Clinical signs and symptoms of male hypogonadism may be subtle, nonspecific, and influenced by the severity and duration of
androgen deficiency, previous testosterone treatment, the patient's age, and comorbidities. Laboratory evaluation of male
hypogonadism may also present challenges.
Male hypogonadism is a clinical syndrome caused by inadequate testosterone (T) production (androgen deficiency) that is accompanied
by a decline in sperm production by the testes.[1] Whereas diminished spermatogenesis with normal T production causes male infertility without manifestations of androgen deficiency,
hypogonadism resulting in androgen deficiency impairs many body functions in addition to spermatogenesis.
Since both the prevalence and incidence of androgen deficiency increase as men age, male hypogonadism is a common disorder.
Its diagnosis, however, may be challenging because in addition to unequivocally low serum T levels, it should be based on
clinical manifestations of androgen deficiency, which may be subtle, nonspecific, and modified by the severity and duration
of androgen deficiency, previous T treatment, the patient's age, co-morbidities, and variations in androgen sensitivity.
This article discusses the basic principles and nuances of diagnosing and evaluating male hypogonadism, the clinical manifestations
of androgen deficiency, and the laboratory measurements most often used to assess a man's androgen status. It describes how
to distinguish primary from secondary hypogonadism and, through a case scenario, addresses some of the major considerations
in patient assessment.
A Case to Consider
A moderately obese 58-year-old black man seeks treatment for erectile dysfunction (ED). He has type 2 diabetes mellitus, complicated
by painful peripheral neuropathy and stage 2 chronic kidney disease with proteinuria, hypertension, dyslipidemia, and asthma.
For the past 3 years, he says, his erections have been poorly sustained and penile rigidity has been insufficient to complete
sexual intercourse, despite normal sexual desire on both his part and that of his wife of 21 years. He also reports having
experienced excessive fatigue, accompanied by reduced physical activity and a 15-lb weight gain over the past year.
Current medications include: metformin, NPH insulin, acetaminophen with codeine, hydrochlorothiazide, atenolol, lisinopril,
simvastatin, niacin, and inhaled salmeterol and fluticasone. A recent asthma exacerbation was treated with a pulse and tapering
doses of prednisone over 2 weeks.
The patient smokes 1 pack of cigarettes daily but no longer drinks alcohol. There is no history of coronary artery disease,
prostate or bladder surgery, or genitourinary trauma.
On physical examination, he is found to have a blood pressure of 124/84 mm Hg, regular pulse of 64/min, respiratory rate of
14/min, body mass index of 35, and pain level of 4/10 (his usual level of perceived chronic neuropathic pain). His mood is
mildly depressed. He has normal skin thickness; mild frontal balding; and normal axillary, chest, and pubic hair. Approximately
3-cm nontender, palpable breast tissue is detected bilaterally. Lung and cardiovascular examinations are unremarkable. Despite
moderate central adiposity, his abdomen is without striae. His penis appears normal. Testicular volumes are normal: 25 mL
on the right and 30 mL on the left (by orchidometer). His prostate is mildly enlarged without palpable nodules or induration.
Hair loss is evident on his lower extremities, and dorsalis pedis and posterior tibial pulses are absent to palpation. He
has reduced vibratory sensation in both feet, but intact sensation to a 10-g monofilament. Thigh muscle bulk is reduced, though
thigh strength is normal.
Laboratory evaluation within the past month reveals: a normal complete blood count; glycosylated hemoglobin, 8.3% (normal,
6% or less); chemistry panel, significant only for a fasting glucose of 198 mg/dL (normal, 70 to 100 mg/dL) and an estimated
glomerular filtration rate of 62 mL/min/1.73 m2 (normal for men his age, 93 mL/min/1.73 m2); total cholesterol, 200 mg/dL (normal, less than 200 mg/dL); triglycerides, 366 mg/dL (normal, less than 150 mg/dL); high-density
lipoprotein cholesterol, 30 mg/dL (normal, above 40 mg/dL in men and above 50 mg/dL in women); low-density lipoprotein cholesterol,
97 mg/dL (normal, less than 100 mg/dL); albumin, 3.2 g/dL (normal, 3.4 to 5.4 g/dL); urine protein excretion, 1022 mg/24 h
(normal, 150 mg/24 h); thyroid-stimulating hormone, 0.89 IU/L (normal, 0.4 to 4 IU/L); prostate-specific antigen, 1.6 ng/mL
(normal, below 4 ng/mL); and total T, 248 ng/dL (normal, 280 to 800 ng/dL).
Is it appropriate to evaluate this patient for hypo-gonadism? Does the patient presented have male hypogonadism?
Prevalence and Incidence of Male Hypogonadism
In community-dwelling white men, the prevalence of biochemical hypogonadism (defined as T level lower than the 2.5th percentile,
or less than 325 ng/dL), was found to be 12%, 19%, 28%, and 49% in men in their 50s, 60s, 70s, and 80s, respectively.[2] By contrast, symptomatic androgen deficiency (defined as serum total T of less than 200 ng/dL or free T of less than 8.9
ng/mL and the presence of at least three signs or symptoms consistent with androgen deficiency) was found to have a crude
prevalence of 6% in a predominantly white cohort of men aged 40 through 70 years.[3] After an average of 9 years of follow-up, when the same men were aged 48 through 79 years, the prevalence rose to 12.3%,
with age-specific prevalences being 7.1%, 11.5%, and 22.8% for men in their 50s, 60s, and 70s, respectively.[3] The crude incidence rate of symptomatic androgen deficiency was 12.3% per 1000 person-years and increased with age: 5.9%,
11.2%, and 23.3% per 1000 person-years in men who were in their 40s, 50s, and 60s, respectively, at baseline.[3]
In a community-based population of white, black, and Hispanic men aged 30 through 79 years, symptomatic androgen deficiency
(defined as serum total T of less than 300 ng/dL, or free T of less than 5 ng/dL and either one suggestive symptom, or at
least two nonspecific symptoms of androgen deficiency, based on the Endocrine Society clinical practice guidelines[4]) was found to have a prevalence of 5.6%, and there were no differences in prevalence rates among the three ethnic groups.[5] In this population, prevalence in men younger than 70 was 3.1% to 7% but increased to 18.4% in men aged 70 and older.
Diagnosis of Male Hypogonadism
Male hypogonadism should be diagnosed only in men who have clinical signs and symptoms that are consistent with androgen deficiency,
and biochemical androgen deficiency confirmed by unequivocally low serum T levels (Figure).[4] The clinical presentation of male hypogonadism depends on the stage of development during which androgen deficiency occurs.
-
Figure 1.
Fetal and Prepubertal Onset
During fetal development, androgen deficiency or defects in androgen action (such as androgen receptor mutations) result in
varying degrees of ambiguous genitalia, depending on the severity of androgen deficiency or androgen resistance. For example,
men with an androgen receptor mutation resulting in complete inactivation of the receptor present as phenotypic females (known
as testicular feminization), due to a lack of androgen action during fetal development of the external genitalia.
Prepubertal androgen deficiency results in delayed puberty and eunuchoidism,[1] which is characterized by infantile genitalia (small penis and testes), long arms and legs (due to failure of long bone epiphyses
to close), poor muscular development, increased body fat, reduced peak bone mass, high-pitched voice, and sparse male-pattern
(axillary, chest, facial, extremity, pubic, and perianal) body hair. Because clinical manifestations of prepubertal androgen
deficiency are usually obvious and associated with psychosocial distress in boys and their families, the diagnosis is rarely
missed. Some boys with prepubertal androgen deficiency, however, do not seek medical care and are not diagnosed until they
are adults.
Adult Onset
The signs and symptoms of androgen deficiency acquired after pubertal development are nonspecific and subtler, in part because serum T levels may not be reduced severely. When substantial androgen deficiency occurs over a prolonged period—for example, in men receiving gonadotropin-releasing hormone (GnRH) agonist or antiandrogen therapy for prostate cancer—clinical signs and symptoms are more obvious and characterize the full spectrum of the adult male hypogonadism syndrome (Table 1). In most men with hypogonadism, however, the nonspecific signs and symptoms of androgen deficiency are less obvious clinically. Furthermore, clinical manifestations may be modified by the severity and duration of androgen deficiency, previous T treatment, patient age, presence of comorbidities, and variations in androgen sensitivity—resulting in more highly variable clinical manifestations, especially in older men, and greater difficulty in making a clinical diagnosis of androgen deficiency.
-
-
Clinical findings that are suggestive of adult male hypogonadism may be classified as sexual manifestations (ED, infertility,
and shrinking or very small [especially less than 5 mL] testes); brain/behavioral manifestations (reduced libido, hot flushes,
and sweating); and physical manifestations (breast discomfort; gynecomastia; loss of axillary, pubic, and facial hair; and,
with more severe, long-standing androgen deficiency, low bone mineral density [BMD], low trauma fracture, and reduced muscle
bulk and strength). In addition, there may be less specific manifestations of androgen deficiency that occur in men with hypogonadism,
including diminished energy and vitality, poor motivation, depressed mood, irritability, sleep disturbance, sleepiness, reduced
concentration and memory, decreased activity and physical performance, mild anemia, and increased body fat.
Confirming Androgen Deficiency
In men with clinical manifestations of androgen deficiency, hypogonadism is confirmed by presence of low serum total T levels.
Serum total T levels exhibit considerable assay and biological variability, and are affected by changes in sex hormone-binding
globulin (SHBG) concentrations, illness, medications, and nutritional deficiency. As with the clinical diagnosis of male hypogonadism,
biochemical confirmation of androgen deficiency presents its own set of challenges.
Initial Measurement of Total T Levels
For the most part, circulating T is bound to serum proteins. Approximately 44% is tightly bound to SHBG, and 54% is weakly
bound to albumin but dissociable into some target tissues. Only about 2% of T is free and circulates unbound to proteins.[6] Unlike SHBG-bound T, free T and albumin-bound T are available to tissues for biological action; hence, the combination of
free and albumin-bound T is referred to as bioavailable T. Total T refers to the combination of free, albumin-bound, and SHBG-bound
T. Because a significant fraction of T is bound to SHBG, measurements of total T are affected directly by alterations in SHBG
concentrations.
Total T is the measurement that is most readily available in local laboratories and most commonly used to assess the adequacy
of androgen status in men and enroll subjects in clinical trials of T treatment of hypogonadal men. Total T measurements are
performed by immunoassay and, more recently, by liquid chromatography tandem mass spectrometry. Automated immunoassays for
total T are performed in most local clinical laboratories and usually are accurate enough to distinguish eugonadal from hypogonadal
men.[7] There is, however, considerable variability in total T measurements for a number of reasons.
Variability in Total T Levels
Assay-to-assay variability in reported total T values may be extreme. For example, total T values measured on the same College
of American Pathologists external quality control sample by different automated assays ranged from 160 to 508 ng/dL, spanning
the hypogonadal to eugonadal range.[7] In addition to assay variability, T levels exhibit biological variability due to episodic secretion from the testes and a
circadian variation in serum total T levels, with peak concentrations in the morning, which is blunted in healthy older men
compared with young men.[8] For this reason, it is recommended that T measurements be performed in the morning, especially in young men.
Even within individual men, there is considerable day-to-day variability in T levels, such that a single total T measurement
does not adequately reflect the average T level in a man. In a study of intraindividual variation in reproductive hormone
levels, 15 of 121 men who underwent repeated blood sampling over 6 months had serum total T levels of less than 250 ng/dL
(that is, in the hypogonadal range) measured on an initial baseline sample. Of these 15 men, only 6 had average T levels less
than 250 ng/dL and 3 had average T levels greater than 300 ng/dL on repeated blood sampling over the subsequent 6 months.[9] Diagnostic accuracy was improved by averaging T values from the first two blood samples (performed 1 to 3 days apart): Five
of 10 men with average total T less than 250 ng/dL on the initial two samples had average T levels greater than 250 ng/dL
but none had average levels greater than 300 ng/dL on repeated blood sampling over the following 6 months. Similarly, in a
clinical trial of T treatment, 30% of men found to be mildly hypogonadal (with a serum total T of less than 300 ng/dL), based
on a single blood sample at a screening visit, were found to have normal total T levels on repeat blood sampling at a subsequent
baseline visit.[10] These findings support the importance of confirming low serum T levels on at least two occasions before diagnosing biochemical
androgen deficiency, especially when T values are only slightly below the normal range.
The threshold T level, below which signs and symptoms of androgen deficiency occur and T replacement is beneficial, is not
known and varies among individuals with age and comorbid conditions, and among affected target organs. So, there is no absolute
value of total T below which clinical androgen deficiency or hypogonadism can be confirmed in all patients.
In general, clinicians should use the lower limit of the normal range for total T (for example, less than 300 ng/dL) established
for a specific assay in the laboratory to confirm biochemical androgen deficiency in men with consistent clinical manifestations.
Normal ranges, however, vary considerably as a result of disparities among the populations that are used to establish them
and assay differences. In some automated clinical assays, the lower end of the normal range for total T is 170 to 200 ng/dL,
which is significantly lower than the 300 ng/dL limit established over the past 30 years using traditional radioimmunoassay
methods, with or without chromatography.[7] Clinicians should question the methodology used by the laboratory or the population upon which statistical ranges are established
any time the lower limit of the normal range for total T assays is below 280 to 300 ng/dL.
Free and Bioavailable T Levels
Alterations in SHBG concentrations occur commonly, especially in men with multiple comorbidities and those taking certain
medications (Table 2). Moderate obesity, nephrotic syndrome, androgens, and anabolic steroids commonly lower SHBG levels and
therefore decrease total T levels. Aging, hepatitis, hepatic cirrhosis, and anticonvulsants commonly raise SHBG levels, increasing
total T levels. In men who have conditions or take medications known to alter SHBG concentrations, free or bioavailable T
assays, which are not affected by changes in SHBG concentrations, should be ordered to confirm biochemical androgen deficiency.
-
-
Calculated free T (computed from total T and SHBG concentrations, using published algorithms) and free T measured by equilibrium
dialysis are accurate methods to assess free T levels and are unaffected by alterations in SHBG.[11] Equilibrium dialysis is considered the gold standard method for measuring free T, but there is excellent concordance between
such measurements and calculated free T levels. Calculated bioavailable T (also computed from total T and SHBG measurements)
and bioavailable T measured by ammonium sulfate precipitation are accurate methods for determining the level of T that is
available to tissues for biological action. Unfortunately, most clinical laboratories do not measure free or bioavailable
T using these accurate methods, so care must be taken to send samples to a reputable commercial reference laboratory.
Automated, direct, free T measurements that rely on analog methods are available in many local clinical laboratories, but
they should not be used because they are affected by alterations in SHBG levels and are not concordant with the equilibrium
dialysis method.[11,12] Also, since analog-based free T methods have recently been shown not to measure free T in the absence of proteins,[13] they offer no advantage over total T measurements.
Conditions That Transiently Lower T Levels
T measurements should not be performed during acute illness, temporary use of certain medication regimens (for example, those
involving CNS-active medications, opioids, or glucocorticoids), use of recreational drugs, states of nutritional deficiency
(for example, starvation as a result of anorexia or an eating disorder), or during extreme exercise, because these conditions
can lower serum T concentrations transiently. A careful history and clinical evaluation is needed to exclude these conditions.
T measurements should be delayed until after recovery from illness, discontinuation of interfering medications or recreational
drugs, and recovery from nutritional deficiency. When interfering medication regimens are not temporary (for example, when
opioid pain medications are prescribed for a chronic pain syndrome), serum T levels may be persistently low and associated
with clinical manifestations of androgen deficiency. In such situations, T measurement may be appropriate.
Screening for and Case Detection of Androgen Deficiency
In some, but not all, epidemiological studies, low T levels (either endogenous or induced by GnRH agonist/antiandrogen therapy)
have been associated with important clinical outcomes, including metabolic syndrome and diabetes mellitus,[14,15] cardiovascular disease and related mortality,[15-18] fractures,[19] falls and reduced physical performance,[20,21] depression,[22,23] mild cognitive impairment and Alzheimer's disease,[24,25] and total mortality.[16,17,26] The long-term benefits and risks of T treatment on these important clinical outcomes are not known. Furthermore, there is
a lack of consensus on case definition and no effective screening strategies. Screening for androgen deficiency in the general
population, therefore, is not justified.
Case detection by measurement of T levels, on the other hand, should be considered within populations of men who have clinical manifestations of androgen deficiency that might improve with T therapy and have been affected by conditions or events that are associated with a high prevalence of androgen deficiency. Such conditions and events include: hypothalamic/pituitary disease; HIV-associated weight loss; chronic organ failure (kidney, liver, lung, and heart failure); type 2 diabetes mellitus; low trauma fracture or osteoporosis; infertility; and long-term use of medications that lower T levels—for example, opioids, glucocorticoids, CNS-active medications, or ketoconazole. There is, however, limited or no evidence regarding the benefits and risks of T treatment in such populations.
Distinguishing Primary from Secondary Hypogonadism
Male hypogonadism may be due to testis dysfunction (primary hypogonadism) or secondary to hypothalamic or pituitary dysfunction
(secondary hypogonadism), resulting in inadequate stimulation of the testes by pituitary gonadotropins, luteinizing hormone
(LH), and follicle-stimulating hormone (FSH). In addition to confirmation of unequivocally low serum T levels in men with
clinical manifestation of androgen deficiency, concomitant measurement of serum LH and FSH levels should be performed to classify
androgen-deficient men as having either primary or secondary hypogonadism.[1]
In men with primary hypogonadism, the serum T level is low in association with high LH and FSH concentrations (Table 3). In
men with secondary hypogonadism, a low serum T level occurs in conjunction with normal or low LH and FSH levels (Table 4).
Normal gonadotropin levels are inappropriate in the setting of low T levels and reduced negative feedback of T, and suggest
hypothalamic or pituitary dysfunction.
-
-
Hypogonadism may be due to defects in both hypothalamic/pituitary and testis function, resulting in a combined primary and
secondary hypogonadism (as can occur with aging, chronic illness, glucocorticoid use, alcohol abuse, and hemochromatosis).
In such cases, a hormonal pattern consistent with either primary or secondary hypogonadism usually predominates. Most men
with hemochromatosis, for example, have a hormonal pattern of secondary hypogonadism with low T, LH, and FSH levels, despite
also having an element of primary testicular dysfunction.
It is important to determine whether hypogonadism is primary or secondary because the latter may be caused by hypothalamic/pituitary
tumors or by infiltrative diseases that can cause tumor mass effects (such as headache, visual field defect, visual impairment,
or cerebrospinal fluid rhinorrhea), excessive secretion or deficiency of other anterior pituitary hormones or hypothalamic
dysfunction (for example, diabetes insipidus), any of which would require medical management in addition to T therapy. Also,
some functional causes of secondary hypogonadism may be transient or reversible (for example, acute illness, medications such
as opioids or high-dose glucocorticoids, and nutritional deficiency). Finally, infertility in men with secondary hypogonadism
is usually treatable with gonadotropin or GnRH therapy.
Further Evaluation of Hypogonadism
In men with hypogonadism who desire fertility, three seminal fluid analyses should be ordered over 2 to 3 months to assess
sperm counts, motility, and morphology. In men with severe androgen deficiency or low trauma fracture, assessment of BMD should
be evaluated by a dual-energy X-ray absorptiometry scan. T treatment has been demonstrated to increase BMD in hypogonadal
men,[27] but there is no evidence that T prevents fractures. T should not be used as primary treatment for osteoporosis, but conventional
therapies (such as bisphosphonates) have demonstrated effect in preventing fractures in hypogonadal men with osteoporosis.[28]
In men with secondary hypogonadism, further evaluation to identify the etiology of hypothalamic/pituitary dysfunction should
be undertaken on an individual basis. This may include: serum prolactin and iron studies to exclude hyperprolactinemia and
hemochromatosis, respectively, or assessment of anterior pituitary function to exclude panhypopituitarism, if clinically indicated
by manifestations of pituitary hormone deficiency or excess, or if occurring in conjunction with severe androgen deficiency
(with a T level less than 150 ng/dL). Also, magnetic resonance imaging (MRI) of the hypothalamus and sella turcica should
be performed to identify a hypothalamic or pituitary tumor or infiltrative disease in men with very low levels of T (less
than 150 ng/dL),[29] LH, and FSH, or isolated LH or FSH elevation, panhypopituitarism, persistent and severe hyperprolactinemia unexplained by
medications, or signs or symptoms of tumor mass effect. Men with idiopathic hypogonadotropic hypogonadism should be examined
for dysmorphic features such as morbid obesity, short stature, anosmia, cleft lip or palate, polydactyly, or kidney abnormalities,
which may be associated with Kallmann syndrome, Prader Willi syndrome, or one of many complex genetic disorders characterized
by this condition. Morbidly obese men with low free T levels should be evaluated for obstructive sleep apnea, which may contribute
to hypogonadotropic hypogonadism.
In men with primary hypogonadism and clinical features of Klinefelter syndrome (very small, firm testes; gynecomastia; infertility;
azoospermia; and disproportionately long legs), a chromosomal karyotype may be obtained to confirm the diagnosis (47,XXY karyotype)
and patients can be referred for genetic counseling.
Case Discussion
There are a number of potential risk factors and etiologies for ED in the patient scenario presented previously. These include
diabetes mellitus (complicated by neuropathy), nephropathy, probable peripheral vascular disease, smoking, hypertension, hyperlipidemia,
medications that may contribute to ED (for example, diuretics, beta-blockers, and opioids), and possibly mild depression.[30]
Low serum T levels are commonly associated with ED.[31] In addition, the patient's codeine use and recent pulse glucocorticoid therapy may have contributed to the low T level, since
both opioid and glucocorticoid use lower serum T levels. Also, in addition to ED, the patient has other manifestations that
are consistent with androgen deficiency, including excessive fatigue and reduced activity, gynecomastia, and reduced muscle
bulk, though these manifestations are relatively nonspecific.
While it is unlikely that T treatment alone will adequately treat ED in a diabetic patient with moderately low serum testosterone
levels, the addition of T therapy to such conventional ED treatments as phosphodiesterase type 5 inhibitors may restore sexual
function in hypogonadal men with ED.[32,33] Furthermore, T treatment may benefit other body functions, for example, increasing muscle mass and strength, as well as BMD.
A diagnosis of hypogonadism in men who present with ED, therefore, may affect overall health management, and it is appropriate
to evaluate such patients for androgen deficiency. In the case presented previously, serum free or bioavailable rather than
total T levels should be used to evaluate whether androgen deficiency is present because moderate obesity, proteinuria with
mild hypoalbuminemia, and a recent pulse of prednisone would decrease SHBG levels, lowering total T concentrations.
The patient's repeat total T was low at 240 ng/dL (normal, 280 to 800 ng/dL). The SHBG level, however, was in the low-normal
range at 25 nmol/L (normal, 10 to 80 nmol/L). Calculated free and bioavailable T were both normal at 53 pg/mL (normal, 34
to 194 pg/mL) and 128 ng/dL (normal, 84 to 402 ng/dL), respectively. Normal free and bioavailable T levels are consistent
with eugonadism rather than hypogonadism, and the patient's low total T is attributable to low-normal SHBG levels, likely
due to moderate obesity, urinary protein loss, and recent glucocorticoid use.
Dr. Matsumoto is a professor in the department of medicine at the University of Washington School of Medicine, Seattle, WA,
as well as the associate director of the Geriatric Research, Education and Clinical Center and the director of the clinical
research unit at the VA Puget Sound Health Care System, all in Seattle, WA.
References
-
