Low bone density and high percentage of body fat among men who were treated with androgen deprivation therapy for prostate carcinoma




Men with prostate carcinoma who are treated with androgen deprivation therapy (ADT) are reported to be at an increased risk of bone loss and weight changes due to the sudden disruption of hormonal levels. In the current case–control study, the authors examined the prevalence and magnitude of low bone density and obesity among men with prostate carcinoma who were treated with ADT.


Sixty-two men with prostate carcinoma who had been receiving ADT for 1–5 years were included as cases. Healthy men (n = 47) with a prostate specific antigen level < 4.0 ng/mL were recruited as controls. Body composition and bone mineral density (BMD) were measured using dual-energy X-ray absorptiometry. The average age was 74.3 years for the cases and 72.8 years for the controls.


The results of the current study demonstrate that prostate carcinoma cases had significantly higher body weight (86.5 kg vs. 80.6 kg), a higher percentage of body fat (30% vs. 26%), and a lower total body BMD (1.12 mg/cm2 vs. 1.17mg/ cm2) compared with controls (P < 0.05). Cases were more likely to be obese (27.4% vs 43%) and have low BMD at trochanter (32.3% vs. 10.6%), intertrochanter (48.4% vs. 29.8%), and total hip measurements (50.0% vs. 25.3%).


The results of the current study indicate that men with prostate carcinoma who are treated with ADT have a significantly increased risk of low bone density and obesity. Cancer 2002;95:2136–44. © 2002 American Cancer Society.

DOI 10.1002/cncr.10967

Prostate carcinoma is the most common malignancy occurring among American men. The prevalence based on 0–22 cumulative years from the time of diagnosis was reported to be 1084/10,000 for white men and 886/10,000 for black men in 1997.1 With advances in the early detection and treatment of prostate carcinoma, the 5-year relative survival rate (adjusted for other causes of death) has increased from a low of 64.0% for men newly diagnosed in 1973 to 92.9% for men newly diagnosed in 1990.2 Even within the category of advanced carcinoma, men with prostate carcinoma appear to have a better absolute cumulative 5-year survival rate (20.8%) than people with breast carcinoma (17.0%), colorectal carcinoma (5.3%), or lung carcinoma (1.5%).3 Androgen deprivation therapy (ADT) has been widely used for the treatment of patients with prostate carcinoma. Although ADT is effective for controlling prostate carcinoma, hypogonadism resulting from ADT is associated with adverse effects such as impotence, hot flashes, cardiovascular morbidity, and, possibly, osteoporosis. To improve quality of life among prostate carcinoma survivors, the adverse effects associated with ADT need to be addressed.

Osteoporosis is not a disease that affects only women; it also is a growing health problem among men. In the U.S., approximately 1.5 million men age > 65 years have osteoporosis, and another 3.5 million men have osteopenia (low bone density), thereby increasing their risk of fractures.4 Hip fractures in older men cause higher morbidity and mortality than those occurring in women.5 Decreasing levels of sex hormones is one of the major risk factors for osteoporosis among men,6, 7 and ADT is a known cause of sudden reduction in sex hormones among men.

It is well known that low bone mineral density (BMD) is the most important determinant of fracture risk in women. Limited data have demonstrated that men with prostate carcinoma may have preexisting low BMD even before ADT,8–11 and ADT causes a remarkable additional decrease in BMD.9, 10, 12–14 The ADT-induced bone loss further worsens the already low bone density status, and may contribute to an increase in the risk of fractures among the ADT-treated patients.15–19 Although loss of BMD after ADT has long been a concern, to our knowledge the majority of the previous BMD studies among men with prostate carcinoma who were treated with ADT included very few participants and did not exclude men with bone metastases, a complication that makes BMD measurements unreliable. Thus, the magnitude and prevalence of low BMD among this high-risk population largely are unknown.

In addition to loss of skeletal mass, another potential consequence of ADT is the change in the body's soft tissue composition. Body composition is related to the risks of many chronic diseases including cardiovascular disease and diabetes.20–23 Reduced muscle mass and increased percent body fat often result from low androgen levels in men.24–29 Decreased muscle mass may lead to reduced muscle strength, which also contributes to risk of bone fracture in men. Although the adverse effects of hypogonadism on body composition have been recognized increasingly,25 to our knowledge research concerning body composition among men with prostate carcinoma after ADT rarely is discussed in the literature. It is likely that ADT-induced hypogonadism, prostate carcinoma prognosis, and changes in lifestyle after a cancer diagnosis could have a combined impact on body composition among ADT-treated patients. Documenting the status of body composition may help in forming effective prevention approaches to reduce the risk of other diseases among men with prostate carcinoma after ADT.

In the current study we measured BMD and body composition among men with prostate carcinoma who were treated with either surgical or chemical ADT and had no radiographic evidence of bone metastases. The major objective of the study was to evaluate levels and frequencies of low BMD and obesity among ADT-treated men with prostate carcinoma.


Study Participants

Office charts and operative case logs were reviewed to identify men with Stage C or D prostate carcinoma (according to the Whitmore-Jewett Staging System) who were receiving ADT at the University of Arizona Medical Center. Treatment modalities included bilateral orchiectomy, luteinizing hormone-releasing hormone (LHRH) analogues (leuprolide acetate or goserelin acetate), and total androgen blockade with an LHRH analogue and nonsteroidal anti-androgen (flutamide).

The duration of therapy was calculated, and men with a minimum of 12 months and ≤ 5 years were recruited into the study. Exclusion criteria included intermittent ADT, active or hormonally refractory prostate carcinoma as demonstrated by a rising prostate specific antigen (PSA) level, known metastases involving the lumbar spine or hip, or prior lumbar spinal or hip surgery. Telephone interviews and an osteoporotic risk questionnaire were used to evaluate other potential causes of secondary osteoporosis. Men with known risk factors for secondary osteoporosis (chronic oral glucocorticoid use for > 6 months, chronic anticonvulsant use, diagnosis of hyperparathyroidism, chronic malabsorption, multiple myeloma, or inflammatory arthritis) were not included in the current study. Approximately 89% of the patients identified fulfilled the inclusion and exclusion criteria and consented to the study. Healthy controls were identified from the Tucson VAMC Prostate Cancer Screening database. These were men with stable and normal PSA levels (< 4.0 ng/mL) within 1 year of enrollment, and no identifiable risk factors for secondary osteoporosis. None of the controls had cancer at the time of enrollment. The response rate was similar in both the cases and the controls. Informed written consent was obtained from all the study participants. The study was approved by the Institutional Review Board.

BMD and Body Composition Measurements

BMD and body composition were measured using dual-energy X-ray absorptiometry (DXA) (QDR 4500W; Hologic, Inc., Waltham, MA). The QDR 4500W DXA system employs two X-ray beams of different energy, which are able to separate soft tissues from bone tissues in the measurements. The QDR 4500W has fan beam capability for total body, spine, and hip application. To ensure long-term stability of the DXA, daily calibrations were performed using a phantom with known density provided by Hologic, Inc. A Hologic-certified medical physician conducted all the measurements. Posterior-anterior lumbar spine scan and proximal femur scans were used to measure the lumbar spine (L1-4) and femoral BMDs. Total body soft tissue composition (including lean soft tissue mass, fat tissue mass, and percent body fat) was assessed using total body DXA scan.

The BMD for each subject was evaluated and T-scores were calculated based on both young male and female references from the National Health and Nutrition Examination Survey (NHANES) III for the appropriate racial or ethnic group.30

equation image

T-scores ≥ -1 were considered normal, whereas scores of < -1 and up to -2.5 were considered osteopenic, and scores < -2.5 were considered osteoporotic. Body composition measurements were categorized into tertiles based on the distribution within the controls. Body mass index (BMI) was classified into categories of < 25.0 kg/m2, 25.0–29.9 kg/m2, and ≥ 30.0 kg/m2 based on the Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults.23

Statistical Analysis

Overall comparisons between cases and controls were evaluated using Student t test for continuous variables and the chi-square test for categoric variables. Comparisons also were made between cases by method of treatment (surgical or medication) using the same statistical tests. Age-adjusted comparisons were examined using the mean least-squares method. The odds of having low BMD were calculated by combining the osteopenic and osteoporotic categories within each group and comparing cases with controls. For the evaluation of obesity, odds ratios were calculated by comparing cases with controls within each BMI category. Linear regression analyses were conducted to study the association between the length of ADT with BMD or body composition, adjusting for age and type of treatment.

Ethnicity was not considered as a risk factor in the current study because only four subjects were black and four subjects were Hispanic. The Hispanic subjects were split evenly between cases and controls. All black subjects were cases. In general, black men have a higher BMD compared with white men. Analyses were conducted with and without the black subjects to evaluate their impact on results. Statistical results were not found to be affected significantly; therefore the decision was made to include these four subjects in the final analysis.


This study included 62 prostate carcinoma cases and 47 controls (Table 1). Age and height were similar in the cases and controls. Based on weight, BMI, and DXA-derived body composition measurements, men with prostate carcinoma were significantly heavier and fatter than the controls. Total body BMD was lower among cases in comparison with controls. The average percent differences in the DXA-derived bone mass and body composition are presented in Figure 1. After controlling for age, the cases were 2–5% lower with regard to BMD than the controls, depending on the skeletal sites. Lean soft tissue mass was approximately 2% lower in cases. Men with prostate carcinoma were 13% and 21% higher, respectively, for percent body fat and for total body fat tissue mass. Among the 62 cases, 33 men received medical ADT and the remainder underwent orchiectomy. On average, men who were treated with surgical ADT were heavier and fatter than those treated with medical ADT (data not shown). After adjusting for age and treatment type (surgical vs. medication), length of treatment was found to be correlated inversely with weight, BMI, total bone mineral content (BMC), and trochanter BMD (P < 0.05). Using data from the NHANES III young female femoral BMD as the reference, 3–9% of the cases were found to have osteoporosis and 9–37% of the cases were found to have osteopenia at femoral neck, trochanter, intertrochanter, or the total hip measurements, whereas none of the controls were osteoporotic (Fig. 2). Figure 3 presents the percentage of osteoporosis and osteopenia among cases and controls when the NHANES III young male values were used as the reference. Men with prostate carcinoma were more likely to be osteoporotic or osteopenic compared with the controls. In Table 2, using a T-score of < -1.0 as the cut-off value for low BMD, the cases demonstrated an increased risk of low femoral BMD compared with the controls using either the male or the female young population as references. However, statistical significance was reached only at trochanter, intertrochanter, and total hip measurements when the male reference population was used.

Table 1. Characteristics of the Study Participants
 Cases (n = 62) Mean (95% CI)Controls (n = 47) Mean (95% CI)
  • 95% CI: 95% confidence interval; BMI: body mass index; BMC: bone mineral content; BMD: bone mineral density.

  • a

    Statistically significantly different at P < 0.05.

Age (yrs)74.3 (72.7–76.0)72.8 (70.9–74.6)
Height (m)1.77 (1.76–1.79)1.77 (1.75–1.79)
Weight (kg)86.5 (82.0–91.0)a80.6 (77.7–83.6)a
BMI (kg/m2)27.4 (26.2–28.6)a25.8 (24.9–26.6)a
Percent body fat (kg)30.2 (28.2–32.1)a25.7 (24.3–27.1)a
Fat tissue mass (kg)26.5 (23.7–29.6)a20.8 (19.3–22.4)a
Lean soft tissue mass (kg)55.6 (53.7–57.6)56.9 (54.9–58.8)
Total BMC (kg)2.7 (2.6–2.8)2.8 (2.7–2.9)
Total BMD (g/cm2)1.12 (1.08–1.15)a1.17 (1.14–1.20)a
Lumbar spine (L1–L4) BMD (g/cm2)1.03 (0.99–1.07)1.06 (1.02–1.11)
Total hip BMD (g/cm2)0.92 (0.88–0.96)0.94 (0.91–0.97)
Femoral neck BMD (g/cm2)0.76 (0.73–0.80)0.76 (0.73–0.79)
Trochanter BMD (g/cm2)0.75 (0.71–0.78)0.78 (0.75–0.81)
Intertrochanter BMD (g/cm2)1.06 (1.02–1.11)1.08 (1.04–1.12)
Figure 1.

Percent difference between age-adjusted dual-energy X-ray absorptiometry measurements (cases vs. controls). Asterisks indicate significant differences between cases and controls, controlling for age. BMC: bone mineral content; BMD: bone mineral density.

Figure 2.

Percent of osteoporotic and osteopenic subjects among cases and controls using female reference data. It is interesting to note that the chi-square tests demonstrated no significant differences between cases and controls. BMD: bone mineral density.

Figure 3.

Percent of osteoporotic and osteopenic subjects among cases and controls using male reference data. Asterisks indicate a significant difference of P < 0.05 between cases and controls using the chi-square test. BMD: bone mineral density.

Table 2. Risk of Low Bone Mineral Density (T-score < −1) by Case Status
BMDMale referencedFemale referenced
Low BMD (%)CrudeAge-adjustedLow BMD (%)CrudeAge-Adjusted
OR95% CIOR95% CIOR95% CIOR95% CI
  • BMD: bone mineral density; OR: odds ratio; 95% CI: 95% confidence interval.

  • a

    P < 0.01.

  • b

    P < 0.05.

Femoral neck          
 Controls65.961.00 1.00 40.431.00 1.00 
 Controls10.641.00 1.00 6.381.00 1.00 
 Controls29.791.00 1.00 19.151.00 1.00 
Total hip          
 Controls25.531.00 1.00 19.151.00 1.00 

Approximately 27% of the cases were obese based on their BMI measurement. Compared with controls, prostate carcinoma cases had an approximately 5.5-fold increase in the risk of being obese when age was adjusted. The cases also had an increased risk of being at the upper tertile for percent body fat (Table 3).

Table 3. Risk of Obesity by Case Status
 Cases (%)Controls (%)CrudeAge-adjusted
OR95% CIOR95% CI
  • OR: odds ratio; 95% CI: 95% confidence interval; BMI: body mass index.

  • a

    P value for trend < 0.01.

  • b

    Tertiles are based on control group distribution.

  • c

    P value for trend < 0.05.

BMI category      
 Normal (< 25.00)31.742.61.001.00
 Overweight (25.00–29.99)41.953.21.090.47–2.531.710.56–5.22
 Obesity (≥ 30.00)–51.475.56a0.73–42.35
Percent body fatb      
 Lower tertile (< 22.93%)14.531.91.001.00
 Mid-tertile (22.93–28.41%)25.834.01.670.56–4.992.120.33–13.80
 Upper tertile (> 28.41%)59.734.13.85a1.32–11.252.72c0.83–8.88


ADT for metastatic prostate carcinoma first was established by Huggins and Hodges in 1941.31 Currently available therapeutic modalities include bilateral orchiectomy, LHRH agonists (e.g., leuprolide and goserelin), and complete deprivation therapy using a combination of an LHRH agonist and a nonsteroidal anti-androgen (e.g., flutamide). Bilateral orchiectomy directly eliminates the gonadal androgen source and has been shown to produce a prompt, 95% reduction in the serum testosterone level within 3 hours postoperatively.32. LHRH indirectly inhibits gonadal androgen synthesis by superstimulating the pituitary gland, resulting in down-regulation of the LHRH receptors, leading to reduced gonadotropin (LH and follicle-stimulating hormone) secretion and reduced gonadal androgen production.33 Hypogonadism has long been found to be associated with low BMD,34, 35 although to our knowledge the mechanism is not fully understood. Androgens may increase bone formation directly through binding to androgen receptors on osteoblasts or by increasing insulin-like growth factor-1 levels. Indirectly, androgen also may alter bone formation via an effect on muscle. Muscle strength is responsive to androgens and has a substantial potential effect on increasing and maintaining bone mass.36 We found a significant correlation between lean soft tissue composition and BMD among our study participants (data not shown). Men with prostate carcinoma had both low BMD and lean soft tissue composition, indicating multiple risk factors for osteoporotic fractures.

On average, cases in the current study were 2.2%, 3.5%, and 4.8% lower, respectively, than controls for total hip, spine, and total body BMD. BMD is an important determinant of the diagnosis of osteoporosis. The World Health Organization (WHO) has defined osteoporosis as a measurement BMD that is ≥ 2.5 standard deviations (SD) below the young average value in women (T-score ≤ −2.5), whereas a BMD that is 1–2.5 SD below the young average value is considered an indication of osteopenia (low bone density) (−2.5 < T-score < −1).37 These criteria have been widely used in postmenopausal women, although their application in men currently is controversial. Recently, the Committee of Scientific Advisors at the International Osteoporosis Foundation published a position paper that recommended that the BMD thresholds used to determine osteoporosis in women should be used in men (rather than similar t-scores, which would yield different absolute BMD levels for men and women) until further research changes this view. They concluded that the relation between BMD and hip fracture risk is similar in men and women,38 and recommended that the WHO criteria should be reserved for DXA-derived hip BMD because BMD assessments at the lumbar spine are susceptible to artificial effects such as osteoarthritis.

Based on the total hip BMD measurement on DXA, we found an increased risk of osteoporosis and osteopenia among cases in comparison with controls when the young male or female values from the NHANES III30 were used as the reference range. The low BMD and excess risk of osteoporosis among men with prostate carcinoma in the current study could be the result of the combination of a low BMD before ADT8–11 and an accelerated bone loss induced by ADT.15, 17 Examining DXA-derived BMD among a group of men with prostate carcinoma before ADT, Smith et al. reported that 29% of the patients had T-scores between −1.0 and −2.5, and another 5% had T-scores <−2.5 as measured by DXA.11 The findings of low BMD in men with prostate carcinoma in these studies were the result of comparisons with young reference populations provided by each DXA manufacturer. It should be noted that the prevalence of osteoporosis or low bone density could be altered depending on the reference populations39 and the technology used. More important, because men do lose bone with normal aging,7, 28, 40 it is not surprising that aging men have low bone density and an increased risk of osteoporosis compared with a young population. To investigate the possible association between low BMD and a diagnosis of prostate carcinoma, future studies need to evaluate the severity of low BMD and the excess risk of osteoporosis in men with prostate carcinoma before ADT using an age-matched reference population as the comparison group.

In contrast, there is less doubt regarding accelerated bone loss after ADT. Bone fractures after orchiectomy18 or chemical castration have been reported in small case studies15, 17 and in a telephone survey study.19 In the study by Daniell,18 nearly 50% of men who survived 9 years after orchiectomy sustained a fracture of the hip or spine, suggesting that osteoporosis is a very severe health problem among prostate carcinoma survivors treated with ADT. A number of studies have evaluated the longitudinal consequences of hypogonadism on bone density in these patients.9, 10, 12–14 Goldray et al.12 studied 17 elderly men (mean age of 72 years) with benign prostatic hyperplasia who were treated with a gonadotropin-releasing hormone (GnRH) agonist, and found a significant decrease in BMD among patients after 6–12 months of ADT. Another study reported similar rates of bone loss in patients treated with GnRH agonists.13 Both studies suggested that GnRH agonist-induced hypogonadism causes high bone turnover and accelerated bone loss in some men. Eriksson et al. studied 27 patients with nonmetastatic prostate carcinoma whose BMD and BMC were measured before and after orchiectomy.10 Their study results demonstrated that BMD and BMC decreased in all areas studied, but the changes were statistically significant only in the distal radius. Daniell et al.14 measured BMD changes among 26 prostate carcinoma patients treated with ADT and found bone loss of 2.4% and 7.6%, respectively, in the first and second years after ADT, whereas the rate of bone loss was reduced to 1.4–2.6% yearly at the total hip in subsequent years. The rate of bone loss was similar in both surgical and medical castration patients in this study, a finding that is consistent with the results of the current study (data not shown). In a 48-week trial of patients receiving leuprolide or a combination of leuprolide and pamidronate, Smith et al.41 found decreases in BMD of 3.3%, 2.1%, and 1.8%, respectively, in the lumbar spine, trochanter, and total hip by DXA, and a decrease of 8.5% in the lumbar spine as measured by quantitative computed tomography scan among men with prostate carcinoma who were treated with leuprolide alone. No significant bone loss was reported in patients receiving both leuprolide and intravenous pamidronate (60 mg every 12 weeks).41 This study provides clear evidence that bone loss and osteoporosis can be prevented effectively by proper treatment in men with prostate carcinoma who have undergone ADT.

In the current study, we found a high prevalence of obesity among the prostate carcinoma cases. Obesity reportedly is associated with an increased risk of prostate carcinoma.42–48 It is possible that the men with prostate carcinoma had a higher percent of body fat at the time of their cancer diagnosis. In addition, ADT also may cause significant weight gain.49 Longitudinal studies should be performed to distinguish the causes of obesity among this patient population. Obesity is a significant risk factor for many chronic diseases, including diabetes and cardiovascular diseases.20–23, 48, 50, 51 Among breast carcinoma patients, it has been found that an increased BMI is associated with a poor prognosis.52 To our knowledge, however, the association between obesity and survivorship among men with prostate carcinoma remains largely unknown.

Increased weight usually is a protective factor for BMD through increasing mechanic stretch and loading on bone. However, the results of the current study demonstrated that in spite of the high body weight observed in the prostate carcinoma cases, their BMDs were significantly lower than those of the controls. This may be due in part to the fact that the increased weight among men with prostate carcinoma mainly is a result of increased fat tissue mass, which has less of an impact on BMD than muscle mass. In fact, men with prostate carcinoma appear to have lower lean soft tissue mass compared with controls, but the group difference in lean soft tissue mass was not found to be statistically significant.

The cross-sectional nature of the current study allows us to examine the excess proportion of low bone density and obesity among ADT-treated men with prostate carcinoma compared with controls. However, the causes of low BMD and obesity among the men with prostate carcinoma could not be elucidated based on the results of the current study. In addition, the current study results should be interpreted with caution because of the small sample size and lack of information regarding some covariates, such as a history of diabetes. Nevertheless, we believe the current study provides important evidence regarding the magnitude and prevalence of low BMD and obesity among men with prostate carcinoma who are treated with ADT. The results of this study may increase the awareness of clinicians and medical researchers regarding the severity of osteoporosis and obesity among prostate carcinoma patients. Clearly, research is needed to reveal causes of low BMD and obesity among men with prostate carcinoma after ADT so that undesirable changes in body composition and bone density can be prevented.

In summary, results of the current study demonstrate that low bone density and obesity were more prevalent among men who were treated with ADT for prostate carcinoma compared with healthy controls. Research and medical attention should be focused on how to prevent and manage bone loss and unhealthy weight changes in men with prostate carcinoma during and after ADT.

Note Added in Proof

Since the completion of the current study, several other cross-sectional53 or longitudinal54–56 studies have been published, supporting the findings of the current study. Collectively, the research in this area suggests that increased body fat mass and reduced lean tissue mass is a common phenomenon among men who are treated with ADT for prostate carcinoma.