgen replacement will be important in this population.

Previous SectionNext Section

Acknowledgments

We thank the nursing and dietary staffs of the General Clinical Research Center at Massachusetts General Hospital, Boston, for their dedication to patient care.

Previous SectionNext Section

Footnotes

  • Financial support: National Institutes of Health (grants AI-51947, DK-54167, and RR-01066), Bethesda, MD, and the Mary Fisher Clinical AIDS Research and Education (CARE) Fund.

  • Received June 26, 2002.
  • Accepted November 26, 2002.
Previous Section
 

References

  1. ↵
    1. Levine AM
    . Evaluation and management of HIV-infected women. Ann Intern Med 2002;136:228-42.
    Abstract/FREE Full Text
  2. ↵
    1. Grinspoon S,
    2. Corcoran C,
    3. Miller K,
    4. et al
    . Body composition and endocrine function in women with acquired immunodeficiency syndrome wasting. J Clin Endocrinol Metab 1997;82:1332-7.
    Abstract/FREE Full Text
    1. Goggin K,
    2. Engelson ES,
    3. Rabkin JG,
    4. Kotler DP
    . The relationship of mood, endocrine, and sexual disorders in human immunodeficiency virus positive (HIV+) women: an exploratory study. Psychosom Med 1998;60:11-6.
    Abstract/FREE Full Text
  4. ↵
    1. Sinha-Hikim I,
    2. Arver S,
    3. Beall G,
    4. et al
    . Measurement of free testosterone levels by a sensitive equilibrium dialysis method during the normal menstrual cycle in healthy women and in HIV-infected women. J Clin Endocrinol Metab 1998;83:1312-8.
    Abstract/FREE Full Text
  5. ↵
    1. Grinspoon S,
    2. Corcoran C,
    3. Stanley T,
    4. Rabe J,
    5. Wilkie S
    . Mechanisms of androgen deficiency in human immunodeficiency virus-infected women with the wasting syndrome. J Clin Endocrinol Metab 2001;86:4120-6.
    Abstract/FREE Full Text
  6. ↵
    1. Nierman DM
    . Mechanick JI. Hypotestosteronemia in chronically critically ill men. Crit Care Med 1999;27:2418-21.
    CrossRefMedlineWeb of Science
  7. ↵
    1. Kew MC
    . Sexual dysfunction in men with chronic liver disease. Hepatology 1988;8:429-31.
    CrossRefMedline
  8. ↵
    1. Wanke CA,
    2. Silva M,
    3. Knox TA,
    4. et al
    . Weight loss and wasting remain common complications in individuals infected with human immunodeficiency virus in the era of highly active antiretroviral therapy. Clin Infect Dis 2000;31:803-5.
    Abstract/FREE Full Text
  9. ↵
    1. Miller K,
    2. Corcoran C,
    3. Armstrong C,
    4. et al
    . Transdermal testosterone administration in women with acquired immunodeficiency syndrome wasting: a pilot study. J Clin Endocrinol Metab 1998;83:2717-25.
    Abstract/FREE Full Text
  10. ↵
    1. Rabkin JG,
    2. Ferrando SJ,
    3. Wagner GJ,
    4. Rabkin R
    . DHEA treatment of HIV+ patients: effects on mood, androgenic and anabolic parameters. Psychoneuroendocrinology 2000;25:53-68.
    CrossRefMedlineWeb of Science
  11. ↵
    1. Clark RA,
    2. Mulligan K,
    3. Stamenovic E,
    4. et al
    . Frequency of anovulation and early menopause among women enrolled in selected adult AIDS Clinical Trials Group studies. J Infect Dis 2001;184:1325-7.
    Abstract/FREE Full Text
  12. ↵
    1. Chirgwin KD,
    2. Feldman J,
    3. Muneyyirci-Delale O,
    4. Landesman S,
    5. Minkoff H
    . Menstrual function in human immunodeficiency virus-infected women without acquired immunodeficiency syndrome. J Acquir Immune Defic Syndr Hum Retrovirol 1996;12:489-94.
    MedlineWeb of Science
  13. ↵
    1. Sinha-Hikim I,
    2. Arver S,
    3. Beall G,
    4. et al
    . The use of a sensitive equilibrium dialysis method for the measurement of free testosterone levels in healthy, cycling women and in human immunodeficiency virus-infected women. J Clin Endocrinol Metab 1998;83:1312-8.
    Abstract/FREE Full Text
  14. ↵
    1. Black DR,
    2. Taylor AM,
    3. Coster DC
    . Accuracy of self-reported body weight: Stepped Approach Model component assessment. Health Educ Res 1998;13:301-7.
    Abstract/FREE Full Text
  15. ↵
    1. Kuczmarski MF,
    2. Kuczmarski RJ,
    3. Najjar M
    . Effects of age on validity of self-reported height, weight, and body mass index: findings from the Third National Health and Nutrition Examination Survey, 1988–1994. J Am Diet Assoc 2001;101:28-34.
    CrossRefMedlineWeb of Science
  16. ↵
    1. Perry GS,
    2. Byers TE,
    3. Mokdad AH,
    4. Serdula MK,
    5. Williamson DF
    . The validity of self-reports of past body weights by U.S. adults. Epidemiology 1995;6:61-6.
    MedlineWeb of Science
  17. ↵
    1. Martin ME,
    2. Benassayag C,
    3. Amiel C,
    4. Canton P,
    5. Nunez EA
    . Alterations in the concentrations and binding properties of sex steroid binding protein and corticosteroid-binding globulin in HIV+ patients. J Endocrinol Invest 1992;15:597-603.
    MedlineWeb of Science
  18. ↵
    1. Rannevik G,
    2. Jeppsson S,
    3. Johnell O,
    4. et al
    . A longitudinal study of the perimenopausal transition: altered profiles of steroid and pituitary hormones, SHBG and bone mineral density. Maturitas 1995;21:103-13.
    CrossRefMedlineWeb of Science
  19. ↵
    1. Wang C,
    2. Swedloff RS,
    3. Iranmanesh A,
    4. et al
    . Transdermal testosterone gel improves sexual function, mood, muscle strength, and body composition parameters in hypogonadal men. Testosterone Gel Study Group. J Clin Endocrinol Metab 2000;85:2839-53.
    Abstract/FREE Full Text
    1. Fairfield WP,
    2. Treat M,
    3. Rosenthal DI,
    4. et al
    . Effects of testosterone and exercise on muscle leanness in eugonadal men with AIDS wasting. J Appl Physiol 2001;90:2166-71.
    Abstract/FREE Full Text
    1. Grinspoon S,
    2. Corcoran C,
    3. Parlman K,
    4. et al
    . Effects of testosterone and progressive resistance training in eugonadal men with AIDS wasting: a randomized, controlled trial. Ann Intern Med 2000;133:348-55.
    Abstract/FREE Full Text
  22. ↵
    1. Bhasin S,
    2. Storer TW,
    3. Javanbakht M,
    4. et al
    . Testosterone replacement and resistance exercise in HIV-infected men with weight loss and low testosterone levels. JAMA 2000;283:763-70.
    Abstract/FREE Full Text
| Table of Contents

This Article

  1. Clin Infect Dis. (2003) 36 (4): 499-506. doi: 10.1086/367642
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 54 (5)
  1. Clinical Infectious Diseases
  1. Alert me to new issues

Most