/00/$03.00/0 Vol. 141, No. 5

Transcription

1 /00/$03.00/0 Vol. 141, No. 5 Endocrinology Printed in U.S.A. Copyright 2000 by The Endocrine Society An Aged Rat Model of Partial Androgen Deficiency: Prevention of Both Loss of Bone and Lean Body Mass by Low-Dose Androgen Replacement* DIRK VANDERSCHUEREN, LIESBETH VANDENPUT, STEVEN BOONEN, ERIK VAN HERCK, JOHANNES V. SWINNEN, AND ROGER BOUILLON Laboratory for Experimental Medicine and Endocrinology (D.V., L.V., E.V.H., J.V.S., R.B.), Onderwijs en Navorsing, Gasthuisberg and Centre for Metabolic Bone Diseases (S.B.), Katholieke Universiteit Leuven, B-3000 Leuven, Belgium ABSTRACT The aim of this study was to evaluate the effects of different doses of androgen replacement, both on body composition and bone, in an aged male orchidectomized rat model. Testosterone was administered by 0.5, 1, and 2.5-cm sc SILASTIC implants (release of, respectively, 11.5, 23, and 55 g/day) to aged (12 months old, 550 g) male orchidectomized Wistar rats during a 15-week experimental period. T 0.5 only partially prevented decrease of ventral prostate and seminal vesicle weight, compared with an intact group that received an empty implant (Intact). The 1-cm implant (T 1) completely prevented decrease of both seminal vesicles and ventral prostate weight. The 2.5-cm implant (T 2.5) was clearly supraphysiological, as demonstrated by significant hypertrophy of both androgen-sensitive organs. Serum testosterone was lower in T 0.5 and T 1 ( ng/ml and ng/ml, respectively) and higher in T 2.5 ( ng/ml), compared with both Intact ( ng/ml) and the baseline group( ng/ml). As expected, orchidectomized rats that received an empty SILASTIC implant had significantly lower bone mineral content HYPOGONADAL MEN HAVE lower bone density (1), lower lean body mass (2), and higher bone turnover (3) than eugonadal men. In these men, androgen replacement increases lean body mass (2) and bone density (4, 5). However, in most human studies, androgen replacement is administered via im injection of testosterone enanthate, and these injections may lead (at least in the first days after the injection) to supraphysiological testosterone concentrations (6). Whether administration of lower doses of testosterone also protects against loss of bone and lean body mass remains to be clarified. This question is relevant because 20 30% of elderly men experience a relative deficiency of testosterone, compared with younger men (7), an observation commonly referred to as partial androgen deficiency of the aged male (PADAM) (8). In these elderly men, testosterone replacement Received October 15, Address all correspondence and requests for reprints to: D. Vanderschueren, Laboratory for Experimental Medicine and Endocrinology, Herestraat 49, B-3000 Leuven, Belgium. dirk.vanderschueren@ uz.kuleuven.ac.be. * The study was supported by the Flemish Fund for Scientific Research grant G and by a fellowship to Dr. Dirk Vanderschueren. Holder of the Merck Sharp & Dohme Chair in Metabolic Bone Diseases at the Katholieke Universiteit Leuven. Senior research assistant of the Fund for Scientific Research Flanders (Belgium). ( 7.9%), apparent density ( 5.7%), and lean body mass ( 10.8%), as measured by dual-energy x-ray absorptiometry, without significant changes in body weight and fat mass, compared with Intact. Also, cancellous ( 50.3%) and cortical ( 1.8%) volumetric density, as measured by peripheral quantitative computed tomography, were decreased in the tibia. Bone turnover, as measured by serum osteocalcin and urinary deoxypyridinoline excretion, was increased in orchidectomized rats that received an empty SILASTIC implant. T 0.5 prevented all changes, not only in bone mineral content, density, and turnover but also in lean body mass. Moreover, there were no significant differences, for all these parameters, between the different doses of testosterone replacement. In conclusion, low-dose androgen replacement does not lead to lower bone mineral density, higher bone turnover, and lower lean body mass in aged male rats, whereas complete androgen deficiency does. Therefore, the threshold concentration of testosterone necessary for prevention of both bone and lean body mass loss in aged male rats is clearly lower than for prostate and seminal vesicles. (Endocrinology 141: , 2000) should avoid supraphysiological concentrations of testosterone that may induce unwanted side effects such as polycythemia, sleep apnea syndrome, and (possibly) prostate hypertrophy (9). Studies in elderly men are also hampered by many concomitant diseases and medications that may affect both lean body mass, bone density, and androgen concentrations (10). The aim of this study was to evaluate the effects of different doses of androgen replacement on body composition and on bone in an aged male orchidectomized rat model. This model was characterized earlier as a representative model for bone studies of androgen replacement in hypogonadal men (11). Animals Materials and Methods Testosterone was administered by a 0.5-, 1-, or 2.5-cm sc SILASTIC brand implant (T 0.5, T 1, and T 2.5, respectively; Dow Corning, Midland, MI) in the cervical region to aged (12-month-old) male orchidectomized Wistar rats (n 10 for each group) during a 15-week experimental period. The effects of androgen replacement in these groups were compared with a group (n 10) of intact animals (Intact), orchidectomized animals (Orch, n 10) who received an empty SILASTIC implant, and a baseline (Base) group (n 10) that was killed at the beginning of the experimental period. All rats received a standard rat diet containing 0.4% calcium and 0.4% phosphate (Hope Farms, Woerden, The Neth- 1642

2 PARTIAL ANDROGEN DEFICIENCY 1643 erlands) and had free access to tap water. Animals were weighed regularly, and urine was collected the week before death, for measurement of deoxypyridinoline excretion, as described (12). Also before being killed, animals were anesthesized using Nembutal ip (60 mg/kg) (Sanofi Pharmaceuticals, Inc. Sante Animal, Brussels, Belgium) for in vivo measurement of lean body mass, bone mineral content (BMC), and density and fat mass, by dual-energy x-ray absorptiometry (DXA). After 15 weeks, the animals were killed under Nembutal anesthesia. Serum was collected from the abdominal aorta and stored at 20 C until measurement of testosterone, leptin, insulin-like growth factor I (IGF-I), cholesterol, and triglycerides. After sacrifice, ventral prostate and seminal vesicles were weighed. The left tibia was used for measurement of volumetric cancellous and cortical density and geometrics by peripheral quantitative computed tomography (pqct) ex vivo. The experiment was conducted after obtaining formal approval by the ethical committee of the Katholieke Universiteit Leuven. SILASTICS The in vitro release of testosterone from the SILASTIC tubes was calculated as follows: 1) Testosterone depots were prepared from polydimethylsilicone tubing (Silclear tubing, Degania Tubing, Degania, Israel, mm) of defined length (1 cm and 2 cm), which was filled with crystalline testosterone (Sigma, St. Louis, MO), sealed at both ends (Dow Corning Corp. Adhesive, type A, Dow Corning Corp.), and kept at room temperature for 24 h; 2) In a preliminary experiment, the in vitro release of the SILASTIC tubes was determined as follows: The SILASTIC tubes were incubated in a 50-ml solution containing 0.9% NaCl, 0.1% sodiumazide, 3% BSA,and 10 mm phosphate buffer, ph 7. Samples were removed at regular intervals and radioimmunoassayed for testosterone during a 15-day period. A linear release of testosterone in the buffer in function of time (r 0.98 for 1 cm and r 0.97 for 2 cm) was obtained. The SILASTIC release during 24 h in vitro was calculated to be 23 g for 1 cm and 44 g for 2 cm. Therefore, the daily dose of testosterone administered in vivo was 11.5 g in the T 0.5 group, 23 g inthe T 1 group, and 55 g in the T 2.5 group. Similar studies have demonstrated that implantation of SILASTIC tubes in orchidectomized rats clearly results in constant testosterone concentrations, both in short-term (13) and long-term (11) experiments. Assays Serum testosterone was measured by a commercial RIA kit (Biosource Technologies, Inc. Europe SA, Nivelles, Belgium). Intra and interassay variations were 4.7 and 8.1%, respectively. Serum leptin was measured by enzyme-linked immunosorbent assay (Assay Designs Inc., Ann Arbor, MI). Intra and interassay variations were 4.8 and 3.5%, respectively. Serum osteocalcin and IGF-I were measured by in-house RIA, as described (14, 15). Urinary deoxypyridinoline (DPD) excretion was measured by HPLC, as described previously (12). Serum cholesterol and triglycerides were measured on a Roche Molecular Biochemicals (Mannheim, Germany) Hitachi 747 automated chemical analyzer. DXA of the whole skeleton, femur, and lumbar vertebrae Total BMC, area, density, lean body mass, and fat mass were measured in vivo by QDR 4500A fan beam x-ray bone densitometer (Hologic, Inc.,Waltham, MA) using a specific software program from the manufacturer for rats weighing g. The BMC and apparent bone mineral density (BMD) of the whole body skeleton (software version v8.19a), lumbar spine (scanned at the levels of the vertebrae L 1 -L 2 -L 3 -L 4 ; regional high resolution, software version v8.16h), and femur (regional high resolution, software version v8.16h) were determined. pqct of the left proximal tibia The defatted tibia was placed in a specially constructed vial containing 70% ethanol and mounted in the Stratec XCT 960A densitometer (Norland Medical Systems, Inc., Fort Atkinson, WI) with a special support. Two CT-slices of each 1 mm of thickness were obtained using a voxel size of mm. Volumetric cancellous bone density was calculated on a slice at 5 mm proximal from the knee joint. Volumetric cortical bone density was measured on a second slice 15 mm from the knee joint. Volumetric cancellous bone density has been calculated using contmode 1, peel-mode 20, and threshold 530, whereas, for cortical bone, contmode 1 and threshold 930 were used. Endocortical and periosteal perimeter represent, respectively, the length of the inner and outer perimeters of the second slice and are expressed in mm. We have previously found that the coefficient of variation of these measurements is 4.2% for trabecular bone and 0.72% for cortical bone. Statistical analysis Data analysis was performed with a software program (NCSS 2000, Kaysville, UT). The results are presented as mean sem. One-way ANOVA was carried out to detect overall differences and if P was less than 0.05, was followed by Fisher s least-significant-difference multiplecomparison test to calculate intergroup differences vs. Intact, Orch, and Base. The Pearson product-moment test was used for the measurement of the correlations. Results Serum testosterone concentrations and ventral prostate and seminal vesicle wet weight As expected, serum testosterone (Fig. 1a) ventral prostate weight (Fig. 1b) and seminal vesicles weight (Fig. 1c) decreased in the Orch group. Serum testosterone was lower in the T 0.5 and T 1 groups and higher in the T 2.5 group, compared with Intact. T 0.5 only partially prevented this decrease of ventral prostate and seminal vesicle weight ( 45 and 52%, respectively), compared with Intact. The 1-cm implant completely prevented the decrease of both seminal vesicle and ventral prostate weight. The 2.5-cm implant was clearly supraphysiological, as demonstrated by significant hypertrophy of both androgen-sensitive organs, compared with Intact. Ventral prostate weight and seminal vesicle weight were significantly higher in T 0.5, T 1, and T 2.5, compared with Orch. Biochemical parameters Serum osteocalcin and urinary DPD increased in the Orch group (Table 1). The T 0.5 group had similar, and thet1and T 2.5 groups had lower osteocalcin concentrations than Intact. All three doses of testosterone replacement, however, significantly reduced serum osteocalcin concentrations, compared with Orch. Also, DPD excretions in the groups with different doses of androgen replacement were significantly lower than in Orch. Urinary DPD in the T 0.5 group, however, remained significantly higher than in Intact. No significant changes in serum lipids, IGF-I, or leptin were observed. As expected, serum leptin was significantly correlated with fat mass (r 0.68, P 0.001) and serum triglycerides (r 0.29, P 0.05). BMC and density of the whole skeleton, as measured by DXA Orch rats had significantly lower BMC (Fig. 2a) and BMD (Fig. 2b), compared with both Intact and Base. Testosterone administration to orchidectomized animals prevented this decrease of BMC and BMD. There were no significant differences between the groups with different doses of androgen replacement, with respect to BMC and BMD.

3 1644 PARTIAL ANDROGEN DEFICIENCY Endo 2000 Vol 141 No 5 Only the T 2.5 group had greater endocortical and periosteal perimeters without gain in cortical thickness, compared with Intact. Cancellous density was lower in Intact, compared with Base. Both cortical and cancellous density of the left tibia were significantly decreased in Orch, compared with Intact. Again, even the lowest dose of testosterone (T 0.5) corrected the density to Intact levels. Administration of higher doses of testosterone did not increase cancellous bone density. Body weight, lean body mass, and fat mass, as measured by DXA Body weight (Table 1) and fat mass (Fig. 2c) were not significantly different between the groups, except for a lower fat mass at baseline. Orch rats had lower lean body mass (Fig. 2d). Lean body mass increased after androgen replacement and was not different between the T 0.5, T 1, and T 2.5 groups and the Intact group. Moreover, lean body mass and BMC were significantly correlated (r 0.54, P ). FIG. 1. Serum testosterone concentration (1a), ventral prostate weight (1b), and seminal vesicles weight (1c) in the Base group, the Intact group, the orchidectomized group without testosterone replacement (Orch), and the orchidectomized groups treated with testosterone (T 0.5, T 1, and T 2.5). Statistical analysis: one-way ANOVA was carried out to detect overall differences and, if P was less than 0.05, was followed by Fisher s least significant difference (LSD) multiple-comparison test., P 0.05, compared with Base; *, P 0.05, compared with Intact;, P 0.05, compared with Orch; NA, not available. BMC and density of the distal lumbar vertebrae and femur, as measured by DXA Again, both BMC and BMD were decreased in Orch rats (Table 2), and androgen replacement in orchidectomized animals prevented bone loss. Differences between the groups with different doses of androgen replacement were not significant. Bone structure, and cancellous and cortical density of the left tibia, as measured by pqct The Intact group had thinner cortices, compared with Base, because of a greater endocortical perimeter (Table 2). Discussion The aged orchidectomized rat was used as an animal model to address the critically important question of the threshold serum concentration of testosterone needed to maintain BMC, density, and lean body mass in aging men. According to earlier work, this model reflects very well the skeletal effects of androgen deficiency and replacement in hypogonadal men (11). However, the skeletal effects of lower- and higher-than physiological doses of testosterone were not yet evaluated in this model. Therefore, in this study, we evaluated both skeletal effects and effects on lean body mass of different doses of testosterone in the aged orchidectomized rat. Assuming that a daily production of 5000 g in adult men (16) would correspond with a daily production of about 30 g in a 500-g rat, the 1-cm SILASTIC implant (23 g/day) would probably correspond with the low physiological range in men. Indeed, serum testosterone concentrations in this group were lower than in Intact. However, seminal vesicle wet weight and ventral prostate weight were not decreased, compared with Intact, suggesting that low serum testosterone concentrations are still sufficient to maintain intact ventral prostate and seminal vesicles weight. Serum testosterone concentrations in the group that received the 0.5-cm implant were only 25% of intact levels. This dose of androgen replacement was clearly subphysiological, as indicated by ventral prostate weight and seminal vesicle wet weight that were, respectively, 45 and 52% lower than in Intact. Serum testosterone concentrations in the T 2, 5 group, finally, were supraphysiological, as evidenced by significant hypertrophy of both androgen-sensitive organs. The 1-cm testosterone implant prevented both the increase of bone turnover and the decrease of bone mass and lean body mass. However, contrary to our expectations, low-dose androgen replacement also prevented both the increase of bone turnover and the decrease of bone mass and lean body mass. Therefore, threshold serum concentrations of testosterone of only 25% of Intact are already sufficient to prevent the loss of bone

4 PARTIAL ANDROGEN DEFICIENCY 1645 TABLE 1. General data Base Intact Orch T 0.5 T 1 T 2.5 P value (ANOVA) Body weight (g) NS Serum IGF-I (ng/ml) NS Serum total cholesterol (mg/dl) NS Serum HDL cholesterol (mg/dl) NS Serum triglyceride (mg/dl) NS Serum leptin (ng/ml) NS Serum osteocalcin (ng/ml) a,b 73 2 c 60 4 b,c 58 3 a,b,c Urinary deoxypyridinoline ( M/mM creat) a,b 36 2 b,c 31 2 c 26 2 c Statistical analysis: one-way ANOVA was carried out to detect overall differences, and if P 0.05 was followed by Fisher LSD multiplecomparison test. Date are reported as mean SEM for baseline (Base) and, at the end of the experimental period, in control rats (Intact), orchidectomized rats (Orch), and orchidectomized rats treated with testosterone (T 0.5, T 1, and T 2.5). Baseline data are indicated in italics. HDL, High-density lipoprotein; NS, not significant. a P 0.05, compared with Base. b P 0.05, compared with Intact. c P 0.05, compared with Orch. FIG. 2. BMC (2a), BMD (2b), fat mass (2c), and lean body mass (2d) in the Base group, the Intact group, the orchidectomized group without testosterone replacement (Orch), and the orchidectomized groups treated with testosterone (T 0.5, T 1, and T 2.5), as measured by DXA. Statistical analysis: one-way ANOVA was carried out to detect overall differences and, if P was less than 0.05, was followed by Fisher s LSD multiple-comparison test., P 0.05, compared with Base; *, P 0.05, compared with Intact;, P 0.05, compared with Orch. and lean body mass in this model. Serum osteocalcin concentrations in the rats that received 1-cm and 2.5-cm testosterone implants were even lower than in Intact, suggesting that the major effect of testosterone is to reduce bone resorption, rather than to stimulate bone formation. Overall, the effects of supraphysiological administration of testosterone were not more beneficial than the effects of physiological administration on either lean body mass or bone mass. However, the increase of periosteal perimeter, compensated by an increase of endosteal perimeter, suggests that testosterone in supraphysiological concentration may also stimulate periosteal bone formation. Translation of these animal data to humans would have important implications for androgen deficiency and replacement: First, PADAM, (which corresponds to the T 0.5 implant in this study) should be regarded as a distinct condition from overt hypogonadism (which corresponds to the Orch group in this study). The effects of PADAM on androgen target organs, such as bone and lean body mass, may be not as important as the well-established detrimental effects of overt hypogonadism (1, 2). Moreover, it follows that whether the beneficial effects of androgen replacement observed in hypogonadal men also apply to men suffering from PADAM can only be determined by properly designed randomized

5 1646 PARTIAL ANDROGEN DEFICIENCY Endo 2000 Vol 141 No 5 TABLE 2. Bone mineral content and density of the distal lumbar vertebrae and femur as measured by DXA; bone structure and bone densities of the tibia as measured by pqct P value (ANOVA) Base Intact Orch T 0.5 T 1 T 2.5 Distal lumbar vertebrae BMC (g) a,b c c c 0.05 BMD (g/cm 2 ) a,b c c c 0.01 Total femur BMC (g) a c c c 0.05 BMD (g/cm 2 ) a,b c c c 0.05 Tibia Periosteal perimeter (mm) c a,b,c 0.01 Endocortical perimeter (mm) a a,b,c 0.01 Cortical thickness (mm) a a a a a 0.05 Cancellous volumetric density (g/cm 3 ) a a,b a,c a,c a,c 0.01 Cortical volumetric density (g/cm 3 ) a a,b a,c c a 0.01 Statistical analysis: one-way ANOVA was carried out to detect overall differences, and if P 0.05 was followed by Fisher LSD multiple-comparison test. Data are reported as mean SEM for baseline (Base) and, at the end of the experimental period, in control rats (Intact), orchidectomized rats (Orch), and orchidectomized rats treated with testosterone (T 0.5, T 1, and T 2.5). Baseline data are indicated in italics. a P 0.05, compared with Base. b P 0.05, compared with Intact. c P 0.05, compared with Orch. clinical trials in PADAM-patients. Thusfar, only a few shortterm studies, including only a small number of elderly men, were reported, without detailed assessment of the effects of androgen replacement on bone (17 19). Similarly, the clinical impact of androgen replacement on lean body mass and body composition in partially androgen-deficient men remains to be clarified (8). In agreement with our data, a recent randomized controlled study in elderly men with only partial androgen deficiency could find no overall significant beneficial effect of androgen replacement on bone density and turnover (20). Only those men with the lowest testosterone concentrations seemed to benefit from androgen replacement. Second, supraphysiological replacement in humans may lead to changes in serum lipids or induce unwanted side effects, such as prostate hypertrophy, polycythemia, or sleep apnea syndrome (9). In this study, supraphysiological replacement (which corresponds to the T 2.5 implant) had no effects on serum lipids or high-density lipoprotein cholesterol, which is in contrast with some studies in men (19, 21). The modest muscle hypertrophy observed in healthy exercising eugonadal men receiving supraphysiological doses of testosterone (22) may not be relevant for either our animal model or the elderly male. Despite these side effects, there was no real benefit on bone or lean body mass, which suggests that there is no indication for supraphysiological androgen replacement in hypogonadal men. Third, a very close correlation between responses in bone and lean body mass to testosterone replacement are also in line with earlier data published in hypogonadal men (5). This may indicate that the positive effect of testosterone therapy on bone depends, in part, on its effects on muscle. Whether the effects of testosterone depend on stimulation of androgen receptors, or estrogen receptors after aromatization of testosterone in bone or muscle, is unknown. It is, however, unlikely that the effects of testosterone are related to changes in serum IGF-I concentrations (which were not different between groups in this study). In conclusion, continuous administration of a low dose of testosterone, a dose that was clearly not able to maintain prostate and seminal vesicles weight, prevented loss of both bone and lean body mass in an aged male orchidectomized animal model. Therefore, it would be interesting to investigate whether a similar low-dose replacement of testosterone would have similar effects in hypogonadal men. Acknowledgments The authors thank Herman Borghs, Jos Nijs, and Herman Peeters for expert technical assistance. References 1. Finkelstein JS 1998 Androgens and bone metabolism. In: Nieschlag E, Behre HM (eds) Testosterone. Action-Deficiency-Substitution. Springer-Verlag, Berlin-Heidelberg-New York, pp Brodsky IG, Balagopal P, Nair KS 1996 Effects of testosterone replacement on muscle mass and muscle protein synthesis in hypogonadal men: a Clinical Research Center study. J Clin Endocrinol Metab 81: Stepan JJ, Lachman M, Zverina J, Pacovsky V, Baylink DJ 1989 Castrated men exhibit bone loss: effect of calcitonin treatment on biochemical indices of bone remodeling. J Clin Endocrinol Metab 69: Behre HM, Kliesch S, Leifke E, Link TM, Nieschlag E 1997 Long-term effect

6 PARTIAL ANDROGEN DEFICIENCY 1647 of testosterone therapy on bone mineral density in hypogonadal men. J Clin Endocrinol Metab 82: Katznelson L, Finkelstein JS, Schoenfeld DA, Rosenthal DI, Anderson EJ, Klibanski A 1996 Increase in bone density and lean body mass during testosterone administration in men with acquired hypogonadism. J Clin Endocrinol Metab 81: Snyder PJ, Lawrence DA 1980 Treatment of male hypogonadism with testosterone enanthate. J Clin Endocrinol Metab 51: Vermeulen A 1991 Androgens in the aging male - clinical review 24. J Clin Endocrinol Metab 73: Kaufman JM, Vermeulen A 1998 Androgens in male senescence. In: Nieschlag E, Behre HM (eds) Testosterone. Action-Deficiency-Substitution. Springer- Verlag, Berlin-Heidelberg-New York, pp Nieschlag E, Behre HM 1998 Pharmacology and clinical uses of testosterone. In: Nieschlag E, Behre HM (eds) Testosterone. Action-Deficiency-Substitution. Springer-Verlag, Berlin-Heidelberg-New York, pp Tenover JS 1992 Effect of testosterone supplementation in the aging male. J Clin Endocrinol Metab 75: Vanderschueren D, Van Herck E, Suiker AMH, Visser WJ, Schot LPC, Bouillon R 1992 Bone and mineral metabolism in aged male rats: short- and long-term effects of androgen deficiency. Endocrinology 130: Vanderschueren D, Jans I, Van Herck E, Moermans K, Verhaeghe J, Bouillon R 1994 Time-related increase of biochemical markers of bone turnover in androgen-deficient male rats. Bone Miner 26: Marchetti B, Poulin R, Plante M, Labrie F 1988 Castration levels of plasma testosterone have potent stimulatory effects on androgen-sensitive parameters in the rat prostate. J Steroid Biochem 31: Verhaeghe J, Van Herck E, Van Bree R, Van Assche FA, Bouillon R 1989 Osteocalcin during the reproductive cycle in normal and diabetic rats. J Endocrinol 120: Verhaeghe J, Van Herck E, Visser WJ, Suiker AMH, Thomasset M, Einhorn TA, Faierman E, Bouillon R 1990 Bone and mineral metabolism in BB rats with long-term diabetes. Decreased bone turnover and osteoporosis. Diabetes 39: Mooradian AD, Morley JE, Korenman SG 1987 Biological actions of androgens. Endocr Rev 8: Urban RJ, Bodenburg Y, Gilkison C, Foxworth J, Coggan AR, Wolfe RR, Ferrando A 1995 Testosterone administration to elderly men increases skeletal muscle strength and protein synthesis. Am J Physiol 269:E820 E Morley JE, Perry HM, Kaiser FE, Kraenzle D, Jensen J, Houston K, Mattamel M 1993 Effects of testosterone replacement therapy in old hypogonadal males: a preliminary study. J Am Geriatr Soc 41: Sih R, Morley JE, Kaiser FE, Perry III HM, Patrick P, Ross C 1997 Testosterone replacement in older hypogonadal men: a 12-month randomized controlled trial. J Clin Endocrinol Metab 82: Snyder PJ, Peachey H, Hannoush P, Berlin JA, Loh L, Holmes JH, Dlewati A, Staley J, Santanna J, Kapoor SC, Attie MF, Haddad Jr JG, Strom BL 1999 Effect of testosterone treatment on bone mineral density in men over 65 years of age. J Clin Endocrinol Metab 84: von Eckardstein A 1998 Androgens, cardiovascular risk factors and atherosclerosis. In: Nieschlag E, Behre HM (eds) Androgens. Action-Deficiency- Substitution. Springer-Verlag, Berlin-Heidelberg-New York, pp Bhasin S, Storer TW, Berman N, Callegari C, Clevenger BA, Philips J, Bunnell T, Tricker R, Shirazi A, Casaburi R 1996 The effects of supraphysiologic doses of testosterone on muscle size and strength in men. N Engl J Med 335:1 7