Differential Effects of Androgens and Estrogens on Bone Turnover in ...

0013-7227/03/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 88(1):204 –210 Copyright © 2003 by The Endocrine Society doi: 10.1210/jc.2002-021036

Differential Effects of Androgens and Estrogens on Bone Turnover in Normal Men BENJAMIN Z. LEDER, KAREN M. LEBLANC, DAVID A. SCHOENFELD, RICHARD EASTELL, JOEL S. FINKELSTEIN

AND

Endocrine Unit (B.Z.L., K.M.L., J.S.F.) and Biostatistics Center (D.A.S.), Department of Medicine, Massachusetts General Hospital, Boston, Massachusetts 02114; and Division of Clinical Sciences (R.E.), Northern General Hospital, Sheffield, United Kingdom The physiologic role of androgens and estrogens in the maintenance of normal bone turnover in men is a fundamental issue in bone biology. To address this question, we randomized 70 men between the ages of 20 and 44 yr to receive one of three treatment regimens. Group 1 (n ⴝ 25) received a GnRH analog (goserelin acetate 3.6 mg by sc injection every 4 wk) alone for 12 wk to suppress endogenous gonadal steroids to prepubertal levels. Group 2 (n ⴝ 22) received goserelin plus transdermal testosterone (Androderm 5 mg topically daily) to normalize circulating testosterone and estradiol levels. Group 3 (n ⴝ 23) received goserelin plus testosterone plus an aromatase inhibitor (anastrozole 1 mg orally daily) to induce selective estrogen deficiency. The selective effects of androgens and estrogens on skeletal homeostasis were then assessed by measuring changes in biochemical markers of bone turnover and analyzing the between-group differences. Bone resorption markers increased in both the hypogonadal group (group 1) and in the group with selective estrogen deficiency (group 3). Urinary deoxypyridinoline excretion increased more in group 1 than in group 3 (P ⴝ 0.023), suggesting a

significant effect of androgens on bone resorption, whereas serum N-telopeptide levels increased more in group 3 than in group 2 (P ⴝ 0.037), suggesting a significant effect of estrogen on bone resorption. Bone formation markers initially declined in all groups and then increased in groups 1 and 3. The between-group comparisons were consistent for all formation markers. Bone formation markers increased more in group 1 than in group 2 (P ⴝ 0.001, 0.037, 0.005 for osteocalcin, carboxyterminal propeptide of type I procollagen, and amino-terminal propeptide of type I procollagen, respectively). Bone formation markers also increased more in group 1 than in group 3 but these differences were not statistically significant (P ⴝ 0.065 0.073, 0.099 for osteocalcin, carboxy-terminal propeptide of type I procollagen, and amino-terminal propeptide of type I procollagen, respectively). Taken together, these data suggest that both androgens and estrogens play independent and fundamental roles in regulating bone resorption in men. These data also suggest that androgens may play an important role in the regulation of bone formation in men. (J Clin Endocrinol Metab 88: 204 –210, 2003)

E

tered to the men with aromatase deficiency, the skeleton matured and BMD increased dramatically (10 –12). Although the phenotypes of men with these genetic mutations imply an essential role for estrogen in male skeletal development, data suggesting a role for estrogen in male bone homeostasis in humans come largely from epidemiological association studies and a single short-term physiological study in which states of selective androgen and estrogen deficiency were created in elderly men for a short period of time (13). To test the hypothesis that estrogen is the major regulator of bone maintenance in men, we studied 70 healthy young men for 12 wk. All men received a GnRH analog to induce temporary hypogonadism, and hormones were selectively replaced in the three groups to create the following physiological states: 1) testosterone and estrogen deficiency (⫺T, ⫺E), 2) testosterone and estrogen sufficiency (⫹T, ⫹E), and 3) selective estrogen deficiency (⫹T, ⫺E). We then measured biochemical markers of bone turnover every 4 wk and analyzed the differences in the changes among groups to assess the relative roles of androgens and estrogens in male bone homeostasis.

STROGEN DEFICIENCY WAS identified as the causative factor in postmenopausal osteoporosis over 60 yr ago (1), and since that time estrogen has been recognized as the principal gonadal steroid regulator of bone turnover in women. Conversely, it had long been assumed that androgens played an analogous dominant role in male bone metabolism. This hypothesis was based largely on the clinical observation that hypogonadism was associated with osteopenia in men (2) and the unambiguous data demonstrating direct androgen action in vitro (3– 6) and in animals (6, 7). Over the past several years, however, the dominant role of androgens in male bone physiology has been increasingly questioned as data have emerged suggesting an important role for estrogens in male skeletal development and homeostasis. First, case reports of men with rare genetic mutations affecting either the estrogen receptor-␣ gene or the aromatase gene suggested a vital role for estrogen in male bone physiology (8 –10). These men, who were unable to either respond to or synthesize estrogen had reduced bone mineral density (BMD), elevated biochemical markers of bone turnover, and unfused epiphyses (8 –10). Furthermore, when estrogen was adminis-

Subjects and Methods Abbreviations: BMD, Bone mineral density; DPD, deoxypyridinoline; NTX, N-telopeptide; OC, osteocalcin; PICP, carboxy-terminal propeptide of type I procollagen; PINP, amino-terminal propeptide of type I procollagen.

Study subjects Ninety-two men between the ages of 20 and 44 yr were recruited using advertisements in accord with institutional guidelines for clinical

204

Leder et al. • Androgens, Estrogens, and Bone Turnover in Men

studies. All subjects were required to have normal serum testosterone, estradiol, LH, and FSH concentrations; normal renal and hepatic function; and normal BMD of the lumbar spine, femoral neck, and total hip. Subjects were excluded from participation if they had any history of congenital or acquired bone disease (osteomalacia, vitamin D deficiency, hyperparathyroidism, Paget’s disease), recent fracture or immobilization, hyperthyroidism, cardiopulmonary disease, malignancy, benign prostatic hyperplasia, major psychiatric disease, or drug or alcohol abuse. Subjects were also excluded if they were using any drug known to interact with the study medications or any drug known to affect bone turnover markers (such as anticoagulants, gonadal steroids, glucocorticoids, anticonvulsants, and suppressive doses of thyroxine, lithium, bisphosphonates, calcitonin, or sodium fluoride). Because the goal of this study was to determine the effects of specific hormonal milieus on bone turnover, it was predetermined that we would not use data from men whose serum testosterone (groups 2 and 3) and estradiol (group 2) levels were less than 2 sd below the mean of the entire study population on more than one visit. Bone turnover markers were not analyzed on the samples from those men (n ⫽ 6). An additional eight men withdrew because of side effects of the study medications and eight for other reasons, leaving 70 men who form the basis of this report. This study was approved by the Human Research Committee of Partners HealthCare Systems, and all subjects provided written informed consent.

Study protocol Subjects were randomized by computer-generated assignment to one of three treatment groups (Fig. 1). Subjects in group 1 received the GnRH analog goserelin acetate (Zoladex; AstraZeneca, Wilmington, DE) 3.6 mg by sc injection every 4 wk for 12 wk. Subjects in group 2 received goserelin acetate 3.6 mg by sc injection every 4 wk plus physiological testosterone and estradiol replacement via an aromatizable topical testosterone delivery system (Androderm 5-mg patch daily; Watson, Corona, CA). Subjects in group 3 received goserelin acetate 3.6 mg by sc injection every 4 wk plus topical testosterone plus a potent thirdgeneration aromatase inhibitor anastrozole (Arimidex; AstraZeneca) 1 mg orally daily. The combined effect of these specific medication regimens was to create groups of subjects that were either testosterone and estrogen deficient [group 1 (⫺T, ⫺E), n ⫽ 25], eugonadal [group 2 (⫹T, ⫹E), n ⫽ 22], or selectively estrogen deficient [group 3 (⫹T, ⫺E), n ⫽ 23]. If a subject had a local reaction to Androderm topical testosterone, the equivalent dose of Testoderm TTS (Alza Corp., Mountain View, CA) topical testosterone was administered. If the local reaction continued, Androgel (5 g packet QD, Unimed Pharmaceuticals, Inc., Deerfield, IL) topical testosterone was administered. Subjects whose diets did not include at least 1000 mg elemental calcium daily were encouraged to increase calcium intake by either diet or calcium supplements. Subjects were seen on the Mallinckrodt General Clinical Research Center (GCRC) at 4-wk intervals for a total of 12 wk (wk 0, 4, 8, and 12). At each visit, a GCRC nurse administered the goserelin injection, and the other medications were administered to each subject according to their assigned group. To monitor compliance, subjects were asked to complete a drug history questionnaire at each visit and return empty medication containers and any unused medication. Morning fasting blood

FIG. 1. Representation of the subject randomization. All treatments lasted 12 wk.

J Clin Endocrinol Metab, January 2003, 88(1):204 –210 205

and second void urine samples were collected before study medication was administered at each visit to assess hormone levels and biochemical markers of bone turnover. Thus, in subjects receiving testosterone replacement, the reported testosterone and estradiol values represent serum concentrations measured approximately 24 h after patch application.

Laboratory methods Gonadal steroid and calcium regulatory hormones. Serum testosterone was measured by RIA using a commercial kit (Diagnostic Products, Los Angeles, CA) with a sensitivity of 4 ng/dl, an intraassay coefficient of variation of 5% within the normal range and 18% in the castrate range, and an interassay coefficient of variation of 7–12%. Serum estradiol was measured using an ultrasensitive competitive RIA after extraction and chromatographic purification (Nichols Institute, San Juan Capistrano, CA). The sensitivity of this assay is 2 pg/ml, and the intra- and interassay coefficients of variation are 5– 6% and 6 –13%, respectively. Serum SHBG was measured using a chemiluminescent immunoassay (Nichols Institute) with a sensitivity of 1 nmol/liter and intra- and interassay coefficients of variation of 3–5% and 3–5%, respectively. Serum intact PTH was measured by a two-site immunoradiometric assay (Nichols Institute) with a sensitivity of 1 pg/ml and intra- and interassay coefficients of variation of 2–3% and 6%, respectively. Serum 25-(OH) vitamin D was measured using an extraction double-antibody RIA (DiaSorin, Inc., Stillwater, MN) with a sensitivity of 1.5 ng/ml and intra- and interassay coefficients of variation of 9 –13% and 8 –11%, respectively. Serum calcium was measured by autoanalyzer. Biochemical markers of bone resorption. Urine deoxypyridinoline (DPD) was measured using a competitive enzyme-linked immunoassay (Quidel Corp., San Diego, CA) with a sensitivity of 1.1 nm bone collagen equivalents and intra- and interassay coefficients of variation of 5–9% and 4 – 8%, respectively. Urine N-telopeptide (NTX) was measured in Sheffield, UK, using an automated chemiluminescent immunoassay (Ortho-Clinical Diagnostics, Rochester, NY) with an interassay CV of 7%. Because these samples were not run in duplicate, no intraassay coefficient of variation was available. Urine creatinine was measured by a dry slide chemistry method (Ortho-Clinical Diagnostics). Serum NTX was measured using a competitive inhibition enzyme immunoassay (Osteomark, Ostex International, Inc., Seattle, WA) with a sensitivity of 1 nm BCE and intra- and interassay coefficients of variation of 6% and 9%, respectively. Biochemical markers of bone formation. Serum osteocalcin (OC) was measured using a double-antibody immunoradiometric assay (Nichols Institute, San Juan Capistrano, CA) with a sensitivity of 0.5 ng/ml and intra- and interassay coefficients of variation of 2– 4% and 3– 6%, respectively. Serum carboxy-terminal propeptide of type I procollagen (PICP) was measured using an equilibrium RIA (DiaSorin, Inc.) with a sensitivity of 1.2 ng/ml and intra- and interassay coefficients of variation of 1– 4% and 1–2%, respectively. Serum amino-terminal propeptide of type I procollagen (PINP) was measured using a RIA (Orion Diagnos-

206

J Clin Endocrinol Metab, January 2003, 88(1):204 –210


tica, Espoo, Finland) with a sensitivity of 2 ng/ml and intra- and interassay coefficients of variation of 5–14% and 2–3%, respectively. Each subject’s samples for serum and urine bone turnover markers were analyzed in the same assay.

Data analysis The primary end points of this study were between-group differences in changes of biochemical markers of bone turnover. Pairwise differences among groups for each marker were assessed using a repeatedmeasures ANOVA using the PROC MIXED procedure in SAS (SAS Institute, Cary, NC). The model included baseline level, visit number, group, and a group-visit interaction. We did not specify the variance covariance matrix for the repeated measurements. For a subset of markers (OC, PICP, PINP), the group-visit interaction was significant, indicating that the treatment effect was different at each visit. Thus, for these markers, the change from baseline at 12 wk was assessed using a simple analysis of covariance. Baseline characteristics of the three groups were compared by ANOVA. P values less than 0.05 were considered statistically significant. Unless specified otherwise, all data are presented as the mean ⫾ sd.

Results

Table 1 shows the baseline clinical characteristics, gonadal steroid levels, and SHBG levels of the study subjects. There were no statistically significant differences among groups. Table 2 shows the baseline biochemical markers of bone turnover and calcium regulatory hormones of study subjects. Subjects were generally well matched in all these areas as well. Baseline urinary NTX excretion was approximately 15% higher in group 1 than in groups 2 and 3, but these differences were not statistically significant (P ⫽ 0.179 for group 1 vs. 2 and P ⫽ 0.129 for group 1 vs. 3). Hormone levels. Figure 2 shows the changes in serum testosterone, estradiol, and SHBG levels during the 12-wk study period. As expected, serum testosterone concentrations did not change significantly from baseline in either groups 2 or 3 but decreased dramatically in group 1 by wk 4 and remained in the castrate range for the duration of the study (P ⬍ 0.001 for change from baseline at wk 4 –12). Changes in serum testosterone levels between groups 2 and 3 were similar (P ⫽ 0.721), whereas serum testosterone levels were significantly lower in group 1 than in groups 2 and 3 (P ⬍ 0.001 for both comparisons). Also as expected, serum estradiol levels decreased into the castrate range in groups 1 and 3 by wk 4 and remained suppressed throughout the study period (P ⬍ 0.001 for change from baseline at wk 4 –12 for both groups). Serum estradiol levels in group 2 remained stable overall but declined transiently at wk 8 (mean ⫾ sd serum estradiol level 62 ⫾ 37 pmol/liter, P ⫽ 0.001 for change TABLE 1. Baseline clinical characteristics and gonadal steroid concentrations

No. Age (yr) BMI (kg/m2) T (nmol/liter) E (pmol/liter) SHBG (nmol/liter)

Group 1 (⫺T, ⫺E)

Group 2 (⫹T, ⫹E)

Group 3 (⫹T, ⫺E)

25 30.1 ⫾ 5.6 25.6 ⫾ 4.3 18.8 ⫾ 5.5 110 ⫾ 48 21.0 ⫾ 10.7

22 34.2 ⫾ 7.9 26.8 ⫾ 4.6 17.4 ⫾ 6.1 103 ⫾ 37 23.6 ⫾ 10.5

23 32.7 ⫾ 6.2 27.4 ⫾ 3.3 17.7 ⫾ 7.6 92 ⫾ 26 20.0 ⫾ 11.8

Values are reported as mean ⫾ SD. No statistically significant differences between groups were observed. T, Testosterone; E, estradiol.

TABLE 2. Baseline biochemical markers of bone turnover and calcium regulatory hormones Group 1 (⫺T, ⫺E)

Serum NTX (nmol/liter) Urine NTX (nmol/mmol) Urine DPD (nmol/mmol) Osteocalcin (ng/ml) PICP (ng/ml) PINP (ng/ml) PTH (ng/liter) 25-(OH)D (nmol/liter) Calcium (nmol/liter)

Group 2 (⫹T, ⫹E)

Group 3 (⫹T, ⫺E)

14.1 ⫾ 2.7 14.1 ⫾ 4.9 14.2 ⫾ 5.2 49.0 ⫾ 18.1 41.8 ⫾ 20.1 41.1 ⫾ 15.5 4.7 ⫾ 1.1

4.8 ⫾ 1.7

4.7 ⫾ 1.2

28.7 ⫾ 7.9 163 ⫾ 50 56 ⫾ 16 34 ⫾ 18 65 ⫾ 25 2.5 ⫾ 0.1

29.0 ⫾ 7.3 173 ⫾ 73 56 ⫾ 18 30 ⫾ 13 70 ⫾ 37 2.4 ⫾ 0.1

30.6 ⫾ 10.2 160 ⫾ 52 54 ⫾ 17 30 ⫾ 8 67 ⫾ 20 2.5 ⫾ 0.1

Values are reported as mean ⫾ SD. No statistically significant differences between groups were observed. T, Testosterone; E, estradiol.

from baseline). Serum estradiol levels declined more in groups 1 and 3 than in group 2 (P ⬍ 0.001 for both comparisons), but serum estradiol levels declined to a similar extent in groups 1 and 3 (P ⫽ 0.649). Serum SHBG levels increased in group 1 (P ⫽ 0.018, 0.056, 0.081 for change from baseline at wk 4, 8, and 12, respectively) but did not change in groups 2 or 3. This increase in SHBG in group 1 was significant only when compared with group 3 (P ⫽ 0.008). Bone resorption markers. Figure 3 shows the changes in biochemical markers of bone resorption during the 12-wk study period. Urinary DPD excretion increased significantly from baseline in group 1 (P ⫽ 0.035 for change from baseline at wk 4 and P ⬍ 0.001 at wk 8 and 12) but did not change in group 2 (P ⫽ 0.734, 0.551, and 0.981 at wk 4, 8, and 12, respectively). In group 3, urinary DPD excretion increased significantly from baseline by wk 8, although this increase was of only borderline significance at wk 12 (P ⫽ 0.215, 0.028, and 0.076 for change from baseline at wk 4, 8, and 12, respectively). Urinary DPD excretion increased more in group 1 than in groups 2 and 3 (P ⬍ 0.001 for group 1 vs. 2, P ⫽ 0.023 for group 1 vs. 3). Although urinary DPD excretion increased more in group 3 than group 2, this difference was not statistically significant (P ⫽ 0.136). Serum NTX levels increased significantly in group 1 (P ⫽ 0.012, P ⬍ 0.001 for change from baseline at wk 4 and 8 –12, respectively). Serum NTX levels also increased in group 3 (P ⫽ 0.082, P ⫽ 0.012, P ⬍ 0.001 for change from baseline at wk 4, 8, and 12, respectively). In group 2, serum NTX was stable at wk 4 and 8 but increased by approximately 12% by wk 12 (P ⫽ 0.011). Serum NTX levels increased more in groups 1 and 3 than in group 2 (P ⫽ 0.002 for group 1 vs. 2, P ⫽ 0.037 for group 3 vs. 2). Serum NTX levels increased to a similar extent in groups 1 and 3 (P ⫽ 0.258). Changes in urinary NTX excretion showed a similar pattern to changes in serum NTX (both in regard to change from baseline and the between-group comparisons), but none of these comparisons was statistically significant, probably because of the higher variability of this marker. Bone formation markers. Figure 4 shows the changes in biochemical markers of bone formation during the 12-wk study period. Serum OC, PICP, and PINP levels decreased initially in all groups and then increased to varying degrees in groups



FIG. 2. Mean (⫾SEM) change in serum gonadal steroid and SHBG levels at 4-wk intervals. The P values refer to the between-group comparisons indicated in the figure.

1 and 3. In contrast, serum OC, PICP, and PINP levels remained significantly suppressed in group 2 (P ⫽ 0.005, 0.040, 0.013 for change from baseline at wk 12 for OC, PICP, and PINP, respectively). The between-group comparisons were also consistent for all formation markers. Serum OC, PICP, and PINP levels increased more in group 1 than in group 2 (P ⫽ 0.001, 0.037, 0.005 for OC, PICP, and PINP, respectively) or group 3, although the differences between changes in groups 1 and 3 were of borderline statistical significance (P ⫽ 0.065, 0.073, 0.099 for OC, PICP, and PINP, respectively). No significant differences were observed in changes in serum OC, PICP, and PINP levels between groups 2 and 3. Discussion

This study explores the relative effects of androgens and estrogens on bone metabolism in men. Our model included three groups of men. Group 1 was deficient in both androgens and estrogens. Group 2 maintained normal adult levels of androgens and estrogens, whereas group 3 was selectively estrogen deficient. Thus, comparisons between groups 1 and 3 indicated the effect of androgens on bone turnover, and comparisons between groups 2 and 3 indicated the effects of estrogens on bone turnover. Urinary DPD excretion, a marker of bone resorption, increased more in group 1 than group 3 demonstrating an independent and physiologically important effect of androgens on bone resorption in adult men. Although urinary DPD excretion also increased more in group 3 than group 2, these changes were of only borderline significance, suggesting, but not definitively demonstrating, an independent effect of estrogens on this mea-

sure of bone resorption. In contrast, levels of serum NTX, another marker of bone resorption, increased more in group 3 than group 2, clearly demonstrating an independent and physiologically important effect of estrogens on bone resorption in adult men. Although serum NTX levels also increased more in group 1 than group 3, these changes were not significant. It is not clear why different measures of bone resorption give different results, although it is possible that gonadal steroids differentially affect either the renal handling of NTX and DPD or the breakdown of collagen from tissues other than bone. Nonetheless, the NTX and DPD data are qualitatively similar in that the values in group 3 were intermediate between groups 1 and 2. Thus, taken together, it appears that both androgens and estrogens play physiologically important roles in the regulation of bone resorption in men. The effects of selective gonadal steroid deprivation on markers of osteoblast activity demonstrated in this study were complex. These markers decreased initially, possibly because of the initial increase in serum testosterone and estradiol levels that occurs upon initiating GnRH analog therapy (14). After this decrease, these markers increased in the hypogonadal group (group 1) as is known to occur during long-term hypogonadism (15). Moreover, there was a trend toward a greater increase in bone formation markers in group 1, compared with group 3. This finding suggests that physiological levels of androgens may normally suppress osteoblast activity, although a larger study may be needed to establish this fact. The findings in this study are largely consistent with

208



FIG. 3. Mean (⫾SEM) percentage change in urine and serum biochemical markers of bone resorption at 4-wk intervals. The P values refer to the between-group comparisons indicated in the figure.

emerging data concerning the roles of androgens and estrogens in male bone physiology. Case reports have shown that men with mutations in either the estrogen receptor-␣ gene or the aromatase gene are osteopenic (8 –10), demonstrating the importance of estrogen in normal male skeletal development. Numerous studies have examined associations between serum testosterone or estrogen levels and BMD in adult men. Although some studies have reported a positive association between serum testosterone levels and BMD in men (16 –18), most have found a stronger association between serum estradiol levels and BMD or fracture (19 –27). The preponderance of the in vitro and animal data suggests crucial roles for both androgens and estrogens in male bone metabolism, beginning with the fundamental observation that both androgen and estrogen receptors are expressed on osteoblasts (28 –30). Aromatase inhibition reduces BMD as effectively as castration in male rats (31, 32), and mice with targeted inactivation of the aromatase gene are osteopenic (33, 34). In addition, mice with targeted inactivation of the estrogen receptor-␣ or both estrogen receptor-␣ and -␤ have cortical (but not trabecular) osteopenia (35). These findings suggest an important independent effect of estrogens on bone homeostasis in animals. Conversely, both aromatizable and nonaromatizable androgens stimulate osteoblast proliferation and differentiation in vitro (3, 4) and prevent osteoblast apoptosis via a nongenotropic mechanism (36). Moreover, in cell culture, both aromatizable and nonaromatizable androgens reduce the resorption activity of osteoclasts (5), and in rodents, administration of either aromatizable or nonaromatizable androgens prevents orchidectomy-induced os-

teoclastogenesis and bone loss (6, 7, 37). Finally, trabecular osteopenia develops in estrogen receptor-␣ and estrogen receptor-␣ and -␤ knockout male mice after orchidectomy, presumably because of the loss of testosterone (38). Taken together these findings suggest important independent effects of androgens on bone turnover. Few studies have examined the differential effects of androgens and estrogens on bone turnover in humans. Bone turnover markers do not change consistently in elderly men given estradiol (39), but C-telopeptide levels increase in elderly men given an aromatase inhibitor (40). These studies are difficult to interpret, however, because testosterone levels decrease with estradiol administration and increase with aromatase inhibition, thus confounding the effects of the primary interventions. Falahati-Nini et al. (13) used combinations of GnRH analog, testosterone, estrogen, and aromatase administration to induce selective estrogen deficiency, selective androgen deficiency, combined gonadal steroid deficiency, or combined gonadal steroid replacement for 3 wk in elderly men. Using that model, they reported that estrogen is the dominant regulator of bone resorption in men, whereas both androgens and estrogens regulate bone formation. Qualitatively, our bone resorption results are similar to those of Falahati-Nini et al. (13), although the changes in urinary DPD excretion in our study clearly indicated an important independent effect of androgens on bone resorption. The effects of gonadal steroids on bone formation markers in the two studies differ more substantially. Whereas Falahati-Nini et al. concluded that both androgens and es-



FIG. 4. Mean (⫾SEM) percentage change in serum biochemical markers of bone formation (or osteoblast activity) at 4-wk intervals. The P values refer to the between-group comparisons indicated in the figure.

trogens are important in maintaining bone formation, we found only an effect of androgens. Although there are many similarities between the designs of these two studies, there are several important differences. First, in the study by Falahati-Nini et al., the different hormonal milieus were maintained for only 3 wk. As is clearly shown in our data, bone formation is not in a steady state at this time. Moreover, bone turnover markers in the study by Falahati-Nini et al. (13) may not have been in a steady state at the time the individual hormonal manipulations were initiated because of supraphysiological increases in gonadal steroids that occur during the initial agonist phase of GnRH analog therapy (14). Finally, Falahati-Nini et al. studied elderly men, whereas we studied young adult men. It is possible that the skeletons of young and old men respond differently to androgens and estrogens, an issue that may have contributed to the subtle differences in the findings or the magnitude of the observed changes. Two limitations of our study deserve mention. First, although we attempted to keep serum estradiol levels stable in group 2 by allowing topical testosterone to be endogenously aromatized to estradiol, mean serum estradiol levels decreased slightly at wk 8 in this group, although they remained within the normal male range. This slight decrease in serum estradiol levels may have occurred because peripheral aromatization accounts for most, but not all, of circulating estradiol in men (41).

Moreover, this transient decrease in serum estradiol levels may explain the slight increase in serum NTX levels in group 2 and may have prevented our ability to detect a statistically significant difference in urine DPD excretion between groups 2 and 3, minimizing our appreciation of an estradiol effect. Importantly, however, this transient decrease in serum estradiol levels in group 2 has no effect on the assessment of the selective effect of testosterone on bone turnover. In addition, we considered including a fourth treatment group (i.e. a goserelin plus estradiol) to assess the effects of selective estrogen replacement in the absence of testosterone. Although including this group would likely have provided additional useful information, the risk of inducing feminizing side effects during 12 wk of selective estrogen administration was thought to be prohibitory. In conclusion, our findings suggest that both androgens and estrogens are independent mediators of bone resorption in young adult men. Although quantifying the effect of each hormone is not possible in the present study, the importance of both androgens and estrogens needs to be considered when assessing the effects of all hormonal therapies or manipulations in men. Although the effects of gonadal steroids on bone formation are more complex, there may be similar selective effects of gonadal steroids on osteoblast function as well. Longer-term studies are now needed to confirm and expand our understanding in this fundamental area.

210



Acknowledgments We are grateful to the nursing staff of the Mallinckrodt General Clinical Research Center for their meticulous performance of the study protocol, Dr. Rosemary Hannon and Gregory Neubauer for help with the analysis of the bone turnover markers, and AstraZeneca and Watson Pharmaceuticals for the generous supply of the study medications. Received July 5, 2002. Accepted October 14, 2002. Address all correspondence and requests for reprints to: Benjamin Z. Leder, M.D., Endocrine Unit, Massachusetts General Hospital, Bulfinch 327, Fruit Street, Boston, Massachusetts 02114. E-mail: bzleder@ partners.org. This work was supported by NIH Grants F32-AR08574 (to B.Z.L.), K23-RR16310 (to B.Z.L.), and K24-DK02759 (to J.S.F.); Massachusetts General Hospital Clinical Research Center Grant RR-1066; and an Endocrine Fellows Foundation grant (to B.Z.L.).

References 1. Albright F, Bloomberg E, Smith PH 1940 Post-menopausal osteoporosis. Trans Assoc Am Phys 55:298 –305 2. Orwoll ES 1998 Osteoporosis in men. Endocrinol Metab Clin North Am 27:349 –367 3. Kasperk CH, Wergedal JE, Farley JR, Linkhart TA, Turner RT, Baylink DJ 1989 Androgens directly stimulate proliferation of bone cells in vitro. Endocrinology 124:1576 –1578 4. Kasperk CH, Wakley GK, Hierl T, Ziegler R 1997 Gonadal and adrenal androgens are potent regulators of human bone cell metabolism in vitro. J Bone Miner Res 12:464 – 471 5. Pederson L, Kremer M, Judd J, Pascoe D, Spelsberg TC, Riggs BL, Oursler MJ 1999 Androgens regulate bone resorption activity of isolated osteoclasts in vitro. Proc Natl Acad Sci USA 96:505–510 6. Bellido T, Jilka RL, Boyce BF, Girasole G, Broxmeyer H, Dalrymple SA, Murray R, Manolagas SC 1995 Regulation of interleukin-6, osteoclastogenesis, and bone mass by androgens. The role of the androgen receptor. J Clin Invest 95:2886 –2895 7. Wakley GK, Schutte Jr HD, Hannon KS, Turner RT 1991 Androgen treatment prevents loss of cancellous bone in the orchidectomized rat. J Bone Miner Res 6:325–330 8. Smith EP, Boyd J, Frank GR, Takahashi H, Cohen RM, Specker B, Williams TC, Lubahn DB, Korach KS 1994 Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N Engl J Med 331:1056 –1061 9. Morishima A, Grumbach MM, Simpson ER, Fisher C, Qin K 1995 Aromatase deficiency in male and female siblings caused by a novel mutation and the physiological role of estrogens. J Clin Endocrinol Metab 80:3689 –3698 10. Carani C, Qin K, Simoni M, Faustini-Fustini M, Serpente S, Boyd J, Korach KS, Simpson ER 1997 Effect of testosterone and estradiol in a man with aromatase deficiency. N Engl J Med 337:91–95 11. Bilezikian JP, Morishima A, Bell J, Grumbach MM 1998 Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. N Engl J Med 339:599 – 603 12. Rochira V, Faustini-Fustini M, Balestrieri A, Carani C 2000 Estrogen replacement therapy in a man with congenital aromatase deficiency: effects of different doses of transdermal estradiol on bone mineral density and hormonal parameters. J Clin Endocrinol Metab 85:1841–1845 13. Falahati-Nini A, Riggs BL, Atkinson EJ, O’Fallon WM, Eastell R, Khosla S 2000 Relative contributions of testosterone and estrogen in regulating bone resorption and formation in normal elderly men. J Clin Invest 106:1553–1560 14. Cockshott ID 2000 Clinical pharmacokinetics of goserelin. Clin Pharmacokinet 39:27– 48 15. Smith MR, McGovern FJ, Zietman AL, Fallon MA, Hayden DL, Schoenfeld DA, Kantoff PW, Finkelstein JS 2001 Pamidronate to prevent bone loss during androgen-deprivation therapy for prostate cancer. N Engl J Med 345:948 –955 16. Kelly PJ, Pocock NA, Sambrook PN, Eisman JA 1990 Dietary calcium, sex hormones, and bone mineral density in men. BMJ 300:1361–1364 17. Murphy S, Khaw KT, Cassidy A, Compston JE 1993 Sex hormones and bone mineral density in elderly men. Bone Miner 20:133–140 18. Van den Beld AW, de Jong FH, Grobbee DE, Pols HA, Lamberts SW 2000 Measures of bioavailable serum testosterone and estradiol and their relationships with muscle strength, bone density, and body composition in elderly men. J Clin Endocrinol Metab 85:3276 –3282 19. Khosla S, Melton 3rd LJ, Atkinson EJ, O’Fallon WM, Klee GG, Riggs BL 1998

20.

21. 22. 23.

24.

25. 26. 27.

28. 29. 30. 31. 32. 33. 34. 35.

36.

37. 38. 39. 40. 41.

Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: a key role for bioavailable estrogen. J Clin Endocrinol Metab 83:2266 –2274 Slemenda CW, Longcope C, Zhou L, Hui SL, Peacock M, Johnston CC 1997 Sex steroids and bone mass in older men. Positive associations with serum estrogens and negative associations with androgens. J Clin Invest 100:1755– 1759 Greendale GA, Edelstein S, Barrett-Connor E 1997 Endogenous sex steroids and bone mineral density in older women and men: the Rancho Bernardo Study. J Bone Miner Res 12:1833–1843 Center JR, Nguyen TV, Sambrook PN, Eisman JA 1999 Hormonal and biochemical parameters in the determination of osteoporosis in elderly men. J Clin Endocrinol Metab 84:3626 –3635 Ongphiphadhanakul B, Rajatanavin R, Chanprasertyothin S, Piaseu N, Chailurkit L 1998 Serum oestradiol and oestrogen-receptor gene polymorphism are associated with bone mineral density independently of serum testosterone in normal males. Clin Endocrinol (Oxf) 49:803– 809 Amin S, Zhang Y, Sawin CT, Evans SR, Hannan MT, Kiel DP, Wilson PW, Felson DT 2000 Association of hypogonadism and estradiol levels with bone mineral density in elderly men from the Framingham Study. Ann Intern Med 133:951–963 Szulc P, Marchand F, Duboeuf F, Delmas PD 2000 Cross-sectional assessment of age-related bone loss in men: the MINOS study. Bone 26:123–129 Khosla S, Melton 3rd LJ, Atkinson EJ, O’Fallon WM 2001 Relationship of serum sex steroid levels to longitudinal changes in bone density in young versus elderly men. J Clin Endocrinol Metab 86:3555–3561 Barrett-Connor E, Mueller JE, von Muhlen DG, Laughlin GA, Schneider DL, Sartoris DJ 2000 Low levels of estradiol are associated with vertebral fractures in older men, but not women: the Rancho Bernardo Study. J Clin Endocrinol Metab 85:219 –223 Colvard DS, Eriksen EF, Keeting PE, Wilson EM, Lubahn DB, French FS, Riggs BL, Spelsberg TC 1989 Identification of androgen receptors in normal human osteoblast-like cells. Proc Natl Acad Sci USA 86:854 – 857 Eriksen EF, Colvard DS, Berg NJ, Graham ML, Mann KG, Spelsberg TC, Riggs BL 1988 Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241:84 – 86 Oursler MJ, Osdoby P, Pyfferoen J, Riggs BL, Spelsberg TC 1991 Avian osteoclasts as estrogen target cells. Proc Natl Acad Sci USA 88:6613– 6617 Vanderschueren D, van Herck E, Nijs J, Ederveen AG, De Coster R, Bouillon R 1997 Aromatase inhibition impairs skeletal modeling and decreases bone mineral density in growing male rats. Endocrinology 138:2301–2307 Vanderschueren D, Van Herck E, De Coster R, Bouillon R 1996 Aromatization of androgens is important for skeletal maintenance of aged male rats. Calcif Tissue Int 59:179 –183 Oz OK, Zerwekh JE, Fisher C, Graves K, Nanu L, Millsaps R, Simpson ER 2000 Bone has a sexually dimorphic response to aromatase deficiency. J Bone Miner Res 15:507–514 Miyaura C, Toda K, Inada M, Ohshiba T, Matsumoto C, Okada T, Ito M, Shizuta Y, Ito A 2001 Sex- and age-related response to aromatase deficiency in bone. Biochem Biophys Res Commun 280:1062–1068 Vidal O, Lindberg MK, Hollberg K, Baylink DJ, Andersson G, Lubahn DB, Mohan S, Gustafsson JA, Ohlsson C 2000 Estrogen receptor specificity in the regulation of skeletal growth and maturation in male mice. Proc Natl Acad Sci USA 97:5474 –5479 Kousteni S, Bellido T, Plotkin LI, O’Brien CA, Bodenner DL, Han L, Han K, DiGregorio GB, Katzenellenbogen JA, Katzenellenbogen BS, Roberson PK, Weinstein RS, Jilka RL, Manolagas SC 2001 Nongenotropic, sex-nonspecific signaling through the estrogen or androgen receptors: dissociation from transcriptional activity. Cell 104:719 –730 Vanderschueren D, Van Herck E, Suiker AM, Visser WJ, Schot LP, Bouillon R 1992 Bone and mineral metabolism in aged male rats: short and long term effects of androgen deficiency. Endocrinology 130:2906 –2916 Lindberg MK, Moverare S, Skrtic S, Alatalo S, Halleen J, Mohan S, Gustafsson JA, Ohlsson C 2002 Two different pathways for the maintenance of trabecular bone in adult male mice. J Bone Miner Res 17:555–562 Taxel P, Kennedy D, Fall P, Willard A, Shoukri K, Clive J, Raisz LG 2000 The effect of short-term treatment with micronized estradiol on bone turnover and gonadotrophins in older men. Endocr Res 26:381–398 Taxel P, Kennedy DG, Fall PM, Willard AK, Clive JM, Raisz LG 2001 The effect of aromatase inhibition on sex steroids, gonadotropins, and markers of bone turnover in older men. J Clin Endocrinol Metab 86:2869 –2874 MacDonald PC, Madden JD, Brenner PF, Wilson JD, Siiteri PK 1979 Origin of estrogen in normal men and in women with testicular feminization. J Clin Endocrinol Metab 49:905–916