Gonadal steroid–dependent effects on bone turnover and bone mineral density in menSevere gonadal steroid deficiency induces bone loss in turnover steroid men; however, the specific roles of androgen and estrogen deficiency in hypogonadal bone loss turnoverr unclear. Additionally, the threshold levels of testosterone and estradiol that initiate bone loss are uncertain. One hundred ninety-eight healthy men, ages 20—50, received goserelin acetate, which suppresses endogenous gonadal steroid production, and were randomized to treatment with 0, 1. An additional cohort of men was randomized to receive these treatments plus anastrozole, turnover steroid suppresses conversion of androgens to estrogens. Thirty-seven men served as controls and turnover steroid placebos for goserelin and testosterone.
Gonadal steroid–dependent effects on bone turnover and bone mineral density in men
Severe gonadal steroid deficiency induces bone loss in adult men; however, the specific roles of androgen and estrogen deficiency in hypogonadal bone loss are unclear. Additionally, the threshold levels of testosterone and estradiol that initiate bone loss are uncertain. One hundred ninety-eight healthy men, ages 20—50, received goserelin acetate, which suppresses endogenous gonadal steroid production, and were randomized to treatment with 0, 1.
An additional cohort of men was randomized to receive these treatments plus anastrozole, which suppresses conversion of androgens to estrogens. Thirty-seven men served as controls and received placebos for goserelin and testosterone. Bone microarchitecture was assessed in men. As testosterone dosage decreased, the percent change in C-telopeptide increased. These increases were considerably greater when aromatization of testosterone to estradiol was also suppressed, suggesting effects of both testosterone and estradiol deficiency.
QCT spine BMD fell substantially in all testosterone-dose groups in which aromatization was also suppressed, and this decline was independent of testosterone dose. Estradiol deficiency disrupted cortical microarchitecture at peripheral sites. Estrogens primarily regulate bone homeostasis in adult men, and testosterone and estradiol levels must decline substantially to impact the skeleton. Gonadal steroids have powerful effects on bone.
During puberty, increases in gonadal steroids stimulate osteoblast activity, causing bone mineral density BMD to increase markedly 1. At midlife in women, gonadal steroids decline to prepubertal levels, which in turn increases bone resorption and leads to rapid bone loss 2. Unlike in women, decreases in gonadal steroids in middle-aged and elderly men are quite modest 3 , 4. If adult men develop severe hypogonadism, however, as happens in men receiving gonadotropin-releasing hormone GnRH agonist therapy for prostate cancer, bone resorption increases and leads to rapid bone loss 5 , 6.
The levels to which gonadal steroids must be reduced to initiate bone loss in men are unknown, however. This issue is clinically important because male aging is associated with modest declines in both gonadal steroids and BMD 7 , but it is not known whether those changes are causally related. Confirmation of causality and identification of thresholds below which skeletal effects are apparent would help to guide evidence-based decisions about testosterone replacement therapy in adult men.
Thus, serum testosterone and estradiol levels are correlated 9 , and the prototypical changes in hypogonadal men — including alterations in body composition, decreases in sexual function, and high turnover bone loss — could be due to androgen deficiency, estrogen deficiency, or both Several findings have been cited as evidence supporting a pivotal role of estrogen in adult male skeletal homeostasis However, they do not definitively establish a causal role for estrogen deficiency in adult male hypogonadal bone loss.
In this study, we utilized a variety of pharmacologic interventions to isolate the specific roles of androgens and estrogens in the pathogenesis of hypogonadal bone loss in adult men and to determine the levels of testosterone and estradiol at which the risk of hypogonadal bone loss begins to increase.
Figure 1 shows details related to recruitment into each of the three cohorts, the randomization of subjects into each of the five testosterone-dose groups in cohorts 1 and 2, and the completion rates for each testosterone-dose group at each study visit. Table 1 shows the baseline characteristics of the three cohorts and Supplemental Table 1 supplemental material available online with this article; doi: There were no significant differences in baseline characteristics between the controls and any of the groups in cohorts 1 or 2.
Reasons for early discontinuation are summarized in Supplemental Table 2. G1, 0 g placebo testosterone gel daily; G2, 1. Figure 2 shows the mean serum testosterone and estradiol levels in relation to testosterone dose in each study group in cohort 1 blue dots , cohort 2 red dots , and the controls black dots. Mean serum estradiol levels ranged from 3. Figure 3 shows the percent change in serum C-terminal telopeptide of type 1 collagen CTX and procollagen type-1 amino-terminal propeptide P1NP in relation both to testosterone dose and to the mean serum testosterone levels during weeks 4—16 in men who received testosterone with cohort 2 and without cohort 1 coadministration of anastrozole.
Within cohort 1 blue dots , serum CTX levels increased significantly more than in the controls black dots only in men who received goserelin plus 0 placebo or 1. Serum CTX even increased substantially more in cohort 2 than in cohort 1 in men who received placebo testosterone Figure 3A , possibly because serum estradiol levels were higher in men who received placebo testosterone alone cohort 1, 3.
Within cohort 2, there was a significant inverse relationship between the testosterone dose and the increase in serum CTX levels.
CTX levels increased more in the groups that received 0, 1. Serum P1NP tended to increase more in groups that received anastrozole than in those that did not, though most of the individual comparisons were not statistically significant.
Within cohort 1, areal BMD of the lumbar spine, total hip, and total body by dual-energy x-ray absorptiometry DXA tended to decline as the dose or level of testosterone declined, though the magnitude of the changes in BMD was small and no change was significantly different from the controls Figure 4, A—F , blue dots.
Specifically, trabecular spine bone loss was detectable 3. There were a few individual group comparisons that were significantly different from the controls, though these differences did not follow any predictable pattern.
Figure 5 shows the changes in bone turnover and BMD in cohort 1 in relation to mean estradiol levels on therapy. Serum CTX levels were stable until serum estradiol levels fell to 5—9. There was a further significant increase in serum CTX levels in men whose estradiol levels were below 5. Subjects in cohort 2 had uniformly low estradiol levels due to anastrozole and are therefore not depicted on this graph.
A potential effect of estradiol on each outcome measure was also assessed by testing for a significant cohort-testosterone dose interaction between cohorts 1 and 2 and by comparing the mean change in each outcome measure between the groups that received any active dosage of testosterone i. Table 2 shows the changes in volumetric BMD vBMD and bone microarchitecture by high resolution peripheral quantitative computed tomography HR-pQCT at the distal radius and distal tibia in a subset of men in cohort 2.
Changes in vBMD at both the radius and the tibia were similar across testosterone-dose groups and were independent of testosterone dose, suggesting that testosterone does not affect vBMD in the setting of estrogen deficiency. However, total vBMD at the tibia declined from baseline within most testosterone-dose groups Supplemental Figure 1 , suggesting an independent effect of estradiol.
Decreases in both cortical and trabecular vBMD contributed to the decline in total vBMD at both the radius and the tibia in men in whom aromatization of testosterone to estradiol was suppressed. There were no significant differences between testosterone-dose groups in the changes of indices of skeletal microarchitecture at either the radius or the tibia Table 2 and Supplemental Figure 2 , again suggesting a lack of testosterone effect in the setting of low estrogen.
Furthermore, changes in indices of microarchitecture were not related to the dosage of testosterone, though there was a nonsignificant trend toward an increase in cortical porosity as the dose of testosterone was lowered. Cortical area decreased and trabecular area increased at both the radius and the tibia when aromatization of testosterone to estradiol was inhibited, suggesting that estrogen deficiency promotes endosteal resorption. There was also a tendency for cortical thickness to decline with estrogen deficiency, particularly at the tibia.
There were no significant changes in trabecular number or trabecular thickness of the radius or the tibia, either in any individual testosterone-dose group or with pooled group analyses.
Results from several types of studies suggest that estradiol plays a significant role in regulating bone metabolism in adult men. The associations are all quite weak, however, generally explaining no more than a few percent of the variation in the data 9 , Case reports of osteoporosis in genetic males with rare but illustrative null mutations of the estrogen receptor ER or androgen receptor AR genes 13 , 14 , or of the aromatase gene 15 , 16 , are frequently cited as evidence that both androgens and estrogens play physiologically important roles in bone metabolism.
However, because these subjects have congenital disorders, it is likely that their skeletal phenotypes reflect effects of gonadal steroids during bone development rather than effects of gonadal steroids on the adult male skeleton. Studies utilizing medications as physiologic probes to produce a state of selective, reversible estrogen deficiency provide the most compelling evidence that estrogen deficiency, rather than androgen deficiency, is responsible for most of the skeletal manifestations in men with adult-onset hypogonadism.
For example, we and others have reported that GnRH agonist-induced suppression of gonadal steroids, coupled with administration of testosterone and a potent aromatase inhibitor to restore serum testosterone levels to the mid-portion of the reference range while estradiol levels remain low , increases bone resorption, though not to the extent observed in men treated with a GnRH agonist alone 17 — In the current study, we extended previously published findings by examining effects of estrogen deficiency over a wide range of testosterone levels.
By administering low doses of testosterone without aromatase blockade as in cohort 1 , we were able to determine the minimal levels of gonadal steroids needed to prevent increases in bone resorption, a finding that should have important clinical implications. By administering a wide range of testosterone doses together with a potent aromatase inhibitor as in cohort 2 , we demonstrated that, as long as estradiol levels remain low, bone resorption increases markedly, even if serum testosterone levels are frankly elevated.
This finding provides compelling evidence of a powerful and independent effect of estrogen deficiency on bone in men. Additionally, our data demonstrate that estradiol deficiency primarily affects cortical bone, particularly by increasing cortical porosity, with little or no effect on trabecular number or thickness.
Although increases in indices of bone resorption with selective estrogen deficiency likely reflect an independent role of estradiol on bone, it is also possible that estrogen deficiency could alter the metabolism of biochemical indices of bone turnover.
Changes in BMD, however, cannot be attributed to theoretical alterations in the clearance of bone turnover markers. Thus, because estradiol deficiency reduced trabecular BMD at the spine, as well as cortical and trabecular vBMD at the distal radius and tibia, the present study provides important new evidence substantiating the role of estradiol deficiency in the pathogenesis of hypogonadal bone loss.
When considered together with our prior findings that estrogen deficiency plays a key role in fat accumulation and sexual dysfunction in hypogonadal men 10 , it is now clear that estrogen deficiency plays an important, if not dominant, role in many of the key clinical features of male hypogonadism.
Classifying features of male hypogonadism based on their relationship to androgen or estrogen deficiency might permit tailoring of therapies for features of male hypogonadism based on their underlying pathogenesis.
Whether androgen deficiency also exerts an independent effect on bone homeostasis in adult men is less clear. While it is feasible to administer testosterone to men with low estradiol levels, we did not administer estradiol to testosterone-deficient men because of concerns that some men would develop gynecomastia if exposed to such a hormonal milieu for several months. Thus, we cannot determine if testosterone deficiency by itself contributes to the skeletal effects of male hypogonadism.
One group did administer estradiol to testosterone-deficient men, albeit for only 3 weeks to minimize safety concerns, and reported a significant increase in bone resorption indices, though the increase was much smaller than when men were rendered selectively estrogen deficient In the current study, the decline in bone resorption with increasing doses of testosterone in cohort 2 that persisted after adjustment for the small increase in estradiol levels is consistent with an independent effect of androgens on bone resorption, though the skeletal effects of androgens appear to be considerably less potent than the effects of estrogens.
The effects of gonadal steroids on bone have also been investigated extensively in animals, though with variable results and conclusions. Bone mass is also reduced in male mice with null mutations in the AR gene, indicating that androgens also exert effects on bone Although these genetic mouse models provide convincing evidence that gonadal steroids have powerful effects on bone, these models most likely reflect effects of gonadal steroids on bone development, not effects on adult bone homeostasis.
Animal studies examining effects of gonadal steroids on adult bone have reached vastly different conclusions. For example, in male rats, BMD of the femur and lumbar vertebrae declined to a similar extent in animals that underwent orchiectomy or that were treated with a potent aromatase inhibitor, which caused selective estrogen deficiency These data suggest that all of the effects of severe acquired androgen and estrogen deficiency on bone density in male rats can be attributed to the effects of estrogen deficiency alone.
Finally, both aromatizable and nonaromatizable androgens can prevent bone loss from orchiectomy in aged male rats, even when combined with an aromatase inhibitor 27 — This finding suggests that all of the effects of gonadal steroids on bone in male rats can be attributed to the effects of androgens. These conflicting results from animal models are difficult to reconcile, though the use of supraphysiological doses of androgens in some studies may confound the interpretation of the physiological effects of gonadal steroids.
To reconcile these findings from animal and human studies, it has been hypothesized that androgens act primarily on trabecular bone while estrogens act primarily on cortical bone.
In male mice with a targeted deletion in the AR gene in osteoblast lineage cells, trabecular bone volume and trabecular number are reduced, demonstrating that androgens exert effects on trabecular bone directly via the AR 30 , In clinical studies, it has been postulated that the effect of androgens on trabecular bone is difficult to detect utilizing biochemical markers of bone resorption because changes in bone resorption markers reflect the effects of gonadal steroids on the entire skeleton, which is mainly composed of cortical bone Furthermore, our findings from HR-pQCT do suggest that changes in cortical bone were due largely to estrogen deficiency.
However, we did not find any evidence of a testosterone effect on trabecular bone, either at the lumbar spine by QCT or at the distal radius or tibia by HR-pQCT.
Instead, we found that trabecular BMD loss was also regulated solely by estrogen. Our data are also consistent with cross-sectional studies that have found that serum estradiol is significantly associated with cortical and trabecular bone microarchitecture at peripheral sites in older men 33 , It remains possible, however, that the regulation of cortical and trabecular bone by gonadal steroids varies throughout the skeleton.
There is no consensus regarding the levels of gonadal steroids at which replacement should begin for skeletal health in hypogonadal men. This difficulty arises, at least in part, because the relationships between gonadal steroid levels and indices of bone health are likely better represented by a continuum rather than a distinct threshold.
Nonetheless, clinicians must ultimately decide whether to give hormone replacement to each hypogonadal man. Although our data suggest that estrogen deficiency plays the dominant pathophysiologic role in hypogonadal bone loss, reliable estradiol assays are not widely available for values in the range seen in hypogonadal men. In those instances, serum testosterone levels appear to serve as a reasonable proxy for estradiol levels. Additionally, the decision to treat an individual patient with testosterone should be based on its potential effects on multiple organ systems, not just bone.
Even though the preponderance of data indicate that estradiol is the major regulator of bone metabolism in adult men, very little is known about the level s of estradiol needed to ensure stability of the skeleton, and current treatment recommendations do not consider estradiol levels.
These discrepancies may be due, at least in part, to the difficulty in measuring low levels of estradiol by radioimmunoassay.