Article
Low Testosterone Levels Are Associated With Poor Peripheral Bone Mineral Density and Quantitative Bone Ultrasound at Phalanges and Calcaneus in Healthy Elderly Men
Biological Research for Nursing 1-6 ª The Author(s) 2014 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1099800414532710 brn.sagepub.com
Jose M. Moran, PhD, MHS1, Raul Roncero Martin, RN, PhD1, Maria Pedrera-Canal, MD1, Javier Alonso-Terron, RN1, Francisco J. Rodriguez-Velasco, RN, PhD1, and Juan D. Pedrera-Zamorano, PhD, MD1
Abstract Variations in sex hormones influence bone health in men. Aging in men is associated with a decrease in testosterone (T) levels. We examined the relationship between T levels and changes in bone health status as measured by quantitative ultrasound (QUS) at the phalanges and the os calcis and by peripheral bone mineral density (pBMD) at the phalanges in healthy elderly Spanish men. We examined 162 men aged 65–88 years and assessed total serum T concentrations. Total serum T < 300 ng/dL was used as the threshold for biochemical T deficiency. The sample was divided into low (n ¼ 66) or normal (n ¼ 96) T levels; both groups were matched for age, weight, height, and body mass index (p > .05 for all the comparisons). All measured bone parameters were higher in the normal serum T group (p < .05). Multiple regression analysis revealed that serum T was an independent predictor of both QUS at the calcaneus and phalangeal pBMD. Our data indicate that T is an independent determinant of QUS at the os calcis and pBMD at the phalanges in elderly Spanish men. Keywords testosterone, bone ultrasound, peripheral dual-energy x-ray absorptiometry, men
Aging in men is associated with mobility difficulties, the accumulation of central fat, decreased muscle mass and strength, increased body fat, and decreased bone mineral density (BMD; Harman, Metter, Tobin, Pearson, & Blackman, 2001). These traits are similar to the abnormalities observed in young men with hypogonadism. Researchers have described a decrease in testosterone (T) and dihydroepiandrosterone levels in men 40–50 years old (Feldman et al., 2002; Harman et al., 2001; Liu et al., 2007). T regulates bone turnover in men via the promotion of bone formation (Tuck & Francis, 2009). Therefore, authors have suggested that T plays an important role in the regulation of bone metabolism and bone loss in elderly men (Martin, 2011). Dual-energy x-ray absorptiometry (DXA) of the hip and spine is the ‘‘gold standard’’ for the diagnosis of osteoporosis; however, peripheral densitometry is a feasible, sensitive, and inexpensive method for the assessment of bone tissue in populations that do not have access to central DXA equipment or in patients with conditions that do not allow a central DXA to be performed. Bone densitometry using peripheral DXA (pBMD),
which measures BMD in the finger, is strongly correlated with the BMD of the hand and forearm (Bouxsein, Michaeli, Plass, Schick, & Melton, 1997). Longitudinal studies have addressed the association of T levels with the results of quantitative bone ultrasound (QUS) at the os calcis (Gennari et al., 2003) and phalanges (Zitzmann, Brune, Vieth, & Nieschlag, 2002). However, limited information is available on the associations between T levels and pBMD, os calcis QUS, and phalanx QUS in elderly men (Morote et al., 2006; Rapado et al., 1999). In the present study, we aimed to investigate the associations between total serum
1
Metabolic Bone Diseases Research Group, School of Nursing, University of Extremadura, Avd., Caceres, Spain Corresponding Author: Juan D. Pedrera-Zamorano, PhD, MD, Departamento de Enfermerı´a, Universidad de Extremadura, Ca´ceres 10003, Spain. Email: [email protected]
Downloaded from brn.sagepub.com at Bobst Library, New York University on May 31, 2015
2
Biological Research for Nursing
T levels and phalangeal pBMD, os calcis QUS, and phalangeal QUS in elderly Spanish males.
Material and Method A total of 162 healthy men with a mean age of 74.25 (standard deviation [SD] 4.97) years participated in this cross-sectional research. We recruited participants in a clinical convenience sample from the general community through advertising in primary care centers. The study was conducted among noninstitutionalized Spanish men aged 65–88 years in the area of Llerena-Zafra, Badajoz, Spain. All participants provided written informed consent. The Office for Protection against Research Risk of the University of Extremadura approved the study. We took a complete medical history and performed a physical examination prior to candidate enrollment in the study. We drew a blood sample from each participant between 8:00 and 10:00 a.m., after overnight fasting. Normality was also established by weight within the range of 70–130% of ideal body weight for height and by the results of a biochemical study of blood glucose, transaminases, g-glutamyl transpeptidase, creatinine, calcium, phosphorus, total proteins, bilirubin, alkaline phosphatase, tartrate-resistant acid phosphatase, and a coagulation study. The study subjects were not taking any medication and reported no disease, including diseases that are associated with abnormalities in mineral metabolism (e.g., diabetes mellitus, liver disease, renal osteodystrophy, or parathyroid, thyroid, or adrenal disease) that could interfere with calcium metabolism. All participants led active lives, but none of the subjects practiced any recreational or professional sports. Height measurements were performed using a Harpender stadiometer with mandible planes parallel to the floor, and patients were weighed on a biomedical balance. Both measurements were performed while subjects wore pajamas without shoes. We calculated body mass index (BMI) by dividing the weight in kilograms by the square of height in meters (kg/m2). We evaluated total serum T using commercial methods (Roche Diagnostics, Indianapolis, IN). The intra- and interassay coefficients of variation (CVs) were 8% and 10%, respectively. Participants were categorized as having low or normal serum T levels, according to the most widely used threshold for low total testosterone level, < 300 ng/dL (10.4 nmol/L), for some of the calculations in this article (Bhasin et al., 2010). Free T levels were not analyzed in this study.
tangentially and measured the amplitude-dependent speed of sound (Ad-SoS) in meters per second through the phalanx. Positioning and repositioning the instrument were easy because the clip is placed just behind the prominences of the lower phalangeal epiphysis. The instrument transmitted at a frequency of 1.2 MHz with 22 W power. Instrument precision was determined from three measurements in eight subjects at time intervals not exceeding 21 days. The CV was 0.77%. The interobserver CV was 1.1%. Os calcis QUS (measured as the broadband ultrasound attenuation in db/MHz) was determined as described previously (Gonnelli et al., 2005) in the left calcaneus using an ultrasonic bone analyzer (CUBA Clinical, McCue Ultrasonic Ltd., Winchester, United Kingdom). Precision was evaluated by performing five replicate scans on each subject. The CV was 1.5%.
Peripheral DXA The pBMD of the middle phalange of the third finger (g/cm2) of the nondominant hand was measured using the accuDEXA device (Lone Oak Medical Technologies, Doylestown, PA), as previously described (Lohiya, Tan-Figueroa, & Iannucci, 2004). The accuDEXA device utilizes DXA technology, which is the standard for bone densitometry. DXA is also designed to assess bone mineral content. The same investigator analyzed all exams. Proper finger placement inside the hand slot was assured, as it is essential for test precision. The QC phantom test provided by the manufacturer was used as an additional quality control check of the accuDEXA system. The CV was 1.1%.
Statistical Analyses All values are expressed as mean + SD. The normal distribution of data was confirmed by calculating the skewness and kurtosis prior to the application of standard tests. The groups were compared using analysis of variance to determine the differences when appropriate. A minimum p value of < .05 was accepted for statistical significance. Regression and correlation analyses were used when appropriate to examine the relationships between continuous variables. Stepwise multiple linear regression analysis was used to estimate the linear relationships between dependent variables (age, weight, height, BMI, and total serum T) and various independent variables. Analyses were performed in SPSS 19.0.
Ultrasound Studies Phalangeal QUS was measured as described previously (Calderon-Garcia et al., 2013; Pedrera-Zamorano et al., 2012; Pedrera-Zamorano, Lavado-Garcia, & Moran, 2013). All of the men underwent an ultrasound study of the second to fifth proximal phalanges of the nondominant hand, and we calculated the mean of all measurements. The ultrasound study was performed using a model DBM Sonic Bone Profiler (Igea, Capri, Italy) equipped with a caliper that closed on the phalanx
Results A total of 162 healthy elderly men participated in the present study. Their mean total T level was 365.60 + 171.11 ng/dL (range 16.90–990.50 ng/dL). We observed no significant differences in the total serum T levels between age-groups (p > .05; see Table 1). A total of 40.7% of the participants (n ¼ 66) exhibited a total T level < 300 ng/dL, while the remaining participants exhibited a total T level of 300–1200
Downloaded from brn.sagepub.com at Bobst Library, New York University on May 31, 2015
Moran et al.
3
Table 1. Total Testosterone Levels (ng/dL) by Age-Group. Total Testosterone (ng/dL) Age-Group (years) 65–69 (n ¼ 45) 70–74 (n ¼ 77) 75–79 (n ¼ 51) >80 (n ¼ 21)
Mean (SD) 335.01 394.78 370.38 302.91
Range
(162.88) (152.50) (192.49) (168.86)
22.30–712.00 79.60–689.70 3.42–990.50 16.90–758.40
Note. SD ¼ standard deviation. No significant differences were observed between groups (p ¼ .135).
ng/dL, which was within the normal reference values for elderly Spanish men. The results of pBMD (g/cm2) at the phalanges and QUS at the phalanges and os calcis are presented by total T-level group in Table 2. There were no significant differences between the two groups in age, weight, height, or BMI (p > .05 in all cases). We did observe significant differences between the low and the normal total T groups across the three techniques used, with lower values of pBMD (6.32%), os calcis QUS (7.92%), and phalanx QUS (1.58%) in the low-serum T group versus the normal serum T level group in all cases (p < .05). All differences remained significant after further adjustment for potential confounders (age, weight, and BMI). Using multivariate analysis, we evaluated the independent contributions to the variance in phalangeal pBMD, os calcis QUS, and phalanx QUS in the total group. Age, height, and BMI were independently related to phalanx QUS; total serum T and weight were related to os calcis QUS; age, height, and total serum T were related to pBMD (p < .01 in all cases; Table 3). When the participants were categorized into low (< 300 ng/dL) and normal (300–1200 ng/dL) total serum T levels, age (r ¼ .269, b ¼ 4.521; p ¼ .045) was a predictor of Ad-SoS for the low T group, and both BMI (r ¼ .313, b ¼ 7.594; p ¼ .003) and age (r ¼ .308, b ¼ 5.514; p ¼ .003) were predictors in the normal serum T level group. Serum total T was a positive predictor of pBMD in the low-serum T group (r ¼ .296, b < .000; p ¼ .016), and age was a negative determinant of pBMD in the normal serum T level group (r ¼ .334, b ¼ .006; p ¼ .001). None of the studied variables were independently related to os calcis QUS when adjusted for total serum T levels.
Discussion Results of the present study demonstrate that total serum T levels are independently associated with pBMD and os calcis QUS in apparently healthy elderly Spanish men. Additionally, phalangeal pBMD, os calcis QUS, and phalanx QUS were significantly higher in the group whose total serum T levels were within the reference values (300–1200 ng/dL). Several researchers have proposed that there is a gradual decrease in total T levels during aging, with an average decrease of 1–2% per year after the age of 40 (Harman et al., 2001). In the current study, we did not detect significant
variation in T level by age. Other authors have described similarly unexpected results (Rhoden, Teloken, Sogari, & Souto, 2002). The age-associated decrease in T levels varies highly across individuals, and enormous variations in T values are observed in men of all ages. Additionally, the average level of total serum T in 75-year-olds is approximately 66% of the level in 25-year-olds, but the average values of free and bioavailable serum T in older men are approximately half of those of healthy young men (Feldman et al., 2002; Harman et al., 2001; Nardozza Junior, Szelbracikowski Sdos, Nardi, & Almeida, 2011; Vermeulen, Kaufman, & Giagulli, 1996; Zmuda et al., 1997). Further investigation is thus required to address the role of the free T fraction rather than of total serum T, which was considered in the present study. The results from the present study on the association between serum T levels and QUS are controversial. We observed higher QUS values in patients with normal serum T levels versus those with low levels; however, several previous studies have addressed this question and have not found an association (Boonen et al., 2011; Gennari et al., 2003; Martinez-Jabaloyas et al., 2011; Vanderschueren et al., 2010). On the other hand, consistent with our observations, additional studies have demonstrated that QUS determinations are significantly lower in men with low levels of bioavailable T (Kuchuk et al., 2007) and that the treatment of hypogonadal men with T significantly increases the levels of QUS compared to untreated patients with secondary hypogonadism (Zitzmann et al., 2002). Due to these inconsistent results, authors have proposed that bone health, as measured by QUS, may be more strongly associated with total and free estradiol than with total T (Vanderschueren et al., 2010). In the present study, we observed that total serum concentrations of T accounted for a significant proportion of the variation in QUS at the os calcis and in phalangeal pBMD. These results support the hypothesis that the bone loss observed in elderly men (at least as determined by os calcis QUS and pBMD at the phalanges) is partially related to hormonal changes, among other contributing factors. Interestingly, total serum T did not predict QUS results at the phalanges. QUS measurements are determined by bone properties (e.g., elastic modulus) more than by BMD, and the results that are obtained by this technique are difficult to extrapolate to other skeletal sites. The finding that total serum T was a positive determinant of QUS at the os calcis may be explained by the fact that the calcaneus is a predominantly (95%) trabecular bone, which may be more sensitive to variations in sex hormones (Vanderschueren et al., 2010) than bone that is predominantly cortical. A number of studies have demonstrated associations between T and BMD measured centrally. Isidori and colleagues (2005) performed a meta-analysis of 1,083 subjects with a mean age of 64.5 years (46.9–77.6 years) and demonstrated that T improved BMD in the lumbar spine by up to 3.7% but not in the femoral neck. Cross-sectional studies have also associated higher BMD with higher T levels (Khosla, Melton, & Riggs, 2001; Vandenput et al., 2007). Lower BMD at the femur has been observed in men with T levels < 7 nmol/L
Downloaded from brn.sagepub.com at Bobst Library, New York University on May 31, 2015
4
Biological Research for Nursing
Table 2. Anthropometric Data and Bone Parameters by Testosterone (T)-Level Group. Variable Age (years) Weight (kg) BMI (kg/m2) Ad-SoS (m/s) BUA (dB/MHz) pBMD (g/cm2)
Low-Serum T Mean (SD)
Normal-Serum T Mean (SD)
p Value
Adjusted p Value*
75.11 (5.81) 82.768 (12.18 31.02 (4.34) 1987.04 (99.28) 71.48 (18.52) 0.592 (0.084)
73.66 (4.23) 81.651 (10.83) 29.94 (3.63) 2018.99 (74.14) 77.63 (17.88) 0.632 (0.081)
.068 .541 .087 .028 .044 .003
< .001 .028
Low Testosterone Levels Are Associated With Poor Peripheral Bone Mineral Density and Quantitative Bone Ultrasound at Phalanges and Calcaneus in Healthy Elderly Men
Biological Research for Nursing 1-6 ª The Author(s) 2014 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1099800414532710 brn.sagepub.com
Jose M. Moran, PhD, MHS1, Raul Roncero Martin, RN, PhD1, Maria Pedrera-Canal, MD1, Javier Alonso-Terron, RN1, Francisco J. Rodriguez-Velasco, RN, PhD1, and Juan D. Pedrera-Zamorano, PhD, MD1
Abstract Variations in sex hormones influence bone health in men. Aging in men is associated with a decrease in testosterone (T) levels. We examined the relationship between T levels and changes in bone health status as measured by quantitative ultrasound (QUS) at the phalanges and the os calcis and by peripheral bone mineral density (pBMD) at the phalanges in healthy elderly Spanish men. We examined 162 men aged 65–88 years and assessed total serum T concentrations. Total serum T < 300 ng/dL was used as the threshold for biochemical T deficiency. The sample was divided into low (n ¼ 66) or normal (n ¼ 96) T levels; both groups were matched for age, weight, height, and body mass index (p > .05 for all the comparisons). All measured bone parameters were higher in the normal serum T group (p < .05). Multiple regression analysis revealed that serum T was an independent predictor of both QUS at the calcaneus and phalangeal pBMD. Our data indicate that T is an independent determinant of QUS at the os calcis and pBMD at the phalanges in elderly Spanish men. Keywords testosterone, bone ultrasound, peripheral dual-energy x-ray absorptiometry, men
Aging in men is associated with mobility difficulties, the accumulation of central fat, decreased muscle mass and strength, increased body fat, and decreased bone mineral density (BMD; Harman, Metter, Tobin, Pearson, & Blackman, 2001). These traits are similar to the abnormalities observed in young men with hypogonadism. Researchers have described a decrease in testosterone (T) and dihydroepiandrosterone levels in men 40–50 years old (Feldman et al., 2002; Harman et al., 2001; Liu et al., 2007). T regulates bone turnover in men via the promotion of bone formation (Tuck & Francis, 2009). Therefore, authors have suggested that T plays an important role in the regulation of bone metabolism and bone loss in elderly men (Martin, 2011). Dual-energy x-ray absorptiometry (DXA) of the hip and spine is the ‘‘gold standard’’ for the diagnosis of osteoporosis; however, peripheral densitometry is a feasible, sensitive, and inexpensive method for the assessment of bone tissue in populations that do not have access to central DXA equipment or in patients with conditions that do not allow a central DXA to be performed. Bone densitometry using peripheral DXA (pBMD),
which measures BMD in the finger, is strongly correlated with the BMD of the hand and forearm (Bouxsein, Michaeli, Plass, Schick, & Melton, 1997). Longitudinal studies have addressed the association of T levels with the results of quantitative bone ultrasound (QUS) at the os calcis (Gennari et al., 2003) and phalanges (Zitzmann, Brune, Vieth, & Nieschlag, 2002). However, limited information is available on the associations between T levels and pBMD, os calcis QUS, and phalanx QUS in elderly men (Morote et al., 2006; Rapado et al., 1999). In the present study, we aimed to investigate the associations between total serum
1
Metabolic Bone Diseases Research Group, School of Nursing, University of Extremadura, Avd., Caceres, Spain Corresponding Author: Juan D. Pedrera-Zamorano, PhD, MD, Departamento de Enfermerı´a, Universidad de Extremadura, Ca´ceres 10003, Spain. Email: [email protected]
Downloaded from brn.sagepub.com at Bobst Library, New York University on May 31, 2015
2
Biological Research for Nursing
T levels and phalangeal pBMD, os calcis QUS, and phalangeal QUS in elderly Spanish males.
Material and Method A total of 162 healthy men with a mean age of 74.25 (standard deviation [SD] 4.97) years participated in this cross-sectional research. We recruited participants in a clinical convenience sample from the general community through advertising in primary care centers. The study was conducted among noninstitutionalized Spanish men aged 65–88 years in the area of Llerena-Zafra, Badajoz, Spain. All participants provided written informed consent. The Office for Protection against Research Risk of the University of Extremadura approved the study. We took a complete medical history and performed a physical examination prior to candidate enrollment in the study. We drew a blood sample from each participant between 8:00 and 10:00 a.m., after overnight fasting. Normality was also established by weight within the range of 70–130% of ideal body weight for height and by the results of a biochemical study of blood glucose, transaminases, g-glutamyl transpeptidase, creatinine, calcium, phosphorus, total proteins, bilirubin, alkaline phosphatase, tartrate-resistant acid phosphatase, and a coagulation study. The study subjects were not taking any medication and reported no disease, including diseases that are associated with abnormalities in mineral metabolism (e.g., diabetes mellitus, liver disease, renal osteodystrophy, or parathyroid, thyroid, or adrenal disease) that could interfere with calcium metabolism. All participants led active lives, but none of the subjects practiced any recreational or professional sports. Height measurements were performed using a Harpender stadiometer with mandible planes parallel to the floor, and patients were weighed on a biomedical balance. Both measurements were performed while subjects wore pajamas without shoes. We calculated body mass index (BMI) by dividing the weight in kilograms by the square of height in meters (kg/m2). We evaluated total serum T using commercial methods (Roche Diagnostics, Indianapolis, IN). The intra- and interassay coefficients of variation (CVs) were 8% and 10%, respectively. Participants were categorized as having low or normal serum T levels, according to the most widely used threshold for low total testosterone level, < 300 ng/dL (10.4 nmol/L), for some of the calculations in this article (Bhasin et al., 2010). Free T levels were not analyzed in this study.
tangentially and measured the amplitude-dependent speed of sound (Ad-SoS) in meters per second through the phalanx. Positioning and repositioning the instrument were easy because the clip is placed just behind the prominences of the lower phalangeal epiphysis. The instrument transmitted at a frequency of 1.2 MHz with 22 W power. Instrument precision was determined from three measurements in eight subjects at time intervals not exceeding 21 days. The CV was 0.77%. The interobserver CV was 1.1%. Os calcis QUS (measured as the broadband ultrasound attenuation in db/MHz) was determined as described previously (Gonnelli et al., 2005) in the left calcaneus using an ultrasonic bone analyzer (CUBA Clinical, McCue Ultrasonic Ltd., Winchester, United Kingdom). Precision was evaluated by performing five replicate scans on each subject. The CV was 1.5%.
Peripheral DXA The pBMD of the middle phalange of the third finger (g/cm2) of the nondominant hand was measured using the accuDEXA device (Lone Oak Medical Technologies, Doylestown, PA), as previously described (Lohiya, Tan-Figueroa, & Iannucci, 2004). The accuDEXA device utilizes DXA technology, which is the standard for bone densitometry. DXA is also designed to assess bone mineral content. The same investigator analyzed all exams. Proper finger placement inside the hand slot was assured, as it is essential for test precision. The QC phantom test provided by the manufacturer was used as an additional quality control check of the accuDEXA system. The CV was 1.1%.
Statistical Analyses All values are expressed as mean + SD. The normal distribution of data was confirmed by calculating the skewness and kurtosis prior to the application of standard tests. The groups were compared using analysis of variance to determine the differences when appropriate. A minimum p value of < .05 was accepted for statistical significance. Regression and correlation analyses were used when appropriate to examine the relationships between continuous variables. Stepwise multiple linear regression analysis was used to estimate the linear relationships between dependent variables (age, weight, height, BMI, and total serum T) and various independent variables. Analyses were performed in SPSS 19.0.
Ultrasound Studies Phalangeal QUS was measured as described previously (Calderon-Garcia et al., 2013; Pedrera-Zamorano et al., 2012; Pedrera-Zamorano, Lavado-Garcia, & Moran, 2013). All of the men underwent an ultrasound study of the second to fifth proximal phalanges of the nondominant hand, and we calculated the mean of all measurements. The ultrasound study was performed using a model DBM Sonic Bone Profiler (Igea, Capri, Italy) equipped with a caliper that closed on the phalanx
Results A total of 162 healthy elderly men participated in the present study. Their mean total T level was 365.60 + 171.11 ng/dL (range 16.90–990.50 ng/dL). We observed no significant differences in the total serum T levels between age-groups (p > .05; see Table 1). A total of 40.7% of the participants (n ¼ 66) exhibited a total T level < 300 ng/dL, while the remaining participants exhibited a total T level of 300–1200
Downloaded from brn.sagepub.com at Bobst Library, New York University on May 31, 2015
Moran et al.
3
Table 1. Total Testosterone Levels (ng/dL) by Age-Group. Total Testosterone (ng/dL) Age-Group (years) 65–69 (n ¼ 45) 70–74 (n ¼ 77) 75–79 (n ¼ 51) >80 (n ¼ 21)
Mean (SD) 335.01 394.78 370.38 302.91
Range
(162.88) (152.50) (192.49) (168.86)
22.30–712.00 79.60–689.70 3.42–990.50 16.90–758.40
Note. SD ¼ standard deviation. No significant differences were observed between groups (p ¼ .135).
ng/dL, which was within the normal reference values for elderly Spanish men. The results of pBMD (g/cm2) at the phalanges and QUS at the phalanges and os calcis are presented by total T-level group in Table 2. There were no significant differences between the two groups in age, weight, height, or BMI (p > .05 in all cases). We did observe significant differences between the low and the normal total T groups across the three techniques used, with lower values of pBMD (6.32%), os calcis QUS (7.92%), and phalanx QUS (1.58%) in the low-serum T group versus the normal serum T level group in all cases (p < .05). All differences remained significant after further adjustment for potential confounders (age, weight, and BMI). Using multivariate analysis, we evaluated the independent contributions to the variance in phalangeal pBMD, os calcis QUS, and phalanx QUS in the total group. Age, height, and BMI were independently related to phalanx QUS; total serum T and weight were related to os calcis QUS; age, height, and total serum T were related to pBMD (p < .01 in all cases; Table 3). When the participants were categorized into low (< 300 ng/dL) and normal (300–1200 ng/dL) total serum T levels, age (r ¼ .269, b ¼ 4.521; p ¼ .045) was a predictor of Ad-SoS for the low T group, and both BMI (r ¼ .313, b ¼ 7.594; p ¼ .003) and age (r ¼ .308, b ¼ 5.514; p ¼ .003) were predictors in the normal serum T level group. Serum total T was a positive predictor of pBMD in the low-serum T group (r ¼ .296, b < .000; p ¼ .016), and age was a negative determinant of pBMD in the normal serum T level group (r ¼ .334, b ¼ .006; p ¼ .001). None of the studied variables were independently related to os calcis QUS when adjusted for total serum T levels.
Discussion Results of the present study demonstrate that total serum T levels are independently associated with pBMD and os calcis QUS in apparently healthy elderly Spanish men. Additionally, phalangeal pBMD, os calcis QUS, and phalanx QUS were significantly higher in the group whose total serum T levels were within the reference values (300–1200 ng/dL). Several researchers have proposed that there is a gradual decrease in total T levels during aging, with an average decrease of 1–2% per year after the age of 40 (Harman et al., 2001). In the current study, we did not detect significant
variation in T level by age. Other authors have described similarly unexpected results (Rhoden, Teloken, Sogari, & Souto, 2002). The age-associated decrease in T levels varies highly across individuals, and enormous variations in T values are observed in men of all ages. Additionally, the average level of total serum T in 75-year-olds is approximately 66% of the level in 25-year-olds, but the average values of free and bioavailable serum T in older men are approximately half of those of healthy young men (Feldman et al., 2002; Harman et al., 2001; Nardozza Junior, Szelbracikowski Sdos, Nardi, & Almeida, 2011; Vermeulen, Kaufman, & Giagulli, 1996; Zmuda et al., 1997). Further investigation is thus required to address the role of the free T fraction rather than of total serum T, which was considered in the present study. The results from the present study on the association between serum T levels and QUS are controversial. We observed higher QUS values in patients with normal serum T levels versus those with low levels; however, several previous studies have addressed this question and have not found an association (Boonen et al., 2011; Gennari et al., 2003; Martinez-Jabaloyas et al., 2011; Vanderschueren et al., 2010). On the other hand, consistent with our observations, additional studies have demonstrated that QUS determinations are significantly lower in men with low levels of bioavailable T (Kuchuk et al., 2007) and that the treatment of hypogonadal men with T significantly increases the levels of QUS compared to untreated patients with secondary hypogonadism (Zitzmann et al., 2002). Due to these inconsistent results, authors have proposed that bone health, as measured by QUS, may be more strongly associated with total and free estradiol than with total T (Vanderschueren et al., 2010). In the present study, we observed that total serum concentrations of T accounted for a significant proportion of the variation in QUS at the os calcis and in phalangeal pBMD. These results support the hypothesis that the bone loss observed in elderly men (at least as determined by os calcis QUS and pBMD at the phalanges) is partially related to hormonal changes, among other contributing factors. Interestingly, total serum T did not predict QUS results at the phalanges. QUS measurements are determined by bone properties (e.g., elastic modulus) more than by BMD, and the results that are obtained by this technique are difficult to extrapolate to other skeletal sites. The finding that total serum T was a positive determinant of QUS at the os calcis may be explained by the fact that the calcaneus is a predominantly (95%) trabecular bone, which may be more sensitive to variations in sex hormones (Vanderschueren et al., 2010) than bone that is predominantly cortical. A number of studies have demonstrated associations between T and BMD measured centrally. Isidori and colleagues (2005) performed a meta-analysis of 1,083 subjects with a mean age of 64.5 years (46.9–77.6 years) and demonstrated that T improved BMD in the lumbar spine by up to 3.7% but not in the femoral neck. Cross-sectional studies have also associated higher BMD with higher T levels (Khosla, Melton, & Riggs, 2001; Vandenput et al., 2007). Lower BMD at the femur has been observed in men with T levels < 7 nmol/L
Downloaded from brn.sagepub.com at Bobst Library, New York University on May 31, 2015
4
Biological Research for Nursing
Table 2. Anthropometric Data and Bone Parameters by Testosterone (T)-Level Group. Variable Age (years) Weight (kg) BMI (kg/m2) Ad-SoS (m/s) BUA (dB/MHz) pBMD (g/cm2)
Low-Serum T Mean (SD)
Normal-Serum T Mean (SD)
p Value
Adjusted p Value*
75.11 (5.81) 82.768 (12.18 31.02 (4.34) 1987.04 (99.28) 71.48 (18.52) 0.592 (0.084)
73.66 (4.23) 81.651 (10.83) 29.94 (3.63) 2018.99 (74.14) 77.63 (17.88) 0.632 (0.081)
.068 .541 .087 .028 .044 .003
< .001 .028
