Androgens and male aging: current evidence of safety and efficacy

Abstract

Many signs of aging, such as sexual dysfunction, visceral obesity, impaired bone and muscle strength, bear a close resemblance to features of hypogonadism in younger men. The statistical decline of serum testosterone in aging men is solidly documented. It has been presumed that the above features of aging are related to the concurrent decline of androgens, and that correction of the lower-than-normal circulating levels of testosterone will lead to improvement of symptoms of aging. But in essence, the pivotal question whether the age-related decline of testosterone must be viewed as hypogonadism, in the best case reversed by testosterone treatment, has not been definitively resolved. Studies in elderly men with lower-than-normal testosterone report improvement of features of the metabolic syndrome, bone mineral density, of mood and of sexual functioning. But as yet there is no definitive proof of the beneficial effects of restoring testosterone levels to normal in elderly men on clinical parameters. Few of these studies meet as yet rigorous standards of scientific enquiry: double-blind, placebo-controlled design of the study. The above applies also to the assessment of safety of testosterone administration to elderly men. There is so far no convincing evidence that testosterone is a main factor in the development of prostate cancer in elderly men and guidelines for monitoring the development of prostate disease have been developed. It is of note that there are presently no long-term safety data with regard to the prostate. Polycythemia is another potential complication of testosterone treatment. It is dose dependent and can be managed with dose adjustment.

Introduction

When Prof. Alex Vermeulen from Gent/Belgium in the 1970s started his pioneering work on the decline of testosterone levels in aging men, his work seemed to most professionals only of theoretical interest (for review:¹). But further studies have substantiated that, with age, a significant percentage of men over the age of 60 years have serum testosterone levels below the lower limits of normal for young adult men (aged 20–30 years) ^{2, 3}. Now, 30–40 years later, the pathophysiological implications of this decline, sometimes amounting to outright testosterone deficiency, have become clearer.

Until a decade ago the ailments of elderly men, such as atherosclerosis, hypertension, diabetes mellitus, lower urinary tract symptoms and erectile dysfunction (ED), were regarded as distinct diagnostic/therapeutic entities but there is a growing evidence that these entities are not disparate and, to improve the health of the aging male, require an integral approach. There is an interdependence between the metabolic syndrome, ED and patterns of testosterone in aging men ^{4, 5}. The main features of the metabolic syndrome are abdominal obesity, insulin resistance, hypertension and dyslipidemia, significant factors in the etiology of erectile functions. The metabolic syndrome is associated with lower-than-normal testosterone levels ^{6, 7}. Testosterone is a determinant of glucose homeostasis and lipid metabolism ⁸ and plays also a significant role in the development and maintenance of bone and muscle mass. Testosterone is not only a factor in libido but exerts also essential effects on the anatomical and physiological substrate of penile erection ⁹. The effects of phosphodiesterase type 5 (PDE5)-inhibitors are suboptimal in the presence of hypogonadal values of circulating testosterone ⁹. A relationship between lower urinary tact symptoms (LUTS) and circulating levels of testosterone has been difficult to prove but the association of LUTS with the metabolic syndrome and with ED is well established in the literature ¹⁰. Testosterone treatment of hypogonadal not only improves features of the metabolic syndrome but also improves LUTS ¹¹. With these recent insights, the health problems of elderly men must be placed in a context that allows an integral approach. On the basis of recent insights, diagnosis of testosterone deficiency should be part of the diagnostic work-up of the above conditions.

Late onset hypogonadism (LOH) vs. classical hypogonadism

Many signs and symptoms of aging, such as sexual dysfunction, visceral obesity, impaired bone and muscle strength, mood disturbances bear a close resemblance to features of hypogonadism in younger men. In younger men improvement of the symptoms of hypogonadism upon testosterone replacement has been convincingly proven. Presently, the statistical decline of serum testosterone in aging men is solidly documented ^{1, 3}. So, it is tempting to associate the above described features of aging to the concurrent decline of androgen action in the elderly, and to assume that correction of the lower-than-normal circulating levels of testosterone will lead to improvement of signs and symptoms of aging. But in essence, the pivotal question whether the age-related decline of testosterone must be viewed as hypogonadism, i.e. a deficiency of testosterone manifesting itself with clinical signs and symptoms of insufficient androgen action, and in the best case reversed by testosterone treatment, has not been definitively resolved. This point has been eloquently made in a paper by Handelsman and Liu ¹², cautioning against a too facile prescription of testosterone to aging men in the hope to improve, or even reverse, signs and symptoms of aging. The authors argue for well-designed, placebo-controlled double-blind studies documenting unequivocally the benefits of restoring testosterone levels in elderly men to the normal range while also closely monitoring the risks of testosterone administration to aging men. The present studies trying to prove the point have a less-than-optimal design (not blinded, not placebo-controlled) and understandably, in the time frame of the study, usually document surrogate markers of a disease and not clinical endpoints. For instance, do men whose features of the metabolic syndrome improve upon testosterone treatment, have a lower risk of type 2 diabetes mellitus and cardiovascular disease? Even more difficult to prove over a relatively short period of a clinical study: do men whose bone mineral density (BMD) improves upon testosterone treatment have a reduction in fracture rate? To document such an effect, long-term studies are needed.

Who is truly testosterone-deficient in old age?

As indicated above, circulating levels of testosterone decline with aging. The vast majority of men will have circulating testosterone values which are 5%–20% below reference values for men. And it has been difficult to draw a clear dividing line between blood levels of testosterone which represent a hypogonadal and a eugonadal state. In addition, blood testosterone thresholds for androgen deficiency symptoms are highly consistent within a person but differ between people ¹³. Hypogonadal men who receive androgen treatment perceive the threshold for androgen deficiency symptoms at highly reproducible blood testosterone levels. This distinctively individual trigger level for androgen deficiency symptoms differs, however, widely between men. The threshold varies from very low to values above the lower limit of the eugonadal reference range. On average it approximates the lower limit of the eugonadal reference range for young men ¹³. The factors which define this symptomatic threshold are as yet unknown but it is reasonable to assume that genetic polymorphisms of the androgen receptor influencing androgen sensitivity play a significant role ^{14, 15}. The impact of the aging process and acquired chronic disease with aging on the threshold for androgen deficiency symptoms are presently uncharted territory ¹³. Another issue is whether an age-specific reference range of testosterone values is appropriate for diagnosing testosterone deficiency in old age. Most population studies show that obesity and/or disease and lifestyle has a profound impact on values of testosterone ^{16, 17}. There is as present no published evidence that use of age-appropriate reference values offers advantages over use of reference values of the general population but it is an important point in the conceptualization of the desirability of testosterone treatment of elderly men.

Thresholds and dose-response relationships of androgen effects

The concept of threshold values of circulating testosterone for manifestation of androgen deficiency symptoms, is only beginning to be supported by empirical evidence. A couple of studies have suggested a threshold for androgen effects on male sexual function, primarily libido, becoming evident at the lower range of reference values of blood testosterone concentrations ^{18, 19}. These studies were, however, not able to define explicitly such a threshold. Conversely, muscle appears to exhibit linear dose-response relationship to testosterone from below to above the eugonadal reference range for blood testosterone concentrations ²⁰. One study has suggested a dose-response relationship for effects of normalization of circulating testosterone on features of the metabolic syndrome ²¹, but an optimal range for those beneficial effects was not established in this study. Whether linear dose-response or threshold models apply to other androgen-sensitive tissues, such as bone and prostate, psychosexual and cardiovascular effects remains to be determined. So, from recent studies of elderly men it has become apparent that spectrum of complaints of testosterone deficiency cannot be related to a specific threshold of testosterone concentrations, but that thresholds vary with the various symptoms of testosterone deficiency ²². In a cohort of men androgen-related loss of libido or vigor became more prevalent when testosterone concentrations fell below ¹⁵ nmol per liter, while depression and diabetes mellitus type 2 (also in nonobese men) were significantly more present in men with testosterone concentrations below 10 nmol L⁻¹. Symptoms related to androgen deficiency in this study could be subdivided in three independent groups: psychosomatic complaints, metabolic disorders, and sexual health problems. Patients suffering from one of these three groups exhibit distinct features in terms of androgen levels, age, and body mass index. So, complaints are not only linked to androgen levels but age and body mass index carried weight as well in the manifestation of signs and symptoms of androgen deficiency ²². To further complicate the matter of the relationship between testosterone levels on the one hand and symptoms of testosterone deficiency on the other, the authors have drawn attention to the multifactorial impact on certain androgen-related functions ²². ED may serve as an example of a composite dysfunctionality in which, apart from testosterone concentrations, arterial endothelial function, neuronal integrity and psychological factors play pivotal roles ^{23, 24}, almost precluding the establishment of a straightforward relationship between testosterone levels and erectile dysfunction. In the study of Zitzmann et al. ²² ED was identified as a composite pathology of metabolic risk factors, smoking, and depression, and only testosterone concentrations below 8 nmol L⁻¹ contributed to that symptom. So, the various symptoms of LOH might start at various circulating concentrations of androgens. With a given plasma testosterone level, some complaints might be present and others not. This has also been confirmed in other studies establishing symptom-specific thresholds of androgen levels ^{1, 2, 25, 26, 27}. It, therefore, comes as no surprise that there is a significant variation among clinics and European countries in their application of threshold values of testosterone signifying hypogonadism which range from 7.5 to 12.0 nmol L^{−1 28}. Almost certainly a factor in this observation is that physicians in different countries will have different concepts of what constitute the core symptoms of hypogonadism. On the basis of the above observations it is clear that the symptoms of testosterone deficiency are not uniformly and predictably related to values of blood testosterone, which may lead to different diagnostic criteria for testosterone deficiency. So, the conclusion seems inevitable that the clinical manifestations of hypogonadism are multifactorially determined and that the diagnosis should not only depend on the measurement of testosterone but proper assessment should comprise somatic and psychological aspects in addition to measurement of testosterone ²⁹.

Diagnosis of LOH

The diagnosis of LOH and, certainly, the decision to provide androgen treatment must be made with caution, taking the specific increment of symptom prevalence in relation to testosterone levels into account. The above being the case, it is virtually impossible to take a blood testosterone value as a sole indication for testosterone treatment. The presenting symptom of hypogonadism in a man may or may not be related to that testosterone value, though the lower the value of blood testosterone, the greater the likelihood. A decision to provide androgen treatment must not only be guided by the levels of blood testosterone but, based on clinical judgment, more the likelihood that symptoms are related to that particular testosterone level. The decision to treat will have an element of arbitrariness unless testosterone is truly low (< 6 nmol L⁻¹) or truly in the normal range (> 15 nmol L⁻¹). A case can be made for a therapeutic trial of testosterone if the interpretation of clinical and laboratory data provide an ambiguous outcome ^{30, 31}. In case of sexual dysfunction this approach was helpful in treating the disorder ³².

In summary, physicians have to be aware that testosterone plays a significant but not all-decisive role in older male patients and that replacement options should be based on symptoms and on hormone concentrations, which should be evaluated on a symptom-specific basis. Physicians treating elderly men should have expertise in the signs and symptoms of testosterone deficiency which have a broad spectrum of psychological, metabolic and sexual symptoms. Naturally, testosterone is not a panacea of the all the mental and somatic problems men encounter in the process of aging, and indiscriminate use in men who present with vague symptoms will have many failures.

In patients at risk or suspected of hypogonadism, a thorough physical and biochemical work-up is necessary. Transient decreases of serum testosterone levels such as those due to acute illnesses should be excluded by careful clinical evaluations and repeated hormone measurement. Hypogonadism (primary or secondary) can occur at all ages including elderly men. Risk factors for hypogonadism in older men may include chronic illnesses (including diabetes mellitus, chronic obstructive lung disease, inflammatory arthritic disease, renal disease, and HIV-related disease), obesity, metabolic syndrome, and hemochromatosis. A serum sample for total testosterone determination should be obtained between 07:00 and 11:00 hours. The most widely accepted parameters to establish the presence of hypogonadism is the measurement of serum total testosterone. There are no generally accepted lower limits of normal. There is, however, general agreement that the total testosterone level above 12 nmol L⁻¹ (350 ng dL⁻¹) does not require substitution. Similarly, based on the data of younger men, there is consensus that patients with serum total testosterone levels below 8 nmol L⁻¹ (230 ng dL⁻¹) will usually benefit from testosterone treatment. If the serum total testosterone level is between 8 and 12 nmol L⁻¹, repeating the measurement of total testosterone with sex hormone-binding globulin (SHBG) to calculate free testosterone or free testosterone by equilibrium dialysis may be helpful. Measurements of serum luteinizing hormone will assist in differentiating between primary and secondary hypogonadism and serum prolactin is indicated when the serum testosterone is lower than 5.2 nmol L⁻¹ (150 ng dL⁻¹) or when secondary hypogonadism is suspected. Since there are known variations between assay methods, it is imperative that the practitioners utilize reliable laboratories and are acquainted with the reference ranges for testosterone from their local laboratory ³¹. Current immunometric methods for the measurement of testosterone can distinguish between hypogonadism and normal adult men. However, the methods based on mass spectrometry are more accurate and precise ³¹, and are increasingly recognized as the method of choice for serum testosterone measurement. The measurement of free or bioavailable testosterone should be considered when the serum total testosterone concentration is not diagnostic of hypogonadism, particularly in obese men or when there are discrepancies between clinical symptoms and the laboratory value of testosterone. Few laboratories measure free testosterone, and a common practice is to calculate free testosterone using algorithms available on the internet. A recent publication draws attention to the potential inaccuracies of these calculations. Measurement of total testosterone, the formulae used to calculate free testosterone and, to a lesser extent, measurement of sex hormone-binding globulin, are potential sources of inaccuracies ³³. There are no generally accepted lower limits of normal for free testosterone for the diagnosis of hypogonadism. However, a free testosterone level below 225 pmol L⁻¹ (65 pg mL⁻¹) can provide supportive evidence for testosterone treatment ³¹. Threshold values for bioavailable testosterone depend on the method used and are not generally available. Equilibrium dialysis is the gold standard for free testosterone measurement. Free testosterone assays based on analog displacement immunoassays are widely available but do not give an accurate measurement of free testosterone; thus they should not be used ³¹. Alternately, measuring serum SHBG levels together with reliable serum total testosterone levels provides the data necessary for calculating free testosterone levels. Calculated free testosterone correlates well with free testosterone by equilibrium dialysis ³¹. Efforts to create standardization of testosterone assays, agreement on standards for testosterone measurement and accurate reference ranges for testosterone by liquid chromatography–mass spectrometry (LC–MS)/MS are being developed. International reference standards, characterization of methodology, and population-based reference ranges for free testosterone by equilibrium dialysis are needed. Consensus on the equilibrium constants for testosterone binding to SHBG and albumin will allow improved calculation of free testosterone ³¹. Salivary testosterone has also been shown to be a reliable substitute for free testosterone measurements but cannot be recommended for general use at this time, since the methodology has not been standardized and adult male ranges are not available in most hospital or reference laboratories ³¹. Again, only longer-term studies can resolve whether these methods of determining testosterone deficiency in elderly men have made contribution to their well-being.

Benefits of testosterone of replacement therapy

Restoring testosterone levels to within the normal range by using testosterone replacement therapy can improve many of the effects of hypogonadism. In the past decade evidence has been produced of the benefit of androgen treatment on multiple target organs of hypogonadal men, and recent studies show short-term beneficial effects of testosterone in older men that are similar to those in younger men ^{34, 35, 36, 37}.

There may be beneficial effects on mood, energy levels and patients' sense of well-being, sexual function, lean body mass (LBM) and muscle strength, erythropoiesis and BMD, cognition and some benefits on cardiovascular risk factors. The following is a summary of studies examing the effects of testosterone replacement in mainly elderly men.

Improving metabolic syndrome and diabetes type 2, cardiovascular disease

Many of the components of the metabolic syndrome (obesity, hypertension, dyslipidemia impaired glucose regulation, and insulin resistance) are also present in hypogonadal men. Lower testosterone levels are associated with surrogate markers for cardiovascular disease, including less favorable carotid intima medial thickness ³⁸ ankle/brachial index as a measure of peripheral arterial disease ³⁹ and calcific aortic atheroma ⁴⁰. Endogenous testosterone concentrations are inversely related to mortality due to cardiovascular disease of all causes. There is a positive correlation between serum testosterone levels and insulin sensitivity in men across the full spectrum of glucose tolerance ⁴¹ and an improvement of insulin sensitivity was noted after testosterone replacement. The effects of testosterone administration on glycemic control of men with diabetes mellitus were until recently much less certain ⁴² but there are recent reports indicating a favorable effect of testosterone in addition to diet and exercise ⁴³. By increasing LBM and reducing fat mass, testosterone therapy modulates insulin resistance and risk of metabolic syndrome ⁴². The mechanism of the fall in lipids might be related to the decrease in the visceral abdominal fat mass under the influence of androgens, which inhibit lipoprotein lipase activity and increase lipolysis ⁴⁴, with improvement of insulin sensitivity and mobilization of triglycerides from abdominal fat tissue. Low androgen levels are associated with an increased risk of cardiovascular disease in men. But data are lacking as to whether higher testosterone levels predict reduced incidence of combined nonfatal and fatal major cardiovascular events ⁴⁵. The inverse correlation between T levels and the severity of coronary artery disease ⁴⁶ may be related to the fact that low androgen levels are accompanied by an accumulation of abdominal visceral fat ^{47, 48} associated with increased cardiovascular risk factors, impaired glucose tolerance, and non-insulin-dependent diabetes mellitus ⁴⁹.

Testosterone replacement therapy does not increase the incidence of cardiovascular disease or events such as myocardial infarction, stroke, or angina ⁵⁰. A meta-analysis ⁵¹ in 2007 concluded that the current available evidence shows no association between testosterone replacement therapy and cardiac events. However, trials of testosterone therapy generally have not been designed or adequately powered to detect effects on clinically significant cardiovascular events ⁵². The outcome of most studies in men report either a favorable or neutral effect of normal testosterone levels on cardiovascular disease in men. The administration of testosterone in physiological concentration increases coronary blood flow in patients with coronary heart disease ⁵³. Beneficial effects on endothelial function ⁵⁴ and myocardial ischemia have also been demonstrated ^{46, 55, 56} but not on cardiovascular mortality ^{53, 54}. Thus, although lower testosterone levels are associated with higher cardiovascular risk and to an extent with mortality in aging men, randomized controlled clinical trials of adequate size and duration are needed to determine whether testosterone therapy will reduce morbidity and mortality from cardiovascular disease in hypogonadal or eugonadal men.

BMD

Osteopenia, osteoporosis, and fracture prevalence rates are greater in hypogonadal younger and older men ⁵⁷. The prevalence of osteoporosis in testosterone deficient males is double that of those with normal testosterone level ⁵⁸. Testosterone plays a major role in BMD ⁵⁹. Bone density in hypogonadal men of all ages increases under testosterone substitution provided the dose is high enough ^{60, 61}, although normal adult bone mass is not reached ⁶². Testosterone produces this effect by increasing osteoblastic activity and through aromatization to estrogen reducing osteoclastic activity. Part of the androgen effects on bone are at least partially indirect, mediated via their aromatization product estradiol ^{63, 64}. Patients with prostate cancer treated with androgen deprivation therapy have an increased risk of osteoporotic fracture. The role of the LOH in aging males in bone fracture rate remains to be established ⁶⁵. The long-term benefit of testosterone requires further investigation. Trials of the effects of testosterone replacement therapy on BMD yielded mixed results. The pooled results of a meta-analysis suggest a beneficial effect on lumbar spine bone density and equivocal findings on femoral neck BMD. Trials of intramuscular testosterone reported significantly larger effects on lumbar bone density than trials of transdermal testosterone, particularly among patients receiving chronic glucocorticoids. None of studies have been large enough to show a fracture risk reduction with testosterone replacement therapy.

Muscle mass and strength

The aging process is accompanied by significant changes in body composition characterized by decreased fat free mass and increased and redistributed fat mass. These changes can impose functional limitations and increase morbidity ^{66, 67}. Maximal muscle strength correlates with muscle mass independently of age ⁶⁸. In men, declining testosterone levels that occur with aging can be a contributing factor to these changes by direct effect on muscle cells by testosterone or by stimulating IGF-1 expression directly and indirectly leading to increased muscle protein synthesis and growth ⁶⁹. Epidemiologic studies have demonstrated a correlation between bioavailable testosterone concentrations and fat-free mass ⁷⁰; however, the correlation with grip strength is not clear ⁷⁰. Testosterone replacement may be effective in reversing age-dependent body composition changes and associated morbidity ⁷¹. Testosterone administration improves body composition: a decrease of fat mass, increase of LBM ^{37, 72, 73}. In most of these studies, the body weight change did not differ significantly. Testosterone therapy was associated with a greater improvement in grip strength than placebo ^{37, 74, 75}. A recent double-blind placebo-controlled study showed benefits of testosterone administration on skeletal muscle performance in elderly men with chronic heart failure ⁷⁶, and another one found positive effects on the prevention of loss of muscle strength of the lower limbs ⁷⁷. Changes in lower-extremity muscle strength and measures of physical function were reported in only a few studies and were inconsistent. Recent cross-sectional studies showed that in aging men there are also positive correlations between testosterone and muscle strength parameters of upper and lower extremities, as measured by leg extensor strength and isometric hand grip strength ⁷⁸. Moreover, testosterone was positively associated with functional parameters, including the doors test as well as “get up and go” test, and 5-chair sit/stand test ⁷⁹. By contrast, some studies have found an increase in LBM but no change in physical function ⁸⁰ or an increase in strength of knee extension or flexion. Although there is a potential role of testosterone in the management of frailty, it is not known whether testosterone replacement improves physical function and other health-related outcomes, or reduces the risk of disability, falls, or fractures in older men with low testosterone levels ⁸¹.

Improved sexual desire, function and performance

The prevalence of ED increases markedly with age ^{82, 83}. Serum-free testosterone was significantly correlated with erectile and orgasmic function domains of the International Index of Erectile Function (IIEF) questionnaire. Compared with younger men, elderly men require higher levels of circulating testosterone for libido and erectile function ^{84, 85}. However, and/or diminished libido with or without a testosterone deficiency, might be related to other co-morbidities or medications ⁸⁶.

Men with ED and/or diminished libido and documented testosterone deficiency are candidates for testosterone therapy. Adequate testosterone treatment can restore venous leakage in the corpus cavernosum which is a frequent etiological factor in ED in elderly men ⁸⁷. Overviews of randomized controlled clinical trials indicate some benefit of testosterone therapy on sexual health-related outcomes ⁸⁸. Long-term follow-up of testosterone replacement in hypogonadal males and a control group indicates that self-assessment of libido was significantly higher in the testosterone-treated group ⁵⁰. Testosterone replacement has also been shown to enhance libido and the frequency of sexual acts and sleep-related erections ^{75, 89}.

Transdermal testosterone replacement therapy, in particular, has been linked to positive effects on fatigue, mood, and sexual function, as well as significant increases in sexual activity ⁹⁰. In the presence of a clinical picture of testosterone deficiency and borderline serum testosterone levels, a short therapeutic trial may be tried. There is evidence that the combined use of testosterone and PDE5 inhibitors in hypogonadal or borderline eugonadal men have a synergetic effect ^{45, 91, 92, 93}. The combination treatment should be considered in hypogonadal patients with ED failing to respond to either treatment alone. Testosterone produces this effect by enhancing the production of nitric oxide synthase.

Lower urinary tract symptoms

In addition to improvement in sexual function, testosterone therapy may also improve lower urinary tract symptomatology/bladder functions by increasing bladder capacity and compliance and decreasing detrusor pressure at maximal flow in men with LOH ^{11, 94}.

Mood and energy and quality of life (QoL)

Men older than 50 years with low free testosterone levels had poorer QoL. Hypogonadal men commonly complain of loss of libido, dysphoria, fatigue, and irritability ^{95, 96}. These symptoms overlap with signs and symptoms of major depression. There is significant inverse correlation between bioavailable testosterone levels (but not with total testosterone) and a depression score in elderly men, independent of age and weight ⁹⁷. There was a reduced libido and reduced feelings of well being and minimal effect on mood in patients with induced testosterone deficiency; the depressive symptoms during the hypogonadal state were reversed by testosterone replacement ⁹⁸. Testosterone replacement therapy has variable effects on mood, energy and sense of well being. The results of placebo-controlled randomized trials on testosterone's effect on QoL and depressive mood were inconsistent across trials and imprecise ^{99, 100}. Testosterone administered to nondepressed eugonadal men at physiological doses, did not result in significant effects on mood ^{101, 102}. In hypogonadal men, testosterone replacement was associated with improved mood and well-being, and reduced fatigue and irritability ^{103, 104, 105}. Randomized controlled trials of testosterone therapy in men without or with underlying chronic illness using a variety of testosterone formulations, report equivocal improvements in QoL measures, including general well-being and fatigue ^{106, 107}. For patients with major depression and/or dysthymia, improvement was equal to that achieved with standard antidepressants with significant improvement in the depression inventory score. This effect may be a direct effect of testosterone or related to positive effects of testosterone on weight and/or other anthropometric indices. Additional studies are needed to assess the effects of testosterone on clinical depression indices in human immunodeficiency virus-infected patients ^{108, 109}. No relationship between testosterone level and depressive symptoms was found in the Massachusetts Male Aging Study. This discrepancy in the results of the effects of testosterone replacement therapy on mood may be explained by the genetic polymorphism in the androgen receptor which defines a vulnerable group in whom depression is expressed when testosterone levels fall below a particular threshold ^{110, 111}.

Cognitive function

Age-related decreases in bioavailable testosterone predicted age-related decline in visual and verbal memory. There is good evidence for a strong correlation between testosterone levels and cognitive performance, such as spatial abilities or mathematical reasoning. Higher bioavailable and free testosterone concentrations have each been associated with better performance in specific aspects of memory and cognitive function, with optimal processing capacity found in men ranging from 35 to 90 years of age, even after adjustment for potential confounders including age, educational attainment and cardiovascular morbidity ^{112, 113}. whereas total testosterone was not ¹¹⁴. However, contradictory findings have also been reported. In men undergoing hormonal therapy for prostate cancer, suppression of endogenous testosterone synthesis and blockade of the androgen receptor resulted in a beneficial effect on verbal memory but an adverse effect on spatial ability ¹¹⁵ and visuomotor showing a slowed reaction times in several attentional domains. While the outcome of these studies looks promising, there is as yet no definitive proof of the beneficial effects of restoring testosterone levels to normal in elderly men on clinical parameters. Trials of testosterone therapy in men to evaluate its effects on measures of cognitive function and memory to date were all relatively small and of a relatively short duration and have shown mixed results ^{116, 117}. Transdermal testosterone treatment in men aged 34 to 70 years improved verbal memory and spatial memory respectively ¹¹⁸ and intramuscular testosterone improved verbal and spatial memory and constructional abilities in non-hypogonadal men with mild cognitive impairment and Alzheimer's disease ¹¹⁹. In one study of healthy men aged 50 to 90 years, intramuscular testosterone alone or in combination with the aromatase inhibitor anastrozole improved spatial memory, whereas verbal memory only improved in testosterone-treated men in the absence of anastrozole, raising the possibility that part of the effect of exogenous testosterone is mediated by its aromatization to estradiol ¹¹⁹. Although the evidence from observational studies is not uniform, lower free testosterone appears to be associated with poorer outcomes on measures of cognitive function, particularly in older men and testosterone therapy in hypogonadal men may have some benefit for cognitive performance.

Improving anemia

Endogenous androgens are known to stimulate erythropoiesis; increase reticulocyte count, blood hemoglobin levels and bone marrow erythropoietic activity in mammals, whereas castration has opposite effects. Testosterone deficiency results in a 10% to 20% decrease in the blood hemoglobin concentration, which can result in anemia ^{120, 121}. Young hypogonadal men usually have fewer red blood cells and lower hemoglobin levels than age-matched controls, whilst healthy older men also may have lower hemoglobin than normal young men.

Safety of testosterone administration to elderly men

The progressive decline of testosterone in aging men is supported by scientific evidence ^{1, 3}. Whether older hypogonadal men will benefit from testosterone treatment and what will be the risks associated with such intervention can only be resolved by sufficiently powered studies. Data on the risks of testosterone administration are needed, particularly on its safety in elderly men ^{31, 122}. It is unlikely that rigorous scientific data with regard to safety of testosterone administration to elderly men will become available soon. Such studies would include 5 000–7 000 men and would require 5–7 years of observation. So, for the time being, smaller scale studies will have to be utilized to garner information on safety. The main side effects of testosterone administration are listed below.

Polycythemia

There is curvilinear relationship in men (not receiving testosterone administration) between plasma testosterone levels and hemoglobin ²². Testosterone exerts its effect on erythropoiesis through a number of mechanisms. Testosterone has an effect on erythropoietin production in the kidney ¹²³ but it has also a direct effect on colony formation of progenitor cells of erythrocytes ¹²⁴. In a study by Wang ⁹⁵, a dose dependent effect of testosterone could be established on hemoglobin and the hematocrit values. This dose-dependency was also apparent from another study ¹²⁵, which compared the effects of transdermal versus intramuscular testosterone; the latter achieved higher plasma levels of testosterone and raised the hematocrit more than transdermal testosterone. In a recent study it was convincingly confirmed that testosterone has a dose-dependent stimulatory effect on hematopoiesis in men. Remarkably, this effect was more pronounced in older men ¹²¹. Also other studies confirmed the relevance of the dose of testosterone ¹²⁶ and of age as factors in the stimulation of hematopoiesis ¹²⁷. Also obesity and shorter CAG repeats appeared to be factors ²².

A higher value of the hematocrit is associated with stroke ^{128, 129}, and coronary heart disease ¹³⁰. However, a relation between increased hematocrit as a result of androgen supplementation as such and an increased risk for stroke or any cardiovascular event in general has not been demonstrated by a large meta-analysis of placebo-controlled trials of testosterone administration to (elderly) men ¹³¹. Polycythemia ia a manageable risk of androgen administration when hemoglobin levels and the hematocrit are monitored and the dose of testosterone is adjusted.

Lower urinary tract symptoms and prostate disease

Several follow-up studies of men receiving testosterone treatment ^{131, 132, 133} have failed to demonstrate an exacerbation of voiding symptoms due to benign prostatic hyperplasia. Complications such as urinary retention in therapy group did not occur at higher rates than in controls receiving placebo.

The occurrence of prostate cancer after testosterone administration to (elderly) men has been reported ^{134, 135, 136, 137, 138}. By contrast, a variety of studies, using various designs and testosterone formulations, over periods ranging from several months to 15 years, in men with a wide range of ages, have not revealed an increased risk of prostate cancer ^{139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154}. A meta-analysis found that testosterone treatment in older men compared to placebo was not associated with a significantly higher risk of detection of prostate cancer ¹³¹, although the frequency of prostate biopsies was much higher in the testosterone-treated group than in the placebo group ¹³¹.

The above applies also to the assessment of safety of testosterone administration to elderly men. There is no convincing evidence that testosterone is a main factor in the development of prostate cancer in elderly men and guidelines for monitoring have been developed which, if rigorously applied, render testosterone administration to elderly men acceptably safe therapy in men without a prior history of prostate carcinoma or without evidence of harboring a prostate carcinoma. There are now at least three publications demonstrating a lack of prostate carcinoma recurrence with testosterone therapy after definitive prostate carcinoma treatment. Two articles have reported no prostate-specific antigen (PSA) recurrence in a total of 17 men, following radical prostatectomy in men with undetectable PSA ^{155, 156}. A third study reported that no cancer recurrence was noted in 31 hypogonadal men treated with brachytherapy with a follow-up of approximately 5 years ¹⁵⁷. These small studies suggest that normalization of testosterone in men who have shown no signs of recurrence of prostate cancer after treatment, could benefit from testosterone replacement.

The case of an 84-year-old man with hypogonadism, with biopsy-proven prostate cancer (Gleason 6 cancer in both lobes), treated with testosterone gel for 2 years has been reported. In this man a decline of serum PSA was noted over the treatment period ¹⁵⁸.

There is a consensus now that administration of testosterone to elderly men is a responsible practice provided certain guidelines of professional bodies are followed with regard to testosterone administration to elderly men ^{31, 122}. This consensus is based on expert opinion and there remains an urgent need for longer-term safety studies.

Cardiovascular disease

Until a decade ago, it was a widely held belief that androgens have an atherogenic effect and thus led to cardiovascular disease, and androgen administration was regarded as adding to the risk of developing cardiovascular disease. Over the last decade several papers have examined the relationship of androgens with cardiovascular disease and concluded that it is no longer tenable to regard testosterone as a culprit in the etiology of cardiovascular disease ^{159, 160, 161, 162, 163}. Recent epidemiological studies have found that low testosterone levels are a predictor of mortality in elderly men ^{164, 165, 166, 167, 168}. Over the last 2 years a large number of review papers have highlighted the significance of depressed levels of testosterone and cardiovascular disease ^{4, 6, 169}.

Studies of testosterone replacement in men have not reported problems with peripheral edema or exacerbation of hypertension or congestive heart failure, but because current data were largely collected from relatively healthy older men, the possible impact of fluid retention on chronically ill or more frail individuals should be considered. Modest, usually transient, leg edema and fluid retention (up to several kilograms in weight gain) is possible, especially within the first few months of testosterone replacement therapy ⁵.

Sleep apnea

Obstructive sleep apnea ayndrome (OSAS) is characterized by snoring, repetitive episodes of upper airway occlusion resulting in hypoxemia and sleep fragmentation and excessive daytime sleepiness ¹⁷⁰. OSAS is also associated with loss of libido and ED ^{171, 172}. It has been estimated that 10% to 60% of patients with OSAS suffer from ED.

Obesity, and more specifically, the metabolic syndrome is common in patients with OSAS ^{173, 174}. OSAS is associated with an increased risk of hypertension, arrhythmia, myocardial infarction, stroke, and sudden death. Sexual dysfunction associated with OSAS may also be explained by the metabolic syndrome ^{175, 176}. Men with OSAS have lowered plasma testosterone levels ¹⁷⁷ but this association was largely explained by adiposity in agreement with the finding that the metabolic syndrome is associated with reduced plasma testosterone values ^{178, 175}. In a large meta-analysis of placebo-controlled trials of testosterone administration to (elderly) men ¹³¹, the frequency of sleep apnea was not significantly different between the two groups. Nevertheless, it is safe to consider obstructive pulmonary disease in overweight persons or heavy smokers as a relative contra-indication.

Conclusion

There is a host of studies in elderly men with lower-than-normal testosterone reporting improvement of features of the metabolic syndrome, of BMD, of mood and of sexual functioning. While the outcome of these studies looks promising, there is as yet no definitive proof of the beneficial effects of restoring testosterone levels to normal in elderly men on clinical parameters. Few of these studies meet as yet rigorous standards of scientific enquiry: double-blind, placebo-controlled design of the study. The endpoints of the studies are usually surrogate markers of disease and not clinical endpoints such as improvement and prevention of cardiovascular disease, diabetes mellitus type 2, osteopenia/osteoporosis, less fractures and less frailty. Admittedly, studies to prove the latter are difficult to undertake in view of the costs and the long duration of such studies. It is difficult to compare these studies because of the intrinsic difficulties in establishing the degree of hypogonadism in the various studies: laboratory outcomes may vary between laboratories and a certain laboratory value of testosterone does not mean the same for various men on the basis of properties of their androgen receptors.

The above applies also to the assessment of safety of testosterone administration to elderly men. There is convincing evidence that testosterone is not a main factor in the development of prostate cancer in elderly men and guidelines for monitoring have been developed which, if rigorously applied, render testosterone administration to elderly men acceptably safe.