Low Sex Hormone-Binding Globulin and Testosterone Levels in Association with Erectile Dysfunction among Human Immunodeficiency Virus-Infected Men Receiving Testosterone and Oxandrolone


Peter Wasserman, MA, Infectious Disease Division, Department of Medicine––New York Hospital Queens, 137-48 Horace Harding Expressway, Flushing, NY, USA 11367. Tel: 718-670-1639; Fax: 718-321-8095; E-mail: pjw9003@nyp.org


Introduction.  Men with acquired immunodeficiency syndrome (AIDS) wasting and hypogonadism are frequently treated with testosterone and oxandrolone, an orally administered anabolic–androgenic steroid hormone. We observed reductions in testosterone and sex hormone-binding globulin (SHBG) levels, in association with complaints of erectile dysfunction, after prolonged exposure to this therapeutic regimen.

Aim.  First description of an association between long-term receipt of oxandrolone with erectile dysfunction, low SHBG and testosterone.

Methods.  Case report of three human immunodeficiency virus-infected hypogonadal male patients receiving treatment for wasting syndrome and hypogonadism, and highly active antiretroviral therapy. All three patients received long-term oxandrolone in addition to testosterone replacement therapy.

Results.  Testosterone and SHBG levels for patients 1, 2, and 3, respectively: total testosterone 183, 71, and 151 ng/dL (260–1,000 ng/dL); free testosterone (not done for patient 3) 58.3 and 26.9 pg/mL (50–210 pg/mL); SHBG 6, 9, and 6 nmol/L (7–50 nmol/L). No other hormonal abnormalities were detected. Following discontinuation of oxandrolone, levels of total testosterone rose, consistent with increase in SHBG. One patient received repeat SHBG assay documenting rise in SHBG level. Patient 2 reported return of libido and early morning erections several weeks after discontinuation of oxandrolone.

Conclusions.  Patients had erectile dysfunction in association with low testosterone and SHBG, in spite of exogenous testosterone replacement. Discontinuation of oxandrolone led to the normalization or improvement of testosterone levels in all three patients with symptomatic improvement in one patient. First pass metabolism of orally administered oxandrolone may decrease hepatic synthesis of SHBG, allowing exogenously supplied testosterone to be excreted. Further work is necessary to elucidate the relationship. Wasserman P, Segal-Maurer S, and Rubin D. Low sex hormone-binding globulin and testosterone levels in association with erectile dysfunction among human immunodeficiency virus-infected men receiving testosterone and oxandrolone. J Sex Med 2008;5:241–247.


Acquired immunodeficiency syndrome (AIDS) wasting was described in 1983 as a syndrome consisting of weight loss, diminished nutrient intake, and profound immunodeficiency with active opportunistic infection [1]. Wasting syndrome remains prevalent in the era of highly active antiretroviral therapy (HAART) and may also occur in the absence of a serious decline in CD4 cell count [2,3]. The onset of wasting syndrome is an AIDS-defining event irrespective of the presence or absence of opportunistic infection or neoplastic disease [4]. In addition, a number of studies have found an association between AIDS, wasting syndrome, and hypogonadism [5–7]. It is now known that human immunodeficiency virus (HIV)-1 may infect the testis, which in turn, may constitute a sanctuary for its continued replication [8]. Both primary and secondary hypogonadism occur in the setting of HIV infection [5,6]. In a recent study of HIV-infected men with weight loss, hypogonadism was a frequent finding despite reported CD4 counts greater than 300 cells/µL and the use of HAART by a majority of patients [9].

The beneficial effects of testosterone replacement therapy on the constitutional symptoms of wasting syndrome and accrual of lean body mass in hypogonadal HIV-infected men are well documented [10,11]. Eugonadal HIV-infected men who received oxandrolone (an oral 17-α-methylated anabolic–androgenic steroid) experienced significant gains in lean body mass while participating in a progressive resistance exercise program [12]. Patients in the intervention and control arms of the trial received testosterone enanthate by intramuscular injection to ensure eugonadal status [12]. As a result of this trial, clinicians may coprescribe testosterone and oxandrolone for the treatment of wasting syndrome. There are no studies to corroborate or refute this practice. Moreover, testosterone and oxandrolone, a synthetic derivative of testosterone, both act on the same unique androgen receptor, suggesting this practice to be of little clinical utility. There are no reports of adverse events related to the long-term usage of anabolic–androgenic hormones in HIV-infected patients. In this article, we describe erectile dysfunction, profound hypotestosteronemia, and low sex hormone-binding globulin (SHBG) associated with long-term receipt of oxandrolone and testosterone replacement in three HIV-infected patients.


Case report of three HIV-infected subjects who were hypogonadal despite exogenous testosterone supplementation. Weight and height were determined by Detecto beam scale/stadiometer (Webb City, MO, USA) and expressed as body mass index (BMI). Waist was measured at the level of the umbilicus. Total and free testosterone levels were assayed by chemiluminescent immunoassay, equilibrium dialysis (Diagnostic Products Corporation, Los Angeles, CA, USA). Antiretrovirals for the treatment of HIV infection are not known to interfere with this assay. SHBG and fasting insulin levels were determined by chemiluminescent immunoassay (Diagnostic Products Corporation). Thyrotropin (TSH), prolactin, luteinizing hormone (LH), and follicle-stimulating hormone (FSH) were assayed by electrochemiluminescence immunoassay (Roche Diagnostics GmbH, Mannheim, Germany). Plasma HIV-1 RNA load were measured by reverse transcriptase-polymerase chain reaction (RT-PCR) assay (Amplicor HIV Monitor, Roche Diagnostic Systems, Basel, Switzerland). CD4 T-lymphocytes cells were determined by flow cytometry.


The subjects identified were three HIV-infected hypogonadal male patients on long-term HAART, receiving treatment for wasting syndrome and hypogonadism. All three subjects were free of hepatitis B and C virus coinfection or other hepatic disease. Patients received routine clinical care and did not undergo therapeutic drug monitoring of serum antiretroviral levels. Patient 1 is a 44-year-old Hispanic male diagnosed with HIV infection in 1990, with nadir CD4 23 cells/µL (2%) and peak viral load (VL) >750,000 copies/mL. Centers for Disease Control (CDC) transmission category was men who have sex with men (MSM). Past medical history is significant for hypertension, diverticulitis, hypertriglyceridemia, peripheral neuropathy, and chronic depression (managed with bupropion and fluoxetine). This patient did not exhibit signs of depressive disorder (e.g., anhedonia or social isolation) during the period described in this report. Hypertension, hypertriglyceridemia, and peripheral neuropathy were managed with atenolol, gemfibrozil, and gabapentin, respectively. He was treated with multiple nucleoside reverse transcriptase inhibitors (NRTIs) from 1992 until 1996 when he began HAART with zidovudine, lamivudine, and ritonavir. Genotypic analysis in 2001 revealed multidrug-resistant HIV. His antiretroviral regimen at the time of the report was zidovudine, lamivudine, abacavir, tenofovir, delavirdine mesylate, and lopinavir/ritonavir with excellent compliance. The most current CD4 and VL were 310 (17%) cells/µL and 1,581 copies/mL, respectively, and wasting persisted despite virologic response. The patient had no history of cigarette smoking or illicit substance or alcohol use. He received oxandrolone 10 mg orally twice daily for 43 months (March 1999 to December 2002), testosterone 5 mg once daily transdermally for 34 months (February 2000 to January 2003), and recombinant human growth hormone (rhGH) 6 mg subcutaneously once daily for 21 months (intermittently between March 1999 and December 2002).

Patient 2 is a 50-year-old white male diagnosed with AIDS in 1999 when he presented with Pneumocystis carinii pneumonia and severe wasting syndrome (BMI 19.7 and greater than 20% weight loss). CDC transmission category was MSM with a known HIV-infected partner. Past medical history is significant for severe peripheral neuropathy, hypertension, gastritis, erectile dysfunction not responsive to phosphodiesterase type 5 inhibitor, and depression (managed with amitriptyline and escitalopram). The patient did not exhibit signs of depressive disorder during the period described in this report. Chronic pain because of peripheral neuropathy was managed with methadone and gabapentin. His nadir CD4 was 87 cells/µL (12%), peak VL >1 million copies/mL. He began HAART in 1999 with zidovudine, lamivudine, abacavir, and efavirenz, and this was his regimen at the time of the report. Response was brisk and durable with most current VL of 239 copies/mL and CD4 665 (30%) cells/µL. The patient smoked cigarettes but did not have a history of substance or alcohol use. The patient received oxandrolone 10 mg twice daily orally for 36 months (December 1999 to January 2003) and testosterone 5 mg once daily transdermally for 39 months (October 1999 through January 2003).

Patient 3 is a 49-year-old white male diagnosed with HIV infection in 1991 with a nadir CD4 count of 247 cells/µL (17.3%) and peak VL of 264,000 copies/mL. CDC transmission category was heterosexual sex. Past medical history is significant for episodes of recurrent Herpes labialis, bacterial pneumonia, and erectile dysfunction for which he received phosphodiesterase type 5 inhibitor. Social history is significant for incarceration and substance use, including cocaine and analgesic medications. The patient was receiving psychotherapy and only exhibited signs of poor anger management during the period described. He began HAART in 1996 with zidovudine, lamivudine, and saquinavir until 1999 when HAART was discontinued. The patient resumed HAART in 2001 and was treated with a series of regimens, including stavudine, lamivudine, abacavir, didanosine, tenofovir, efavirenz, nevirapine, and nelfinavir. The patient had self-interrupted his most recent HAART regimen at time of report because of gastrointestinal intolerance. His most recent CD4 was at 315 (16.6%) cells/µL with a VL of 3,440 copies/mL. The patient received oxandrolone 2.5 mg orally twice daily for 30 months then 10 mg twice daily for 7 months (late 1999 to November 2003). Testosterone cypionate was administered intramuscularly, initially 200 mg once monthly progressing to biweekly, for 41 months (March 2000 to January 2002).


All three patients presented with complaints of erectile dysfunction, decreased morning erections, and decreased libido. Testosterone levels were significantly decreased in all patients despite provision of exogenous hormone (Table 1). LH and FSH were low or within normal range in all three patients consistent with secondary (central) hypogonadism. Further serologic analysis revealed low SHBG levels in all patients. No patient demonstrated hepatic transaminase elevations greater than grade 1 or pattern of elevation suggestive of hepatocellular toxicity reaction (defined as elevation alanine aminotransferase (AST) and aspartate aminotransferase (ALT) three times the upper limit of normal and total bilirubin 1.5 times the upper limit of normal) (Table 1). It should be noted that mild isolated transaminase elevations are common in this therapeutic population. Exogenous testosterone was continued but oxandrolone was discontinued in all patients. Significant intercurrent change in BMI did not occur (Table 1). Patient 2 demonstrated BMI >25.0 at both time points because of a highly muscled frame. No patient demonstrated waist circumference consistent with abdominal obesity as defined by the National Cholesterol Education Program Adult Treatment Panel III guidelines [13]. Follow-up analysis documented an increase in testosterone levels (Table 1). SHBG normalized in patient 3 at 18 nmol/L 7 weeks after discontinuation of oxandrolone. Patient 2 reported return of libido and improvement in erectile function and return of early morning erections within weeks of discontinuing oxandrolone. All three patients demonstrated improved serum total testosterone during the dosing interval consistent with restoration of SHBG concentrations. Among the three patients, levels of total testosterone were greater in the patients with longer time from discontinuation of oxandrolone treatment.

Table 1.  Testosterone and other endocrine parameters in the three subjects
TestUnitsPatient 1Patient 2Patient 3Reference range
  • *

    Supplementation regimen prior to discontinuation of oxandrolone.

  • Supplementation regimen following discontinuation.

  • Oxandrolone 10 mg orally twice daily. Testosterone 5 mg transdermally daily.

  • §

    Oxandrolone 2.5 mg orally twice daily (then 10 mg twice daily). Testosterone cypionate 200 mg intramuscularly monthly, then bimonthly.

  • Testosterone 5 mg transdermally daily.

  • **

    Testosterone 200 mg intramuscularly bimonthly.

  • ††

    Ten weeks after discontinuation of oxandrolone.

  • ‡‡

    Twelve weeks after discontinuation of oxandrolone.

  • §§

    Seven weeks after discontinuation of oxandrolone.

  • ¶¶

    Eight days after receipt of 200 mg testosterone cypionate intramuscularly.

  • SHBG = sex hormone-binding globulin; LH = leutenizing hormone; FSH = follicle stimulating hormone; TSH = thyroid stimulating hormone; BMI = body mass index; AST = aspartate aminotransferase; ALT = alanine aminotransferase; GGT = gamma-glutamyl transferase; NA = not available.

Total testosterone(ng/dL)183330711392151187260–1,000 ng/dL
Free testosterone(pg/mL)58.361.426.9370.5N/A35.850–210 pg/mL
Free testosterone(%)–2.7%
SHBG(nmol/L)6NA9NA6187–50 nmol/L
LH(mIU/mL)8.000.312–12 mIU/mL
FSH(mIU/mL)5.90.350.981–8 mIU/mL
Prolactin(ng/mL)8.317.97.464.1–18.4 ng/mL
TSH(µIU/mL)0.81.841.80.27–4.2 µIU/mL
Fasting insulin(µIU)13.38.6N/A<17.0 µIU
BMI 23.422.623.623.127.827.820.0–25.0
Waist circumference(cm)82.685.194.699.0 
AST585427NA325710–42 U/L
ALT282241NA629810–60 U/L
GGT8213188NA50587–64 U/L
Bilirubin, total0.90.90.3NA0.80.90.2–1.0 mg/dL


Men with AIDS wasting syndrome and hypogonadism frequently receive testosterone replacement and oxandrolone therapy. We observed a symptomatic decrease in total and free testosterone despite transdermal or intramuscular replacement therapy in patients receiving oxandrolone long term. In addition, we noted low levels of SHBG. SHBG is a steroid-binding plasma glycoprotein primarily synthesized in the liver as well as in prostatic and breast tissues [14]. It binds testosterone and 5 α-dihydrotestosterone with high affinity (estradiol with low affinity) and can mediate androgen and estrogen signaling at the cell membrane level via receptor-mediated signal transduction [15]. The activation of the cell membrane level SHBG receptor is believed to amplify or modulate sex hormone action, which is mediated at the level of DNA transcription by the binding of nuclear hormone receptors to DNA.

Plasma SHBG concentration is decreased by a number of factors including androgens, glucocorticoid, and increased insulin levels [16,17]. The regulation of SHBG synthesis by androgens is not clearly understood. In men, SHBG synthesis increases with age and mediates slower declines in total testosterone levels (as compared to free testosterone levels) [18]. Basal levels of SHBG synthesis may be modulated by body weight (obesity or cachexia), disease states and administration of exogenous hormone [15,17,19,20]. Growth hormone (GH) has been shown to decrease SHBG binding activity, consistent with an anabolic state [21]. Insulin-like growth factor-1 (IGF-1), a potent mitogen upregulated by GH and downregulated by undernutrition, may link nutritional status with the onset of an anabolic state by its downregulation of SHBG, resulting in an increase in free testosterone.

SHBG may be upregulated by the administration of oral conjugated estrogen, a finding consistent with direct estrogen effect on hepatic synthesis of this globulin [22]. Elevations in serum estradiol have been observed in HIV-infected men with normal total testosterone levels receiving combination antiretroviral therapy [23]. Estradiol elevations were believed to be due to lipodystrophy with abdominal fat accumulation (as compared to primarily peripheral lipoatrophic lipodystrophy) and mediated by adipocyte aromatase [23]. Our patients demonstrated normal or slightly above normal range (when highly muscled) BMI consistent with normal range body fat content at the time of the report. No patient demonstrated waist circumference consistent with abdominal obesity. Estradiol was not assayed because the patients demonstrated low or low normal SHBG.

Potent oxandrolone-mediated reduction of SHBG concentrations has previously been reported in children with Turners syndrome and constitutional delay of growth and puberty [24,25]. Recently, Grunfeld et al. noted suppression of SHBG, total and free testosterone levels at 12 weeks in a prospective trial of oxandrolone for HIV-infected men with wasting syndrome [26]. Patients were not receiving exogenous testosterone in addition to oxandrolone.

SHBG levels can be either normal or elevated at all stages of HIV infection and are highly correlated with levels of total testosterone in HIV-infected patients [27–30]. Alterations in corticosteroid (dexamethasone and fluticasone propionate) and ethinyl-estradiol pharmacokinetics and/or pharmacodynamics by protease inhibitors and non-NRTIs have been described. Antiretroviral effects on testosterone metabolism have not been documented in the literature. Earlier works with oral testosterone undecanoate and oral micronized estradiol-17β (in HIV-uninfected subjects) demonstrated that the regulatory effects of exogenously administered androgens or estrogens on SHBG are weak or absent when administered percutaneously rather than orally [31–33].

Erectile dysfunction associated with low SHBG and testosterone levels among patients receiving oxandrolone, has not been previously described in HIV-infected patients. We describe functional and biochemical abnormalities in three HIV-infected adults electively receiving oxandrolone. In our three subjects, total testosterone levels were markedly decreased in spite of testosterone replacement therapy. Common to all subjects was the long-term usage of oxandrolone. One of our subjects received subcutaneous rhGH at pharmacologic doses, which could have theoretically downregulated SHBG synthesis by increasing IGF-1 levels. However, discontinuation of oxandrolone in all subjects (and rhGH in patient 1) led to the retention of total (bound) testosterone at or above normal levels in patients 1 and 2, and increase in patient 3. Increases in total testosterone levels are highly consistent with rise in SHBG levels. Patient 2 reported subsequent improvement in erectile function with receipt of intramuscular testosterone at replacement dosage (changed from daily transdermal formulation at replacement dosage). Our findings suggest that SHBG-bound testosterone may function as a reservoir for the maintenance of free testosterone levels. When SHBG levels are suppressed in the setting of hypogonadism, exogenously supplied unbound testosterone may be hepatically cleared, resulting in decreased testosterone levels.

Withdrawal of oxandrolone resulted in the return of erectile function in addition to systemic retention of exogenous testosterone in patient 2. This suggests that the loss of replacement testosterone because of oxandrolone-mediated diminishment of SHBG synthesis, lead to erectile dysfunction in this patient. In addition, lack of previous response to phosphodiesterase type 5 inhibitor is consistent with erectile dysfunction because of hypogonadism [34,35]. It should be noted that all three patients had other conditions associated with erectile dysfunction, e.g., peripheral neuropathy, hypertension, psychiatric disease, or substance use. Lack of improvement in erectile function in patients 1 and 3 may have been because of underlying psychological and/or medical conditions.

Only one androgen receptor has been identified to date. This suggests that at equivalent doses all anabolic–androgenic steroids mediate the same tissue-specific effects [36]. These cases suggest intravenous formulations, even at supraphysiologic dosing, may be safer and more efficacious than oral formulations because of the absence of first pass metabolism. Recent work documents preservation of sexual function in patients receiving megestrol acetate by intramuscular provision of testosterone enanthate [37]. Our case report further supports this approach, which addresses reduced energy intake, a primary cause of wasting syndrome, in addition to endocrinopathy. Duration of therapy with oxandrolone, if an oral formulation must be used, should be limited, and testosterone levels closely monitored. If testosterone levels are decreased during receipt of oxandrolone, the measurement of SHBG may help in identifying a possible etiology in the face of concurrent use of an oral 17-α-methylated anabolic–androgenic steroid. Lipoatrophic-lipodystrophy patients receiving “salvage regimens” whose clinical presentation may be similar to wasting syndrome (decreased BMI and extremity circumferences) present a special problem. It may be difficult to limit the duration of anabolic–androgenic steroid use in these patients who request a “remedy” for their stigmata. However, long-term use of orally administered anabolic-androgenic steroids may lead to decreased levels of SHBG, with decreased free and total testosterone levels, leading to persistent erectile dysfunction and loss of accumulated gains in lean body mass. Persistent erectile dysfunction, in addition to negative body image because of wasting, may ultimately result in depression and loss of adherence to antiretroviral therapy.

Conflict of Interest: None declared.