Androgens and Bone: Clinical Aspects – Estrogens and Hypogonadism
Androgens have myriad actions on the skeleton throughout life. During adolescence those effects clearly promote skeletal growth and the accumulation of mineral mass, and for many years there has been hope that these anabolic effects will be useful in the prevention and therapy of metabolic bone disease in later life. In fact, the essential nature of the effects of androgens in bone remains uncertain, and the clinical usefulness of androgens is clearly defined in only a few situations. Nevertheless, there is an increasing interest in how the actions of androgens are integrated into the broad scheme of bone metabolism and how those effects can be adapted for prevention and therapy.
Adolescence is associated with profound increases in bone mass in both sexes. Both axial and appendicular bone mass increase (Gilsanz etal., 1988; Lu et al., 1994), with the addition of almost half of the total adult skeleton during this brief time. In boys, the rapid increase in indices of bone formation and skeletal mass during this period is closely linked to pubertal stage (Krabbe et al., 1979; Bonjour et al., 1991) and to testosterone levels (Krabbe et al., 1979,1984; Riis etal., 1985).These data are consistent with the precept that testicular androgen secretion plays a role in the genesis of the adolescent increase in bone mass. In addition, however, the increase in adrenal androgens that occurs in the prepubertal period (adrenarche) (Odell, 1989) may also affect bone mass. Longitudinal bone growth (via epiphyseal action) has been reported to accelerate during adrenarche (Parker, 1991). Bone mass accretion certainly occurs before sexual development begins (Slemenda et al., 1994) and could be influenced by the actions of adrenal androgens.
The overall result is a male skeleton that is larger in most dimensions, thus conferring a considerable biomechanical advantage. Total body mineral content is 25-30% greater in men (Rico etal., 1992). Both the diameter and cortical thickness (and hence the total mass and mineral content) of long bones is greater in men (Kelly et al., 1990b; Fehily et al., 1992). Vertebral size is also larger in men, even when other elements of body size (height, weight) are controlled (Gilsanz et al., 1994). These gender differences in bone size are not matched by differences in the essential composition of bone because the true volumetric density of the bone in men and women is essentially the same (Bonjour et al., 1994).
That these pubertal changes in the male skeleton are related to androgen action is suggested by several kinds of evidence. Genetic males with complete androgen resistance appear to have low bone density (Soule et al., 1996) and a skeletal mass similar to that of women (Munoz-Torres et al., 1995). Moreover, the presence of androgen insufficiency, for instance in patients with isolated gonadotrophin deficiency, results in abnormally low bone mass even when corrected for bone age (Finkelstein et al., 1987). Treatment of these patients with testosterone before epiphyseal closure results in a rapid increase in bone mass (Finkelstein etal., 1989; Arisaka etal., 1991; Devogelaer et al., 1992). Finally, the short-term administration of testosterone to prepubertal boys quickly causes an increase in calcium retention and incorporation into bone (Mauras et al., 1994). Nonaromatizable androgens have been reported to increase linear growth (Stanhope et al., 1988; Keenan et al., 1993), but their effects on bone density aren’t yet clear.
Not only do androgens seem to be involved in normal bone mass development, but the actual timing of the onset of puberty is fundamentally important as well. It has been known that constitutionally delayed puberty can result in somewhat greater adult height (Uriarte et al., 1992), but bone mass is reduced in these patients even at the conclusion of sexual maturation (Finkelstein et al., 1992, 1996). In delayed puberty, bone density is also reduced even when adjusted for bone age (Bertelloni et al., 1995). Constitutional delay of puberty is a common condition, and this reduction in peak adult bone mass may have important implications for eventual osteoporosis and fracture risk. More evidence for the importance of the timing of puberty comes from experience with the therapy of hypogonadal men who have not yet undergone puberty. In these patients, there is a brisk increase in bone mass in response to testosterone therapy, but the final bone mass developed is impaired (Finkelstein et al., 1989). A similar response is seen in boys with constitutional pubertal delay treated with testosterone (Bertelloni etal., 1995).
TABLE I. Vertebral Di mensions in 30 Boys and 30 Girls Matched for Age, Height, and Weight
Cross-sectional area (cm2)
7.72 ± 1.24
6.69 ± 0.99
1.83 ± 0.14
1.86 ± 0.18
8.33 ± 1.46
7.49 ± 0.99
1.89 ± 0.14
1.93 ± 0.17
9.12 ± 1.46
8.12 ± 0.98
1.94 ± 0.13
1.97 ± 0.16
8.39 ± 1.35
7.50 ± 0.98
1.89 ± 0.13
1.92 ± 0.17
All these findings support a crucial role of pubertal androgens in the development of adult bone mass in the male, probably as a result of both direct and indirect actions at the skeletal level. Because androgens are important in the development of peak bone mass, there is great potential for the use of androgens as therapeutic agents during this period of life. As is discussed elsewhere in this volume, the mechanisms by which androgens affect bone, particularly during growth, are probably integrally related to the effects of estrogens and the activity of aromatases.
III. Estrogens versus Androgens in Puberty
The dynamic interplay between androgens and estrogens at the skeletal level during puberty is not well understood but is probably quite important. That estrogen action is essential for the development of peak bone mass in men has been highlighted by reports of interesting young men with estrogen deficiency. Smith et al. (1994) described a young adult patient with an abnormality in the structure of the estrogen a-receptor (thus rendering him functionally estrogen-deficient, at least in regard to the a-receptor) who had a delayed bone age, tall stature, and profound osteopenia. Men with estrogen deficiency resulting from aromatase deficiency have also been recently noted to have a very similar phenotype (Morishima et al., 1995; Bilezikian et al., 1998). The common abnormalities in bone age and stature in these estrogen-deficient patients are to be expected because estrogens are essential for normal closure of growth plates in both sexes.
The aberrations noted in bone mass are more difficult to understand, particularly because they exist in the face of testosterone levels that are actually higher (even strikingly higher) than in normal men. In one man with aromatase deficiency, estrogen treatment was noted to close the open epiphyses and result in an increase in bone mass (Bilezikian et al., 1998). Whether the increase in bone mass resulted from the accumulation of mineral at the growth plates or was the result of a general skeletal effect of estrogen is unknown, but it is of fundamental importance in the understanding of the respective roles of androgen and estrogen. If in fact the estrogen therapy resulted in an increase in bone mass unrelated to epiphyseal closure, estrogen would assume a more pivotal role in skeletal maturation in men and would shift therapeutic attention (for instance, in hypogonadal adolescents) to ensuring the adequacy of estrogen as well as androgen action.
IV. Age-Related Declines in Androgen Levels in Adult Men: Contribution to Bone Loss
Aging is associated with a clear decline in testicular and adrenal androgen levels in men (Gray et al., 1991; Vermeulen, 1991), and there is considerable interest in the possibility that androgen replacement may attenuate that loss and reduce fracture rates in the elderly. Although there have been attempts to evaluate this hypothesis, most efforts have been indirect and less than conclusive. For instance, a variety of cross-sectional studies have examined the relationship between androgen levels and bone mass in older men. Some have suggested a significant association between androgen concentrations and bone mass (McElduff etal., 1988; Kelly etal., 1990b; Murphy et al., 1993; Rudman et al., 1994; Ongphiphadhanakul et al., 1995), but others have not been able to substantiate an interaction between bone mass and either testosterone or adrenal androgens (Meier et al., 1987; Barrett- Connor etal., 1993;Drinka etal., 1993; Wishart et al., 1995). There continues to be much speculation about the role of the age-related decline in androgen levels in the development of bone loss in older men, and controversy surrounds the issue of whether androgen replacement is useful in the prevention of osteoporosis in older men.
V. Estrogens in Adult Men
The possibility that there is an important role for estrogens in the maintenance of adult bone mass in men must be considered. It is feasible that estrogens function in men to maintain bone mass, presumably in concert with the direct effects of androgens. Recent data suggest that estrogen levels are significantly associated with bone mass in elderly men (Slemenda et al., 1997) and that reduced estrogen levels may be related to the development of low bone mass in older men (Bernecker et al., 1995). A fascinating series of male-to-female transsexuals indicates that in the presence of antiandrogen treatment or castration, estrogen treatment is capable of maintaining or increasing bone mass (Lips et al., 1989; Van Kesteren et al., 1996). In fact, recent publications have suggested that serum concentrations of estrogens may be more closely related to bone mass in older men than are androgens (Greendale et al., 1997; Slemenda et al., 1997). The correlations between estrogen levels and bone density were significant but were quite weak, and estrogen explained a minor proportion of the overall variation in bone density. Despite the previous assumption that androgens provided the dominant gonadal steroid effect on bone in men, this new information demands that the role of estrogens, and interactions between sex steroids, be further examined. These controversies remain unresolved and clearly affect the question of the usefulness, and possible mechanisms of action, of androgen replacement therapy in older men with low bone mass.
VI. Hypogonadism in Adult Men
After peak bone mass is achieved during adolescence, androgens continue to be vitally important in the maintenance of bone health. Abnormal gonadal function is well known to incur risk for bone loss, and a wide variety of causes of gonadal failure result in osteopenia (Orwoll and Klein, 1995). Hypogonadism is present in a substantial portion (15%) of men evaluated for severe osteoporosis (Kelepouris et al., 1995), and hip fracture occurs more commonly in the presence of hypogonadism (Stanley et al., 1991). Both cortical and trabecular osteopenia are present, but cancellous bone loss seems most intense. Resulting changes in cancellous architecture are not well characterized, but in most reports there is evidence of reduced trabecular number (Francis and Peacock, 1986; Jackson et al., 1987). In some series, the degree of bone loss correlates with the level of serum testosterone (Foresta et al., 1983; Horowitz et al., 1992), but that is not a consistent finding. A threshold level of serum testosterone below which bone mass begins to decline, even in the absence of other skeletal stressors, has not been established. This issue is of considerable importance given the very wide range of testosterone levels in adult men and the lack of consensus concerning the definition of hypogonadism. Finally, because hypogonadism in men is usually characterized by at least some degree of estrogen as well as androgen deficiency, a component of hypogonadal bone disease may be related to a lack of estrogen action. Although this issue is not well understood, it remains likely that androgen deficiency is a major determinant of the observed changes in bone metabolism (vide supra).
Androgen insufficiency may be an important component of several other forms of metabolic bone disease as well. For instance, in subjects treated with glucocorticoids, testosterone levels can be substantially reduced (Reid et al., 1985) and may contribute to bone loss. Similar interactions may result in remodeling alterations in patients with renal insufficiency, alcoholism, chemotherapy, and the like (Orwoll and Klein, 1995).
The mechanisms by which androgen withdrawal leads to bone loss in men have only begun to be explored. In the most direct assessment of the events following the onset of hypogonadism, Stepan et al. (1989) examined changes in bone mass and biochemical indices of remodeling in a group of men undergoing castration. In the 1 to 3 years these men were followed after orchidectomy, vertebral bone loss was rapid (~7%/year) and progressed in conjunction with clear evidence of an increase in bone turnover. Essentially the same changes occur after the institution of GnRH agonist therapy in adult men (Goldray et al., 1993). Thus in the early stages of hypogonadism there appears to be an increase in remodeling and bone resorption, just as is seen in animal models of gonadal insufficiency. This concept is supported by the data indicating androgen action is important in the suppression of cytokines active in the stimulation of osteoclastogenesis (Bellido et al., 1995). Bone loss seems to be most intense in the several years after castration, with a subsequent slower phase of loss—a process remarkably similar to that seen in women after menopause. This pattern also is observed in castrate animals (Gunness and Orwoll, 1995). More information is needed concerning the histomorphometric character of acute androgen deficiency and the effects on specific skeletal compartments. For instance, it is unclear to what extent an androgen-dependent depression of bone formation accompanies the increase in bone resorption, especially in the slower phase of bone loss that probably follows the rapid loss that occurs immediately after the onset of hypogonadism. Whether abnormalities in cortical bone metabolism are as profound as those reported to occur in cancellous bone is also uncertain.
Commonly, the diagnosis of hypogonadism is made in the subacute or chronic phases of the disorder, and most evaluations of the nature of hypo-gonadal bone disease involve patients with long-standing reproductive ab-normalities. In these patients the mechanisms responsible for osteopenia are particularly unclear, and the histological pattern of the bone disorder is not well described. There are some small retrospective series available, and most other reports are uncontrolled and involve men in whom the hypogonadism was of varied causation and duration. For example, in a study of 13 men with long-standing hypogonadism, Francis and Peacock (1986) found that bone remodeling and formation rates were reduced, and 1,25-(OH),vitamin D levels and intestinal calcium absorption were low in those with fracture. With testosterone therapy, l,25-(OH)zvitamin D levels increased, and there was a suggestion that indices of bone formation increased. Similarly, Delmas and Meunier (1981) found decreased rates of formation in a small group of hypogonadal men, and formation was low in a single case report by Baran (Baran et al., 1978), but vitamin D levels were not described. These contributions certainly raise the question of whether the osteopenia in patients with long-term androgen deficiency is, to a major extent, the result of a defect in bone formation.
In contrast, Jackson et al. (1987) described the histomorphometric character of chronically hypogonadal men without vitamin D deficiency and found no apparent defect in formation but rather an increase in bone resorption. They suggested that earlier findings of a defect in bone formation were more the result of insufficient vitamin D action than of gonadal insufficiency. Actually, in all the available reports, there is considerable subject heterogeneity, and the nature of remodeling is quite variable. In the face of this considerable patient diversity, inadequate controls, and the presence of other confounding medical conditions, no firm conclusions can be drawn concerning the remodeling defect induced by long-standing hypogonadism in men. If it is similar to that in postmenopausal women, it would be expected that the initial period of increased remodeling would be followed by one of relatively lower remodeling rates and slower rates of change in bone mass. Since the descriptions of osteoporosis by Fuller Albright half a century ago, a reduction in bone formation has been postulated to be a primary cause of hypogonadal bone disease. In fact, this precept remains essentially unproved.
VII. Androgen Therapy: Potentially Useful Androgen Effects
Despite the relative paucity of research concerning androgen action in bone, there are several well-known effects, both direct and indirect, that may prove beneficial. An understanding of these actions was derived primarily from observations in animals or during human adolescence, but they have the potential to be translated into therapeutic terms.
A. Growth Promoting Effects
During adolescence, androgens appear to exert important effects on the skeleton. By the end of puberty, men have greater bone mass than women, a difference that is most marked in the cortical compartment (greater cortical diameter and thickness) (Martin, 1993; Bonjour etal., 1994; van der Meulen et al., 1996). Studies of the effects of testosterone on skeletal calcium accumulation during childhood strongly suggest that this is a direct effect of gonadal steroids (Mauras et al., 1994), and in animals orchidectomy reduces periosteal bone formation, an effect which is reversed with androgen therapy (Turner et al., 1990). The observation that, in genetic males with complete androgen insensitivity (androgen receptor dysfunction), skeletal size is similar to that in normal females strongly suggests that androgens play a major role (Vanderschueren, 1996). Nonaromatizable androgens promote skeletal growth (Stanhope et al., 1988; Keenan et al., 1993).
However, there may also be indirect effects that account for a larger skeleton in men. For instance, before and through puberty, boys have greater muscle and total body mass than do girls, and androgen therapy in adolescents with delayed puberty increase fat-free mass (Arslanian and Suprasong- sin, 1997). The resultant increase in mechanical force exerted on the skeleton has been postulated to play a major role in the determination of bone mass (van der Meulen et al., 1996; Gilsanz et al., 1997). Androgens may also affect the growth hormone/IGF-1 axis (Parker et al., 1984; Keenan et al., 1993; Tai-Pang et al., 1995; Benbassat et al., 1997), that in turn may influence skeletal development. Finally, estrogen (derived from the aromatization of androgen) is very important for skeletal maturation (Smith et al., 1994) and may contribute to male skeletal growth (Vanderschueren, 1996).
Because bone size and cortical thickness have profound effects on bio-mechanical strength and fracture resistance, these positive effects of androgens are of great potential use. The therapeutic impact of those actions may be more obvious during skeletal growth, for instance in the therapy of pubertal forms of hypogonadism in boys. However, if the potential for androgen action on cortical thickness or bone size continues into adulthood, those effects could be useful in the prevention and therapy of common age-related disorders as well (especially osteoporosis). That androgens may continue to exert those actions is suggested by the observation that long bone dimensions continue to increase during adulthood in men, presumably due to periosteal bone accretion, and that this increase is more marked in men than in women (Ruff and Hayes, 1988).
B. Suppression of Bone Resorption
In a manner that seems very similar to that of estrogen, androgens seem to exert a moderating effect on cancellous osteoclastic bone resorption. Increases in osteoclastic activity follow quickly after castration in males (Turner et al., 1989, 1990; Gunness and Orwoll, 1995) and appear to be prevented by nonaromatizable androgens (Turner et al., 1990). In vitro, androgens prevent the increases in cytokine generation that in large part mediate osteoclastogenesis and resorption after gonadectomy in males (Bellido et al.,
1995) . Although there is considerable uncertainty about how the direct effects of androgens in bone cells are intertwined with the effects of estrogens derived from the aromatization of androgens, the effectiveness of nonaromatizable androgens in mediating these cellular effects points strongly to a primary androgen action.
C. Bone Formation
Androgens are the prototypical anabolic agents, and there has been considerable speculation that the effects of androgens in the skeleton may be in part the results of a stimulation of bone formation (Orwoll, 1996). There are androgen receptors in osteoblasts (Colvard et al., 1989; Orwoll etal., 1991) and evidence that androgens affect osteoblast activity (Fukayama and Ta- shjian, 1989; Kasperk et al., 1990; Bellido et al., 1995; Weinstein et al., 1997). During pubertal development, and in the periosteal space, there is considerable support for the contention that androgens enhance bone formation (vide supra). In addition to possible effects on bone growth, some have speculated that androgens may promote increased osteoblastic new bone formation in cancellous areas. However, peak cancellous bone density is similar in males and females, and in most studies the benefits of androgen administration on trabecular bone mass has been modest. Certainly, there is yet little substantial evidence of a major anabolic effect. On the other hand, some reports suggest the possibility of impressive gains in bone density during androgen therapy of hypogonadal men. In view of the clear effects of androgens to reduce the increased rate of bone remodeling (and the rates of both resorption and formation) induced by castration, it is difficult to design experiments to examine directly the influence of androgens on bone formation in isolation. Thus the issue remains unresolved. If, in fact, androgens have positive effects on trabecular bone accumulation, there would be an obvious therapeutic potential.
D. Androgens and IGF-I
Androgens may exert many of their complex effects on bone metabolism via actions on cytokines and growth factors in the skeletal microenvironment. In addition, systemic levels of some of these substances are modulated by androgens (perhaps via conversion to estrogens) (Hobbs et al., 1993; Weissberger and Ho, 1993), and may also affect bone health. IGF-1 has potent actions on bone, and serum levels are increased by androgens (Mauras et al., 1987). If circulating IGF-1 levels have positive effects on bone, the increase stimulated by androgens may be beneficial.
E. Androgens and Muscle Strength
Muscle strength has long been considered responsive to androgens, and recent objective data support that contention. Because strength has been associated with increased bone mass and a reduction in fall propensity, this affect of androgens could be useful in reducing fracture risk. Certainly there remain unresolved issues (the effects of androgens on strength in physiologic concentrations, the usefulness of androgens in improving strength in the elderly, etc.), but the potential benefits of androgens acting on the skeleton via an effect on muscle may be substantial.
VIII. Androgen Replacement in Adolescent Hypogonadism
Because adolescence is such a critically important part of the process of attaining optimal peak bone mass, it is also especially vulnerable to disruption by alterations in gonadal function. Even constitutional pubertal delay is associated with a reduction in peak bone mass development, despite eventual full gonadal development (Finkelstein etal., 1992,1996). The impairment in bone mass in adolescence with organic hypogonadism (hypogonadotropic hypogonadism) is similar to patients with this form of hypogonadism studied later in life, suggesting that the detrimental effect suffered in adolescence is the major cause of osteopenia (Finkelstein et al., 1987). In view of the major effects of androgens on the skeleton during growth (whether direct or indirect, as discussed earlier), the response to therapy of gonadal dysfunction during this time would be expected to be brisk.
Although studies are few, this would appear to be the case (Arisaka et al., 1995). Finkelstein et al. (1989) reported that treatment of hypogonadal men with testosterone elicited the most robust skeletal response in those who were skeletally immature (open epiphyses). In young men considered to have constitutional delay of puberty, testosterone therapy results in a clear increase in bone mass, but whether this provides a solution to the problem of low peak bone mass in these patients is not yet known (Bertelloni et al., 1995). All this information suggests that the diagnosis of frank hypogonadism during childhood or adolescence carries with it the risk of impaired skeletal development and that there is an opportunity to improve bone mass with androgen therapy. In fact, from a skeletal perspective, it appears that therapy should be initiated before epiphyseal closure to maximize bone mass accumulation. Issues that are unresolved include whether bone mass can be normalized with therapy, the most appropriate doses and timing of therapy, and the source of the beneficial effects (androgen versus estrogen, growth factor stimulation, etc.).
IX. Androgen Replacement in Hypogonadal Adult Men
Androgen therapy in hypogonadal men has been shown to affect bone mass positively, at least in most patient groups (Finkelstein et al., 1989; Diamond etal., 1991; Devogelaer et al., 1992; Orwoll and Klein, 1995). For instance, Katznelson et al. recently reported an increase in spinal BMD of 5-6% in a group of adult men with hypogonadism treated with testosterone for 18 months (Katznelson et al., 1996), although there was an insignificant increase in radial BMD. As in the experience reported by Katznelson et al., the increase in density following testosterone replacement generally appears to be most apparent in cancellous bone (e.g., lumbar spine), although the literature is not particularly consistent in this regard. Most reports indicate that the increase in bone mass with testosterone therapy can be expected to be modest in the short term (up to 24 months), but Behre et al. (1997) noted an increase in spinal trabecular BMD of >20% in the first year of testosterone therapy in a group of hypogonadal men, with further increases thereafter. The most marked increases were observed in those with the lowest testosterone levels before therapy. In men treated for at least 3 years, bone density was found to be at levels normally expected for their ages. Although the experience remains small, there is a suggestion that, in older men with hypogonadism, the response to therapy can be expected to be similar to that in younger adult patients (Morley et al., 1993; Behre etal., 1997).
The cellular mechanisms responsible for improvements in bone mass are unclear. As discussed earlier, in the early phases of androgen deficiency (e.g., after castration) there appears to be a phase of increased remodeling and resorption, so that therapy may be beneficial because of an inhibitory effect on osteoclastic activity. However, in most available clinical studies, the patient populations treated with androgens have had well-established hypogonadism and were characterized by an array of remodeling states. In these subjects the cellular effects of androgen replacement are not well known. In some reports, testosterone therapy appeared to result in an increase in cancellous bone formation (Baran et al., 1978; Francis and Peacock, 1986), but in other series there appeared to be no clear remodeling trend induced by therapy (Finkelstein et al., 1989). Most recently, several groups have reported that biochemical indices of remodeling decline in response to testosterone replacement (Katznelson et al., 1996; Wang et al., 1996b), which is what might be predicted if sex steroid deficiency results in an increase in remodeling, with bone loss on that basis. Interestingly, some reports also suggest that osteocalcin and/or alkaline phosphatase levels may increase with androgen therapy (Morley et al., 1993; Wang et al., 1996b; Guo et al., 1997), perhaps signaling an increase in bone formation.
In addition to the generally positive effects of androgen replacement therapy in hypogonadal men, additional benefits may be gained from the increases that have been noted in strength and lean body mass in these patients (Morley et al., 1993; Katznelson et al., 1996; Wang et al., 1996b; Sih et al., 1997). Because lean body mass and strength have been correlated with bone mass and a reduced propensity to fall, they may further serve to promote bone health and reduce fracture risk.
Despite the generally positive tenor of most studies of the skeletal effects of testosterone replacement, in some patient groups (for instance those with Kleinfelter’s syndrome), the advantage associated with androgen therapy is questionable. The available studies report very mixed results (Kubler et al., 1992; Wong et al., 1993). This may be because the level of androgen deficiency in Kleinfelter’s syndrome (as in the case of some other causes of hypogonadism) is quite variable. These findings suggest the need to consider carefully the potential benefits of androgen replacement in each patient individually.
A. Doses/Routes of Administration
The most efficacious doses and routes of androgen administration for the prevention/therapy of bone loss in hypogonadal men remain uncertain. In part, this is because the specific testosterone levels necessary for an optimal effect have not been defined. Current practice is to attempt to ensure testosterone concentrations similar to those of normal young men. Whether the pulsatile pattern of testosterone exposure characteristic of intramuscular administration is more or less conducive to skeletal health than the more stable pattern produced by transdermal administration is unknown. In some studies, transdermal testosterone therapy appeared to be as effective as intramuscular administration in promoting bone mass (Behre et al., 1997). Novel routes of administration are being examined (e.g., buccal preparations) and appear to result in positive skeletal effects (Wang et al., 1996b).
B. Follow-up of Treated Patients
The follow-up of hypogonadal men treated with testosterone, although not well codified, should certainly include careful monitoring for adverse effects. The risk of prostate disease in androgen-treated men is unknown, but regular prostate evaluations are necessary to ensure that any development of benign or malignant disease is detected early in its course. The development of errythrocytosis is not uncommon, particularly with intramuscular testosterone administration, and complete blood counts at 6 to 12 month intervals are useful to detect its appearance. Other problems that have been postulated to be of concern in androgen-treated men are hyperlipidemia and sleep apnea (Swerdloff and Wang, 1993).
In terms of skeletal effects, therapeutic success may be assessed via follow- up bone mass measures. In view of recent reports, increases in bone density can be anticipated at several skeletal sites in the average patient (Katznelson et al., 1996; Guo et al., 1997). Although the role of biochemical markers of remodeling is controversial, the available data suggest that an adequate androgen effect should be accompanied by a fall in indices of bone resorption, an effect that should be especially useful if resorption markers are increased at base line (Katznelson et al., 1996; Wang et al., 1996b; Guo et al., 1997). Markers of bone formation may be more difficult to use at present in routine clinical situations; some reports suggest that increases follow therapy, whereas others support a decline (Katznelson et al., 1996; Wang et al., 1996a; Guo et al., 1997). The response may depend on the specific marker. Clearly, clinicians deciding on a follow-up strategy must be aware of the uncertainty currently inherent in the field and the vagaries of using the tools available (i.e., issues of measurement precision).
C. Unresolved Issues
There remain many additional unresolved issues concerning the role of androgen treatment in the prevention/therapy of osteoporosis in hypogonadal men, including:
• The degree of hypogonadism (level of testosterone) at which adverse skeletal effects begin to occur is undefined, and hence it is difficult to decide upon the usefulness of therapy in many men with borderline levels of serum testosterone.
• Because hypogonadism in men results in deficiencies of estrogen as well as testosterone, and because testosterone therapy results in increases in serum estrogen (as well as androgen) levels, the relative roles of estrogen versus testosterone in affecting skeletal health in hypogonadal men are unclear. It is unknown whether it is useful to assess estrogen concentrations in the diagnosis of hypogonadal bone disease in men, or whether using estrogen levels to monitor the success of testosterone therapy is beneficial.
• In general, the available treatment studies are of relatively short duration, and it is unclear how long any increases in bone mass can be sustained and what eventual treatment effect can be expected.
• The increase in bone mass that appears to accompany testosterone therapy is of uncertain usefulness in preventing fractures.
• Whether pretreatment age, duration of hypogonadism, degree of osteopenia, remodeling character, and associated medical conditions affect the therapeutic response is relatively unknown.
• Potential adverse effects of androgen therapy (e.g., prostate, lipid) are not well delineated.
X. Androgen Replacement in Aging Men
Old age is associated with a panoply of physical changes in men, many of which have been speculated to be related, either directly or indirectly, to the decline in androgens that accompanies aging (Lamberts et al., 1997). A few small trials of androgen administration in older men have suggested that there may be beneficial effects (increased strength and improved body composition) (Tenover, 1992; Morley et al., 1993; Sih et al., 1997), and some reports, as of yet inconclusive, indicate that bone mass or biochemical indices of remodeling may improve (Tenover, 1992; Morley et al., 1993). Whether androgen replacement therapy can prevent or reverse bone loss in aging men is of enormous importance, but it remains uncertain. Until there is more definitive data available concerning both advantages and disadvantages, testosterone replacement should not be used in elderly patients unless there is convincing evidence for androgen deficiency. This decision is difficult in many older men who have symptoms that can be associated with androgen deficiency but that are also common in the aged regardless of gonadal status (weakness, loss of libido or sexual ability, etc.). The identification of hypogonadism in this group is made especially challenging by the expected decline in androgen levels with age and the dirth of data concerning the levels (threshold concentrations) that are associated with adverse effects on bone.
XI. Androgen Therapy in Eugonadal Men
It has been hypothesized that androgens may have positive effects on bone formation and resorption (vida supra). The threshold level of androgens necessary to provide maximal skeletal benefits is unknown, and some have speculated that testosterone supplementation would benefit osteoporotic men even in the face of normal testosterone levels. The experience with this approach has been very limited, but Anderson et al. (1997) recently found in an uncontrolled trial that testosterone supplementation was associated with an increase in bone density and a reduction in biochemical markers of remodeling in a group of osteoporotic, eugonadal men. This approach remains very much of uncertain benefit, and until its advantages are documented in controlled trials, it cannot be recommended. This is particularly true in view of the lack of knowledge concerning the potential adverse effects that may be associated with testosterone supplementation.
XII. Androgen Therapy in Secondary Forms of Metabolic Bone Disease in Men
A variety of systemic illnesses and medications are associated with lowered testosterone levels (Gray et al., 1991), and it has been postulated that relative hypogonadism may contribute to the bone loss that also accompanies many of these conditions. For instance, renal insufficiency, glucocorticoid excess, post-transplantation, malnutrition, and alcoholism are all associated with osteopenia and with low testosterone concentrations. Although there is little experience with testosterone supplementation in these patients, it may offer advantages to skeletal health as well as to other tissues (muscle, red cells, etc.). In a randomized study of crossover design, Reid et al. (1996) reported that testosterone therapy apparently improved bone density (and body composition) in a small group of men receiving glucocorticoids. Similarly, testosterone therapy apparently improved forearm bone mass in a small group of men with hemochromatosis (treated simultaneously with venesection) (Diamond et al., 1991). The number of patients affected by conditions associated with low testosterone levels is potentially quite large, and more information is needed to understand the role of androgen replacement in the prevention/therapy of concomitant bone loss.
XIII. Therapy with Other Androgens
Several androgenic compounds may have effects on bone mass, and of most current interest are adrenal androgens. DHEA has become widely available as a supplement and has been the source of considerable recent debate. DHEA levels fall dramatically with aging in both sexes (Lamberts et al., 1997), and because DHEA can act as a precursor for both androgens and estrogens, there is some reason to expect that its administration may have effects on bone metabolism and bone mass. A variety of attempts have been made to link those changes with alterations in bone mass associated with age (Barrett-Connor et al., 1993), but to date the information available is both inconclusive and incomplete. Labrie et al. (1997) reported that cutaneous DHEA therapy in post-menopausal women was associated with an increase in spinal bone density, with reductions in some markers of bone remodeling (but increases in others). Similar studies are available in animals. There remains no long-term, well-controlled trials of DHEA supplementation in any coherent subject group, and until that information is available, no confident recommendations can be made. Of concern, there is also no available data concerning the potential risks associated with long-term therapy.
XIV. Research Directions
The skeleton is an androgen-responsive tissue, and theoretically there are a number of mechanisms by which androgen action may be important for skeletal health in both men and women. Clinical studies show that androgen therapy provides benefit for the prevention and therapy of bone loss in hypogonadal men, and added gain may result from increases in muscle strength, particularly in older men at risk for falls. Nevertheless, the potential usefulness of androgen therapy, even in men, remains unclear. Major issues to be clarified include the appropriate criteria for patient selection, the specifics of dosing and drug delivery, the nature of short- and long-term adverse effects, and the impact of therapy on fracture rates. Some of these issues can be clarified only with large-scale intervention trials. Androgens may have usefulness in women as well, but adverse effects loom as a more difficult problem. In both men and women, a greater understanding of the molecular mechanisms of androgen action in the skeleton may provide means to harness the beneficial effects of androgens without the disadvantages, for instance by the development of compounds with tissue-specific actions.