Risks of Testosterone Replacement TherapyBerlin, Louise Loh, John H. As men age, their serum testosterone concentrations decrease, as do their bone densities. Because bone density is also low in hypogonadal testosterone therapy and bone density, we hypothesized that increasing the serum testosterone concentrations of men over 65 yr to those found therayp young men would increase their bone densities. We randomized men over 65 yr of age to wear either a testosterone patch or a placebo patch double blindly for 36 months. We measured bone mineral density by dual energy x-ray absorptiometry before and therxpy treatment. Ninety-six men completed the entire month protocol.
Testosterone Replacement Therapy and Bone Mineral Density in Men with Hypogonadism
In both men and women, a decrease in bone mineral density BMD is a major symptom of hypogonadism. Although the effects of estrogens on osteoporosis in women are well documented, comparatively little is known about the effects of long term testosterone substitution on BMD in hypogonadal men. Therefore, we studied BMD in 72 hypogonadal patients 37 men with primary and 35 men with secondary hypogonadism under testosterone substitution therapy that continued for up to 16 yr.
Thirty-two of these men were also seen before initiation of therapy. At annual intervals, trabecular BMD of the lumbar spine was measured by quantitative computed tomography, a true volumetric and reproducible method for long term serial BMD measurements. Serum levels of testosterone increased to the normal range in all androgen-treated hypogonadal men.
The most significant increase in BMD was seen during the first year of testosterone treatment in previously untreated patients, when BMD increased from Long term testosterone treatment maintained BMD in the age-dependent reference range in all 72 hypogonadal men, independent of the type of hypogonadism. Transdermal testosterone patches applied to the scrotum were as effective in normalizing BMD as im testosterone enanthate injections. In summary, testosterone therapy increases BMD in hypogonadal men regardless of age.
The greatest increase is seen during the first year of treatment in previously untreated patients with low initial BMD. In hypogonadal men, BMD can be normalized and maintained in the normal range by continuous, long term testosterone substitution.
Case-controlled studies have demonstrated that in hypogonadal men this reduced BMD is associated with a significant increase in bone fractures 3.
Along with lower testosterone levels a clear decrease in BMD 6 , 7 and an increase in hip and spine fractures 8 have been demonstrated in aging males, similar to the well-described changes in postmenopausal women. Large multicenter clinical trials have shown that estrogen replacement therapy prevents loss of BMD and decreases the incidence of bone fractures, which in untreated women are caused by the decrease in endogenous estrogens after menopause 9 — As comparatively little is known about the beneficial effects of long term testosterone substitution therapy on bone in men 2 , we investigated BMD changes in hypogonadal men treated with testosterone preparations for up to 16 yr.
In addition, we compared the effects of standard im testosterone injection therapy to those of modern transdermal testosterone application. Adult male hypogonadal patients, aged 18—74 yr, who received effective androgen substitution therapy and underwent serial measurements of lumbar spine BMD by quantitative computed tomography QCT after giving informed consent were included in the evaluation. Patients with constitutional delay of puberty, other endocrine diseases causing lowered BMD e.
In 32 of these patients initial QCT measurement was carried out before initiation of androgen substitution therapy. The other 40 patients had received effective androgen therapy for 3. Hypogonadism was treated by exogenous testosterone or, in seven patients with secondary hypogonadism, with hCG or pulsatile GnRH.
Testosterone preparations were administered im, transdermally, or orally. Intramuscular substitution therapy was applied to 52 patients by mg testosterone enanthate Testoviron Depot , Schering, Berlin, Germany , injected in most cases every 3 weeks range, 2—4 weeks. Morning blood samples for testosterone measurements were drawn between — h.
To test for effective testosterone substitution, blood was taken at time points indicating average testosterone serum levels during substitution therapy In patients treated with transdermal testosterone patch or oral testosterone undecanoate, blood was taken 3—6 h after administration 12 , Serum testosterone levels were measured before initiation of therapy and then at yearly intervals. Testosterone levels before initiation of therapy, after 1 yr of treatment, and at the last QCT measurement of all individual patients were included in the statistical analysis.
Serum testosterone levels were measured by RIA as described previously The detection limit for testosterone was 0. The intra- and interassay coefficients of variation were 5. The intra- and interassay coefficients of variation for estradiol were 6. A calibration phantom was used in all studies, initially a liquid Cann-Genant-calibration standard 15 January to March and afterward a solid calibration standard Image Analysis, Columbia, KS; since April To achieve comparable results, data were adjusted by cross-calibration analysis as previously described Ten-millimeter thick midvertebral slices of L2, L3, and L4 were obtained with the gantry angled parallel to the vertebral end plate using a low dose technique.
Trabecular BMD was measured in an oval and a round region of interest, respectively. Single energy QCT was used to improve the long term precision of BMD measurements and to reduce total radiation exposure of the patients during longitudinal follow-up The known bone marrow fat error of single energy compared to dual energy QCT has been shown to be of limited impact on follow-up BMD measurements 17 , A total of BMD measurements 32 before testosterone therapy and during therapy were performed during the study period in 72 hypogonadal patients.
Age-adjusted reference ranges for trabecular BMD 19 and reference ranges for fracture risk 20 , 21 were applied for estimation of the therapeutic efficacy of testosterone substitution therapy. Comparisons between BMD and testosterone, respectively, at the first examination, after 1 yr of therapy, and at the last examination were performed by ANOVA for repeated measures.
Multiple linear regression and partial correlation analysis were applied for evaluation of the association of testosterone, estradiol, age, and body mass index with BMD and increase in BMD after therapy, respectively. Independent data at the first QCT measurement in the individual patients were used for the analysis.
To adjust for different age, etiology of hypogonadism, different androgen pretreatment, and duration of testosterone substitution therapy in the individual patients, these variables were included in the ANOVA as covariates. When necessary, analysis was performed on logarithmically or, for percentage data, arcsine-transformed data.
Computations were performed using the statistical software package SPSS, version 6. Of the 32 patients whose initial QCT measurement preceded initiation of androgen therapy, the 15 patients with secondary hypogonadism had lower pretreatment serum levels of testosterone 3.
Effective testosterone substitution therapy resulted in a uniform increase in BMD from In patients who had been effectively treated for at least 1 yr before their first QCT measurement, BMD increased significantly from However, this increase of Increase in BMD during long term testosterone substitution therapy up to 16 yr in 72 hypogonadal patients.
Circles indicate hypogonadal patients with first QCT measurement before initiation of testosterone substitution therapy, squares show those patients already receiving testosterone therapy at the first QCT.
The dark shaded area indicates the range of high fracture risk, the unshaded area shows the range without significant fracture risk, and the light shaded area indicates the intermediate range where fractures may occur 20, This indicates that, on the average, long term androgen therapy maintains the BMD achieved during initial therapy, but does not further significantly increase BMD. Circles indicate hypogonadal patients with first QCT measurement before initiation of testosterone substitution therapy; squares show those patients already receiving testosterone therapy at the first QCT.
In all individual hypogonadal patients, effective testosterone substitution therapy of longer than 3-yr duration increased BMD to the previously described age-dependent reference range BMD, age, duration of therapy, sex hormones, and BMI in patients treated with im or transdermal testosterone.
Previous studies in hypogonadal men, which in most publications included only a few patients, showed discrepant effects of testosterone therapy on BMD.
No change in trabecular BMD was seen in the older patients, aged 24—52 yr, with closed epiphyses even after treatment of up to 31 months. In 8 patients with hyperprolactinemic hypogonadism, normalization of testosterone due to different therapies only caused an increase in cortical BMD, whereas no change in trabecular BMD was detected during an observation period of 6—48 months Treatment of 6 hypogonadal patients with mg of a testosterone ester mixture Sustanon every 3 weeks for 24 months increased trabecular BMD of the spine by Treatment of 14 hypogonadal men with mg Sustanon every month caused an increase in cortical BMD of 5.
In a recent study treatment with sublingual testosterone cyclodextrin at a dose of 5 mg, 3 times daily, in 34 hypogonadal men for 6 months did not change the BMD of total body, hip, or spine when measured by dual energy x-ray absorptiometry In part, the varying effects of androgen replacement on BMD in hypogonadal men seen in previous studies can be explained by different techniques of BMD measurements, different durations of treatment, and different testosterone preparations applied.
In our study we included only patients receiving long term testosterone substitution therapy for at least 1 yr who were treated effectively with established and well described androgen substitution therapies 12 , In addition, we applied a true volumetric method for BMD measurement at the same site in all patients and used the single energy QCT technique to improve long term precision during serial measurements 15 — In the present evaluation we demonstrated that decreased BMD due to hypogonadism can be restored to the age-dependent reference range by effective, long term androgen substitution therapy.
The largest increase was seen in patients with initial low BMD during the first year of treatment. Our results are similar to those obtained in postmenopausal women receiving estrogen replacement therapy, where therapy is most effective during the first year of treatment, and the magnitude of the BMD increase is greatest in those women with low initial BMD In our study we did not assess the incidence of fractures in hypogonadal patients.
However, it was shown previously that testosterone deficiency is an important risk factor for fractures in men 3 , 21 , 30 , and in women it has been clearly demonstrated that long term, continuous estrogen hormone replacement therapy that increases BMD reduces the individual risk of bone fractures 9. In contrast to previous studies we demonstrated that effective androgen therapy increases BMD in hypogonadal patients independent of age.
Another patient with idiopathic hypogonadotropic hypogonadism was first diagnosed at the age of 61 yr. Statistical analysis in all 72 patients of our study revealed that testosterone therapy is an age-independent, highly significant factor influencing BMD.
Hypogonadal patients with reduced BMD will benefit from long term, effective androgen substitution therapy, and thus, age per se should not preclude patients from appropriate treatment. It might be argued that relatively high or even supraphysiological serum testosterone levels are necessary to increase BMD in hypogonadal patients. Most previous clinical studies used im testosterone preparations that result in supraphysiological testosterone levels shortly after injection, as demonstrated by detailed pharmacokinetic analysis For this reason we compared the efficacy of im testosterone with transdermal scrotal testosterone systems that lead to physiological serum levels of testosterone in the lower normal range and even mimic normal diurnal variation Statistical analysis revealed no significant difference between both testosterone treatments regarding the BMD achieved.
Thus, as is the case for other biological effects 31 — 33 , for treatment of reduced BMD, restoration of testosterone to the normal physiological range seems to be sufficient without the need for pharmacologically high testosterone levels. The exact mechanism by which androgens affect BMD is not yet fully elucidated In vitro studies and clinical trials have shown that nonaromatizable androgens increase bone formation 2 , In addition, it has been demonstrated in vitro 10 and in male patients with estrogen receptor mutations 36 or aromatase deficiency 37 that estrogens also play an important role in bone mineralization in men.
Recently, it was demonstrated by histomorphometry that testosterone treatment of female to male transsexuals after bilateral ovariectomy maintains BMD Available data indicate that both androgens and estrogens have important direct effects on bone. It is well accepted that testosterone substitution therapy should be performed with androgen preparations releasing natural testosterone into the general circulation, which can be converted to the active metabolites dihydrotestosterone and estradiol In our evaluation we cannot discriminate between pure androgenic and estrogenic effects of testosterone therapy on BMD.
However, statistical analysis revealed that serum levels of testosterone show the strongest correlation with BMD, and that serum estradiol that is converted from testosterone has no further testosterone-independent association. In summary, our study demonstrates that long term, effective androgen therapy significantly increases BMD in hypogonadal men.
Intramuscular testosterone injections and transdermal testosterone are equally effective. Independent of age, the highest increase in BMD is seen during the first year of treatment and in those patients with initial low values of BMD.
Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Close mobile search navigation Article navigation. View large Download slide. Testosterone deficiency as a risk factor for hip fractures in men: Age, disease, and changing sex hormone levels in middle-aged men: