Introduction

Decreases in serum testosterone occur naturally as part of the aging process. These hormonal changes are associated with a variety of clinical outcomes, including alterations in sexual function. This paper reviews recent research with a focus on sexual health in elderly men.

Erectile dysfunction (ED) is progressively being recognized as an early warning sign of cardiovascular and small vessel arterial disease. In addition, low serum testosterone has been associated with development of metabolic syndrome, a precursor for diabetes mellitus [1]. For example, in a convenience sample of 241 men with ED, 41.5 % had testosterone deficiency, 36.5 % had metabolic syndrome, and 19.5 % had both [2]. A similar study of 1,046 men with mean age 55.8 ± 8.4 years (range 41–93), identified prevalence of moderate to severe ED at 20.0 %; 16.1 % had hypogonadism with serum testosterone <10.4 nmol/L; and 31.3 % had metabolic syndrome [3]. Hypertension, hyperglycemia, heart disease, and low high-density lipoprotein levels were all independent predictors of ED, and all metabolic syndrome elements were independent predictors of hypogonadism. These data support the argument that these entities exist as a clinical triad in a sizable proportion of men. Because of this, the role sexual health assessment as part of a complete evaluation for men’s health is becoming better appreciated as a measure of overall health, and an early indicator of comorbidity and pathophysiology [4].

Definitions

Hypogonadism is defined by the reduction in serum testosterone and other circulating androgens. This can be primary, with changes arising directly in the testicles, or secondary, with changes occurring in the hypothalamic-pituitary-testicular axis. When it occurs in elderly men, it is typically referred to as late onset hypogonadism. The reductions in serum testosterone levels tend to occur gradually, and clinical symptoms can be slow to manifest in many cases. The term ‘andropause’ has been used to describe this process analogous to menopause in women. However, the term is not particularly accurate, because not all men develop symptomatic hypogonadism with advancing age.

One of the challenges for both research and clinical care in older men is lack of a clearly accepted definition of hypogonadism based on serum testosterone levels. The Endocrine Society has published guidelines regarding evaluation and treatment of men with androgen deficiency. Most of these recommendations are based on lower limits of normal testosterone established in young healthy men, of 280–300 ng/dL (9.8–10.4 nmol/L) [5•]. They recommend against routine screening of the general male population, but suggest targeted evaluation and intervention based on associated symptoms.

A few studies have sought to better characterize the condition, particularly in the context of sexual effects. Clinical symptoms include loss of morning erections, low sexual desire and erectile dysfunction, in conjunction with low levels of serum testosterone (< 8–11 nmol/L) [6]. In a study of 675 healthy men, Lackner et al. measured serum testosterone levels and compared them to a variety of sexual health parameters using validated survey instruments [7•]. They identified that although decreased serum testosterone levels were associated with multiple symptoms, age was a strong confounding variable. In fact, no definitive testosterone levels were significantly identified for any sexual health parameters measured, except for psychological symptoms including depression, after correcting for subject age. Erectile dysfunction and other sexual problems were reported even in men with normal serum testosterone. It is well known that there are diurnal variations in testosterone levels, and the most accurate specimens are generally obtained in the morning.

Etiology and Effects of Hypogonadism in Elderly Men

Testosterone levels decrease with advancing age in men. This is due in part to decreased testosterone production from testicular Leydig cells [8]. Levels of luteinizing hormone (LH) tend to remain unchanged or slightly increase with normal aging. In addition, there is an increase in sex-hormone binding globulin (SHBG) with aging, which results in a greater percentage of protein-bound testosterone, and lower levels of circulating free testosterone. Increased aromatization of testosterone in adipose tissue with conversion to estradiol may also be involved [9].

Diminished testosterone leads to reductions in muscle and bone mass and overall physical function in older men. There has been recent interest regarding whether hormone replacement in select older adults might help prevent or potentially reverse these effects. The question remains as to whether this might help improve not only these physical and functional outcomes, but also overall longevity [10].

The associations of hypogonadism on sexual health have been somewhat more controversial. Although there has been relatively consistent data linking low testosterone to diminished libido, demonstrated effects on erectile function have been variable. A study of 120 men who underwent comprehensive geriatric assessment including measurements of erectile function and lower urinary tract symptoms (LUTS) using validated instruments showed a relationship in these parameters [11]. Mean age was 73.8 ± 5.9 years, and there was a significant reduction in both free and total testosterone with advancing age (p = 0.021). Decreased serum testosterone levels were correlated with worse erectile function as measured by the International Index of Erectile Function (IIEF) (r = 0.66, p < 0.001). LUTS was also correlated, although more weakly, with hypogonadism (r = −0.23, p = 0.016). Other measurements of health including obesity, osteoporosis, cardiovascular disease, cognition, and activities of daily living (ADLs) were not statistically different between normal testosterone and hypogonadal groups. Results from Demir et al. supported these findings, and also showed ED was the strongest predictor of severe LUTS in a series of 190 men [12]. Prevalence of metabolic syndrome was also statistically higher in men with severe LUTS (p = 0.009).

In contrast, Marberger et al. reported little association between serum testosterone levels and sexual function measured using validated survey instruments in the Reduction by Dutasteride of Prostate Cancer Events (REDUCE) Trial [13]. In this 4-year, prospective, randomized, double-blind, placebo-controlled trial which examined safety and efficacy of daily dutasteride therapy in 8,000 men, testosterone levels and total prostate volume were not found to be independent predictors of any sexual function variables assessed.

Another study in 1,776 men in China examined hormonal status and other predictive variables with regard to ED [14]. Levels of sex hormone binding globulin (SHBG) increased proportionally with age, and were statistically associated with greater severity of ED (p < 0.001). Total testosterone was associated with ED after adjusting for age (OR 1.02, 95 % CI 1.00–1.04).

Epidemiology

Estimates suggest more than 4.5 million elderly American men may have late onset hypogonadism [15]. An internet survey of over 10,000 men in the United States and United Kingdom showed 80 % reported moderate or severe scores consistent with hypogonadism [16••]. Although mean respondent age was 52, the researchers noted that many men in their 40s reported similar symptoms. These data suggest that hypogonadism may be underreported and underdiagnosed in older men.

Sexual Health

Libido

Hypogonadism has frequently been identified as a cause of impaired libido. Hyde et al. reported results of a cross-sectional, population-based study of 3,274 men aged 75–95 (mean 82) in Western Australia [17•]. Sexual health complaints were highly prevalent in this cohort with 49.4 % reporting erectile dysfunction, 47.7 % loss of interest in sex, 38.7 % with inability to reach orgasm, and 20.4 % with anxiety about sexual performance. Overall, 72.0 % of the men reported at least one sexual problem. Associations with other comorbid diseases, depression and insomnia were also identified. Low testosterone levels were associated with lack of libido, but not other sexual complaints including ED in this group.

Erectile Function

Recent research has identified erectile dysfunction can be an early sign of underlying comorbidities in older adult men. One of the more common is diabetes. Liu et al. examined the relationships between overactive bladder (OAB) and ED in a group of 453 diabetic men attending an outpatient clinic in Taiwan [18]. Mean age was 60.6 years. Overall, 25.4 % had OAB, 10.2 % also leaked urine (OAB wet), and 81.9 % had ED, with 28.3 % reporting severe ED. Significant associations between OAB and ED were identified in this group, with urgency incontinence having the strongest relationship to ED. The relationship between clinical factors and diabetes is complex, and may be related to changes in microvasculature which influence both erectile and bladder function.

Metabolic syndrome, often considered a precursor to diabetes mellitus, has also been associated with hypogonadism and ED. Fillo and colleagues reported an ED prevalence of 73 % in 167 men with abdominal obesity [19]. Of this group with ED, 84 of 122 patients (68.9 %) had serum testosterone levels <14 nmol/L, and 49 (40.2 %) had levels <10 nmol/L. Rates of diabetes mellitus, hypertension and dyslipidemia were significantly higher in those with more severe hypogonadism. These are potentially modifiable risk factors that can often be treated with weight loss, diet, smoking cessation, and other lifestyle modifications.

Guay et al. examined nearly 1,000 men and compared serum testosterone levels with various comorbid conditions [20]. They identified significant relationships between hypogonadism and hypertension, tobacco abuse, sleep apnea, and work stress. The researchers concluded a wide variety of comorbidities are associated with hypogonadism, and that men with underlying conditions should have a serum testosterone level checked, particularly if they have any sexual symptomatology.

The impact of hypogonadism on erectile function has been somewhat controversial. In a national observational study of 1,340 men aged 22–81 years in Spain, age was the most important predictive factor for development of erectile dysfunction [21]. Of the total cohort, 33 % had a total serum testosterone <12 nmol/L, and 13.5 % had a level <8 nmol/L. In this study, there appeared to be a direct relationship between diminished serum testosterone and clinical severity of ED. However, obesity, smoking and diabetes were also highly prevalent in this population, and certainly contributed to the multifactorial nature of the disease.

In a randomized, placebo-controlled trial, Spitzer et al. examined outcomes for erectile function in a group of men with ED and hypogonadism [22]. Outcomes were assessed using the erectile function domain (EFD) of the International Index of Erectile Function (IIEF). Subjects who received a combination of transdermal testosterone and sildenafil had statistically equivalent improvement compared to control subjects who received sildenafil alone. One study limitation was that use of testosterone alone to improve erectile dysfunction was not examined. A similar trial with combined testosterone replacement and tadalafil yielded comparable results [23]. Men who were more severely hypogonadal had greater improvements overall compared to men with higher baseline testosterone levels. In a pilot study of 29 men aged 36–75 (mean 59), Garcia et al. showed administration of testosterone improved libido in hypogonadal men with metabolic syndrome who had not responded well to oral phosphodiesterase type-5 inhibitors [24]. There was a trend toward better erectile function, which reached a plateau at 30 weeks of combination treatment.

A small study examined the impact of testosterone replacement not only on sexual function in men taking hormones, but also on their female sexual partners [25]. Libido in hypogonadal men improved with normalization of serum testosterone. Of note, women partners also reported improvements in sexual function with less pain and more satisfaction following hormonal replacement in the men.

Iacono and colleagues studied the relationship between serum testosterone and anatomic changes in a group of 47 men with ED stratified into older (≥65) and younger (<65) age groups [26]. Erectile function was measured using IIEF and nocturnal penile tumescence testing, and all subjects underwent corpora cavernosum biopsies. On histological examination, corporal fibrosis with >52 % collagen deposition was identified in a substantial portion (74 %) of men with organic ED. Hypogonadism was associated with these findings, suggesting a possible relationship between endocrine and structural factors. Causality could not be demonstrated because of study design, but this raises questions about the possible utility of testosterone replacement to prevent corporal fibrosis and other structural alterations in penile anatomy associated with ED and aging.

Nam and colleagues examined the relationships between hormone levels and clinical characteristics of Peyronie’s disease in a case–control study [27]. Men with Peyronie’s’ disease and hypogonadism had worse penile curvature, larger penile plaques, and worse overall erectile function compared to men with normal serum testosterone concentrations. These data were corroborated in a study of 106 men with Peyronie’s disease and 99 healthy controls [28]. Hypogonadal men had worse penile curvature and larger penile plaques compared to men with normal serum testosterone. These data support arguments that low testosterone may exacerbate both structural and functional outcomes in men with Peyronie’s disease. It was interesting that administration of exogenous hormone was associated with better efficacy of intraplaque verapamil therapy compared to men who received placebo instead.

The relationship between hormonal status, comorbidity and ED is complex. Collins et al. examined a cohort of 90 men with moderate to severe chronic obstructive pulmonary disease (COPD) [29]. Of this group, 40 % were hypogonadal based on validated questionnaires. Measurements were taken at baseline, and subjects were followed a median of 4.8 years. Hypogonadism, depressive symptoms, and lack of a sexual partner were all identified as independent predictors of ED, which in turn was associated with worse quality of life. Mortality was linked to severity of COPD and comorbidities (p = 0.006), but not to ED (p > 0.05).

Orgasm and Ejaculation

The role of testosterone concentration on orgasm and ejaculation has not been extensively studied. Masturbation can be an important form of sexual expression, particularly in elderly men who may not have a regular sexual partner. A prospective cohort study of 2,786 older men found the overall prevalence of at least monthly masturbation was 61.9 % [30]. After adjusting for age, there was a direct positive correlation between serum testosterone levels and masturbation activity. Increased rates of guilt feelings were reported in men with hypogonadism compared to those with normal hormonal levels. It is important for clinicians to include a discussion of masturbation as a viable form of sexual expression in elderly men, rather than just focusing on penetrative sexual activity.

Depression and Mental Health

Depression and other affective disorders have been linked to hypogonadism in elderly men [31]. It is hypothesized this may be due to hormonal influences on secretion and binding of serotonin and other psychoactive neurotransmitters. A recent randomized, double-blind, placebo-controlled study examined effects of testosterone replacement in older men with dysthymia or minor depression. Subjects were treated for 12 weeks with either testosterone or placebo gel, followed by an additional 12 weeks of open-label extension in which all subjects received testosterone [32]. Using validated measures of depression, subjects on testosterone had statistically significant improvements compared to those receiving placebo (p = 0.024), and a higher rate of remission of subthreshold depression (52.9 %) compared to placebo (18.8 %) (p = 0.041). At conclusion of the open-label extension, men in the original treatment group had sustained improvements, and the prior control group improved to equivalent levels. Although the study did not specifically examine the relationship with sexual health, it is conceivable improvements in mental health could be associated with potential improvements in libido and sexuality.

Cardiovascular Health and Sexuality

Low testosterone has been associated with an increased risk for development of coronary artery disease in older men [33]. This interaction is complex and multifactorial. Nitric oxide (NO) has been identified as an important mediator in both cardiovascular health and erectile function. The phosphodiesterase type-5 inhibitors (PDE5i) are an important class of drugs designed to increase NO levels in tissues and improve sexual function. The role of testosterone in biochemical production and signaling of NO will continue to be an important focus for translational research [34].

Hormonal replacement in elderly men has been shown to yield potential cardiovascular and hematologic benefits. Alterations in erythrocyte membrane lipids can enhance red cell flexibility, which could potentially reduce thrombosis risk [35]. Because exogenous testosterone administration can lead to increased red cell production, patients should be monitored closely for development of erythrocytosis, which could lead to stroke and other cerebrovascular disease. Another study demonstrated that replacement of dehydroepiandrosterone (DHEA) led to clinical improvements in arterial stiffness, which could also reduce cardiovascular disease in older adults [36]. Similarly, testosterone replacement appeared to increase pulse width velocity, a measure of arterial elasticity, in a small group of men treated for hypogonadism [37].

Frailty

Frailty is a common syndrome in older adults, and is characterized by at least three of the following parameters: decreased grip strength, diminished gait speed, easy fatigue and exhaustion, unintentional weight loss, and low levels of physical activity [38]. Sarcopenia is a key feature of the condition, and this has been associated with hypogonadism in elderly men [39]. The role of endocrine biomarkers in diagnosis of frailty is a topic of ongoing research.

Testosterone replacement has been shown to reduce sarcopenia associated with hypogonadism in elderly men. Although prior research has shown that administration of growth hormone (GH) alone does not appear to improve muscle strength, combined administration of GH and testosterone may be synergistic. A recent study exploring this concept concluded additional research is needed to examine long-term efficacy and safety of combination therapy in elderly men [40]. Atkinson et al. demonstrated in a group of 30 men with an established diagnosis of either frailty or intermediate frailty that testosterone administration helped preserve muscle mass and prevent additional deterioration of muscle structure and function [41]. This raises questions about testosterone as a modifiable risk factor for frailty in elderly men [42].

Clinical Hormone Replacement

Several different forms of testosterone replacement are currently available. These include intramuscular injections, topical patches, transdermal gels, buccal absorbable compounds, and subcutaneous implantable pellets. Oral administration should be avoided due to potential hepatotoxicity. Administration of exogenous testosterone in elderly men has been associated with improvements in bone density and cognition and mood. However, documented improvements in cardiovascular health, frailty and sexuality are somewhat more controversial [43].

Reported rates of prescription for testosterone products have increased by 170 % in the past 5 years [6]. A multinational survey of physicians indicated decreased libido, depression and obesity were considered clinical indicators of hypogonadism [44]. Overall, 70 % of doctors surveyed considered severity of patient symptoms more important than serum levels in making the decision to prescribe testosterone replacement. However, 11 % of potential patients did not receive hormonal therapy due to concerns about possible prostate disease. This indicates there is still a need for additional professional education about efficacy and safety of testosterone replacement, and methods for monitoring patients undergoing treatment.

The American Geriatrics Society (AGS) recently published an update of the Beers Criteria, a listing of potentially inappropriate medications in older adults [45••]. Testosterone and methlytestosterone were both included in this recommendation of medications to generally avoid in older adults. The rationale included potential for cardiac problems, and contraindications for men with an established personal history of prostate cancer. However, they did indicate these could be useful for treatment of moderate to severe hypogonadism in elderly men, and treatment should be individualized based on clinical needs.

Bhattacharya et al. examined effects of twelve months of testosterone replacement in a cohort of 849 hypogonadal men [46••]. Subjects were stratified into younger (< 65) and older (≥ 65) groups. The groups were similar with regard to duration of hypogonadism prior to treatment (mean 1 year), compliance with therapy, and both free and total testosterone levels at baseline and 12 months. Although prostate specific antigen (PSA) levels were higher in the older group at 12 months (2.18 ± 2.18 ng/mL versus 1.14 ± 0.84 ng/mL), this was not statistically significant (p = 0.1), and PSA levels were statistically equivalent between groups at follow-up. Testosterone replacement was well tolerated in both younger and older hypogonadal men. In this same cohort, transdermal testosterone replacement therapy was associated with improvements in multiple parameters of sexual function using the Brief Male Sexual Function Inventory [47].

Depression and other affective disorders have been linked to sexual dysfunction. Testosterone replacement has been shown to improve mood and reduce depressive symptoms [48•]. Although antidepressive medications can often be effective to improve mood, some classes of these drugs have been shown to exacerbate reduction of libido and the occurrence of ED. There is concern that concomitant use of antidepressant medications could also reduce effects of testosterone replacement. In a recent prospective, double-blind, placebo-controlled study of 63 men with major depressive disorder who were taking serotonergic antidepressants, Amiaz et al. showed simultaneous administration of testosterone led to significantly improved erectile function compared to placebo (p = 0.001) [49]. Even ejaculatory function significantly improved in this group compared to controls (p ≤ 0.03). Results were maintained even after adjusting for degree of improvement in depressive symptoms.

The combination of diet and exercise with testosterone replacement has been examined. One intervention study examined 56 men (mean age 52.3 ± 7.8 years) with partial androgen deficiency syndrome [50]. Subjects received intramuscular testosterone injections twice weekly, a low-glycemic balanced nutrition diet, and participated in 30–60 minutes of structured daily exercise. Statistically significant reductions in weight, body mass index (BMI), total body fat, fasting glucose, and total cholesterol were observed at 6–18 months. While this study was designed to examine effects of this type of therapy on cardiac health, it raises important questions about possible utility to improve other aspects of men’s health with aging including sexual function. It will be interesting to see if patients are able to sustain compliance with interventions and improvements for longer duration.

There has been interest in possible use of phytotherapies for hormonal manipulation in elderly men. Although popular among the lay public, there has been relatively limited scientific research on the efficacy and safety of herbal compounds to treat sexual dysfunction due to hypogonadism. Shah et al. published a trial of 78 men randomized to receive either a proprietary polyherbal preparation versus placebo [51]. After 12 weeks, men in the group receiving herbal preparation had significantly improved global and subscore measurements on all IIEF components compared to controls (all, p < 0.001). Patients and their sexual partners also had significant improvements based on the Erectile Dysfunction Inventory of Treatment Satisfaction (EDITS) scores compared to controls (all, p < 0.001). Thirty-five of 39 men (90 %) in the treatment group but only one of 39 in the placebo group (3 %) wanted to continue their assigned therapy. The study did not indicate specific phytotherapeutic agents utilized. In another trial, younger male subjects (age 30–50 years) were given a proprietary product containing pine bark extract and l-arginine aspartate [52]. Subjects receiving active therapy had statistically significant improvements in erectile function and serum testosterone levels compared to controls. Results were maintained through 6 months. However, potential utility in elderly men is unknown as they were not included. The results of these studies are intriguing, and suggest additional investigation of phytotherapeutic agents may be warranted.

The potential risk of incident prostate cancer is one of the main concerns about testosterone replacement in elderly men. Feneley and Carruthers published a longitudinal study of 1,365 men with symptomatic hypogonadism who received testosterone replacement [53]. Mean age was 55 (range 28–87), and some subjects had been monitored up to 20 years. They identified 14 new cases of prostate cancer, which accounted for one case per 212 years with a total of 2,966 man-years of therapy. This incidence of prostate cancer was identical to the general population over the same time period. Mean time to diagnosis was 6.3 years, and all tumors were clinically localized and amenable to treatment with curative intent. The authors concluded that with proper monitoring, testosterone therapy appears to be safe for treatment of symptomatic hypogonadism. These results were supported by a retrospective review of 2,291 men seen in a single outpatient urology clinic [54]. Serum testosterone levels did not appear to be directly related to PSA except in the most severely hypogonadal men where PSA was also quite low. The results were maintained when stratified by age, and were only statistically significant in young men. Khera et al. demonstrated administration of testosterone in severely hypogonadal men resulted in greater elevation of PSA levels compared to men who were less severely hypogonadal at baseline [55]. Clinicians must be aware of this and monitor patients carefully after initiation of androgen replacement therapy.

Use of exogenous testosterone replacement in men with a personal history of prostate cancer is quite controversial. Leibowitz reported a series of 96 men given testosterone replacement after prostate cancer treatment [56]. Of this group, 41 (42.7 %) had PSA progression while on hormonal therapy, but radiographic progression was only identified in 7 of these men. The majority (58.3 %) ultimately stopped testosterone replacement, and 59 % had a subsequent reduction in PSA. A total of 31 men (32.3 %) remained on testosterone replacement at 36.7 months. Data from the Baltimore Longitudinal Study of Aging suggested increased serum free testosterone was associated with an increased risk of aggressive prostate cancers even in elderly men [57]. Data from these studies raise questions about utility of hormonal replacement to relieve hypogonadal symptoms in men with a history of prostate cancer, and whether potential benefits really outweigh potential risks.

Future Research Recommendations

Relationships between hypogonadism, LUTS and ED are intriguing and warrant additional investigation. Orabi et al. have postulated four pathophysiological mechanisms that might explain this interaction [58]. These include changes in bioavailability of nitric oxide, hyperactivity of α1-adrenergic receptors, pelvic atherosclerosis, and sex hormones including testosterone. They argue additional research on the relationship between these conditions is needed, and that it may better elucidate the role of testosterone in these processes. The potential utility of testosterone and other biomarkers as indicators for development of frailty are also targets for future research work [39, 42]. Finally, the use of selective androgen receptor modulators (SARMS), which could influence functional outcomes by targeting the androgen receptor while simultaneously avoiding some potential adverse events associated with direct testosterone replacement, deserve additional investigation [59].

Conclusion

Late onset hypogonadism is a common clinical condition in elderly men. It is most likely underdiagnosed and underreported based on epidemiological data. It has been linked to a number of negative outcomes including osteoporosis, heart disease, depression, sarcopenia, and development of frailty in the geriatric population. Effects on sexual health can be significant, and studies have shown testosterone replacement therapy may be of benefit to improve a variety of sexual symptoms. Potential risks and benefits need to be carefully considered and therapeutic decisions made individually with each patient. Appropriate monitoring must be continued for men undergoing hormonal replacement.