- Email: [email protected]
Usefulness of Bone Quantitative Ultrasound in Management of Osteoporosis in Men
Journal of Clinical Densitometry, vol. 4, no. 3, 231–237, Fall 2001 © Copyright 2001 by Humana Press Inc. All rights of any nature whatsoever reserved. 1094-6950/01/4:231–237/$11.75
Original Article
Usefulness of Bone Quantitative Ultrasound in Management of Osteoporosis in Men Andrea Montagnani, MD, Stefano Gonnelli, MD, Chiara Cepollaro, MD, Mario Mangeri, MD, Roberto Monaco, MD, Luigi Gennari, MD, and Carlo Gennari, MD Institute of Internal Medicine, University of Siena, Siena, Italy
Abstract In order to evaluate the usefulness of QUS at the phalanx in the diagnosis of osteoporosis and in the prediction of fracture risk in males, we studied 182 subjects (age 61.2 ± 9.4 yr), of which 22 had had a previous nontraumatic bone fracture. In all subjects, bone mineral density (BMD) at the lumbar spine and femur was measured by dual X-ray absorptiometry (DXA). Moreover, QUS parameters [amplitude-dependent speed of sound (AD-SOS), fast-wave amplitude, signal dynamic, and bone transmission time (BTT)] were assessed at the phalanxes using the DBM Sonic 1200 (IGEA). According to World Health Organization (WHO) criteria, all the patients were divided into two groups: 62 osteoporotic and 120 nonosteoporotic. All QUS parameters were significantly lower in osteoporotic than in nonosteoporotic patients. Receiver operating characteristic (ROC) analysis showed a moderate ability of AD-SOS, BTT, and ultrasound bone profile index (UBPI) in distinguishing between healthy and osteoporotic men. Among osteoporotic patients, BMD values were lower in patients with fracture than in those without fracture. AD-SOS and BTT were significantly reduced in men with fracture. Furthermore, in a regression analysis, only BTT and DXA parameters were predictive of fracture. Moreover, performing a multivariate regression analysis BTT entered before BMD at the lumbar spine and at Ward’s triangle. In conclusion, our data show that QUS parameters are reduced in osteoporotic males; however, only BTT was comparable to DXA parameters in the prediction of fracture risk in men. Key Words: Male osteoporosis; quantitative ultrasound; fracture.
Introduction
occurs at menopause; however, some epidemiologic studies have pointed out that osteoporosis is becoming an important health problem in men as well, since up to 20% of symptomatic vertebral fractures and 30% of hip fractures occur in the male population (1). The lifetime risk of symptomatic vertebral and femoral fracture for a 50-yr-old Caucasian man has been estimated to be lower than the lifetime risk for an age-matched woman: 2 and 3%, respectively, as opposed to 11 and 14% (2). On the other hand, men are affected by higher mortality and morbidity after hip fracture, as reported by Seeman (3).
Osteoporosis is characterized by a reduction in bone density associated with skeletal fragility and an increased risk of fracture after minimal trauma. Osteoporosis is generally regarded as a female disease related to the decline in estrogen levels that
Received 11/20/00; Revised 11/30/00; Accepted 02/02/01. Address correspondence to Dr. A. Montagnani, Institute of Internal Medicine, University of Siena, Policlinico Le Scotte, Viale Bracci, 53100 Siena, Italy. E-mail: [email protected]
231
232 Fracture risk is determined by bone structure, connectivity, and bone mass, characteristics that are different in the male and female skeleton. In fact, men reach a higher peak of bone mass at puberty in comparison with women, and they show a higher bone cortical width owing to a continuous apposition during puberty and a lesser bone loss with aging (3). Moreover, bone loss is different between the two sexes: women show a reduction in trabecular number with aging, whereas men lose bone by thinning trabeculae (4) while maintaining a normal bone connectivity. These data could explain the fact that, although bone mineral density (BMD) is predictive for fracture in both men and women (5,6), there is a substantial difference in fracture risk between males and females at similar levels of bone mineral content. This could suggest that bone quality and bone architecture, as well as bone mass, play a critical role in the determination of bone fracture (7,8). For some years quantitative ultrasound (QUS) has been used to assess skeletal status. The majority of the experience derives from studies carried out in the female population and suggests that QUS at several skeletal sites can predict fracture risk in postmenopausal women, independently of BMD measurements (9–11). This suggests that QUS can provide information not only on bone density, but also on qualitative aspects of bone, such as elasticity, connectivity, and homogeneity (12). On the other hand, the information on the usefulness of QUS in men is limited to a few studies. Some studies were performed by measuring the heel (13,14) and the phalanx in healthy men (15,16) or in patients with osteoporosis or other metabolic bone disease (17,18). Therefore, studies aimed at evaluating ultrasound (US) as a predictor of fracture risk in males could be of great interest. The aim of the present study was to evaluate the usefulness of QUS parameters at the phalanx in diagnosing osteoporosis in men and in discriminating between male patients with and without bone fracture.
Materials and Methods Subjects Among 500 male subjects ages 50–75 yr, contacted by mailing for an epidemiologic study, 245 agreed to to under a complete study of skeletal status at the Journal of Clinical Densitometry
Montagnani et al. Centre for Diagnosis and Therapy of Osteoporosis at the University of Siena. Each subject gave informed consent to participate in the study. All men were interviewed on their clinical history, and those with diseases (such as immobilization, chronic renal failure, hyperparathyroidism, hyperthyroidism, or Cushing disease) or under medication (such as corticosteroids, thyroid hormones, anticonvulsants, or antiacids) affecting bone metabolism or who were currently being treated for osteoporosis were excluded from the study. Moreover, subjects who consumed >200 g/wk of alcohol and those who smoked more than 10 cigarettes per day were also not considered for the study.
Measurements One hundred eighty-two subjects participated in the study. Among these, 22 had a history of a previous nontraumatic fracture, prevalently at the spine, whereas in the study were not planned an X-ray study of the spine. In all subjects, BMD was measured at the lumbar spine (g/cm2) and at all sites of the femur (femoral neck BMD, trochanter BMD, intertrochanter BMD, Ward’s triangle BMD, total BMD, g/cm2) using the QDR 4500 (Hologic, Waltham, WA). In patients with fracture at the spine, measurement was performed on vertebrae without fracture. Moreover, in the same subjects, QUS parameters were assessed at the phalanxes with the DBM Sonic 1200 (IGEA srl, Italy). This device measures the amplitude-dependent speed of sound ([AD-SOS], m/s) and three parameters characterizing the graphic trace: bone transmission time ([BTT], µs), which is the time lapse between the moment of the first peak and that when the US signal reaches the velocity of 1700 m/s; fast-wave amplitude ([FWA], pixel), which is the amplitude of the first wave of US signal; and signal dynamic (SDy, mV/µs2), which is obtained from the mathematical synthesis of area under the first two peaks. Finally, the software dedicated to the DBM Sonic device performs a mathematical synthesis of the latter three parameters in a unique parameter, called ultrasound bone profile index (UBPI) (15): UBPI = 1/[1 + exp(–0.0018 SDy – 0.056 FWA – 1.1467 BTT + 3.03)]. A study of precision of QUS parameters was performed by measuring five healthy subjects for five consecutive days. Volume 4, 2001
Bone US at Phalanx in Men with or Without Fracture
233
Table 1 Clinical Characteristics and Densitometric and Ultrasonographic Parameters in Healthy and Osteoporotic Patients Normal Subjects n Age (yr) BMI Lumbar spine BMD (g/cm2) Femoral neck BMD (g/cm2) Trochanter BMD (g/cm2) Intertrochanter BMD (g/cm2) Total femur BMD (g/cm2) Ward’s triangle BMD (g/cm2) AD-SOS (m/s) BTT (µs) UBPI a b
120 59.2 ± 6.3 26.4 ± 4.1 1.035 ± 0.182 0.833 ± 0.145 0.687 ± 0.123 1.005 ± 0.150 0.890 ± 0.090 0.580 ± 0.118 2055.6 ± 71.3 1.73 ± 0.19 0.78 ± 0.14
Osteoporotic Patients 62 57.0 ± 11.9 25.2 ± 3.2 0.803 ± 0.17a 0.664 ± 0.15a 0.602 ± 0.10a 0.956 ± 0.16b 0.812 ± 0.13b 0.447 ± 0.13a 1980.9 ± 107.3a 1.6 ± 0.2a 0.62 ± 0.24a
p < 0.001 between groups. p < 0.01.
According to World Health Organization (WHO) criteria (19) men were considered osteoporotic when they had a BMD value lower than –2.5 as T-score at either the lumbar spine or femur. One hundred twenty subjects were included in the healthy group, whereas 62 subjects were defined as osteoporotic.
Statistical Analyses Statistical analyses were conducted using analysis of variance to compare data between different groups. To evaluate the discriminative power of densitometric and QUS parameters, we calculated receiver operating characteristics (ROC) curves, while a logistic regression analysis was performed to calculate the odds ratios (ORs) of densitometric and QUS parameters in predicting fracture in males. All statistical analyses were carried out using SPSS software.
Results Precision Study The coefficients of variation (CVs) of US parameters showed different results: AD-SOS and BTT had a CV of 0.5 and 0.8%, respectively; UBPI showed a precision of 2.2%, whereas both FWA and SDy evidenced Journal of Clinical Densitometry
a poor precision, for which reason the last two parameters were not considered in the study.
Diagnosis of Osteoporosis The mean ± standard deviation (SD) values of BMD at lumbar femoral sites and of QUS parameters at the phalanxes for the healthy and osteoporotic groups are reported in Table 1. All parameters were significantly different between the two groups. The ROC curves relative to healthy and osteoporotic patients are plotted in Fig. 1. QUS parameters at the phalanxes showed similar ability to diagnose osteoporosis, with the areas under fitted curves of 0.63 for AD-SOS, 0.62 for BTT, and 0.61 for UBPI.
Prediction of Fracture Among the osteoporotic patients, 22 subjects had a history of previous low-trauma fracture, and they showed BMD values lower than those of osteoporotic patients without fracture (Table 2). In patients with osteoporosis and bone fracture, AD-SOS and BTT were significantly (p < 0.05 and p < 0.001, respectively) lower than in patients without fracture, whereas UBPI did not show a significant difference between the two groups (Table 2). The areas (±standard errors [SEs]) under ROC curves of dual X-ray Volume 4, 2001
234
Montagnani et al.
Fig. 1. ROC curve analysis for QUS parameters in diagnosis of osteoporosis in men. The discriminatory ability of AD-SOS, BTT, and UBPI is expressed by the area under ROC curves (ROC AUC).
Table 2 Clinical Characteristics and Densitometric and Ultrasonographic Parameters in Osteoporotic Patients with and Without Low Traumatic Fracture Osteoporotic patients with fracture n Age (yr) BMI Lumbar spine BMD (g/cm2) Femoral neck BMD (g/cm2) Trochanter BMD (g/cm2) Intertrochanter BMD (g/cm2) Total femur BMD (g/cm2) Ward’s triangle BMD (g/cm2) AD-SOS (m/s) BTT (µs) UBPI
22 58.7 ± 9.7 24.5 ± 2.9 0.690 ± 0.13 0.590 ± 0.07 0.551 ± 0.08 0.858 ± 0.10 0.729 ± 0.07 0.358 ± 0.05 1930.3 ± 108.9 1.49 ± 0.19 0.55 ± 0.18
Osteoporotic patients without fracture 40 59.5 ± 10.1 25.0 ± 3.5 0.864 ± 0.16a 0.705 ± 0.16b 0.629 ± 0.11b 0.989 ± 0.17c 0.852 ± 0.13c 0.495 ± 0.14c 1995.7 ± 109.7b 1.72 ± 0.20a 0.60 ± 0.25
a
p < 0.001 between groups. p < 0.05. c p < 0.01. b
Journal of Clinical Densitometry
Volume 4, 2001
Bone US at Phalanx in Men with or Without Fracture
235
Table 3 Values of Area Under ROC Curves for QUS Parameters in Discriminating Between Fractured and Unfractured Male Osteoporotic Patients Area under the curve
SE
0.70 0.60 0.74 0.70 0.59 0.71 0.79 0.62
0.084 0.093 0.081 0.085 0.094 0.084 0.062 0.075
Lumbar spine BMD Femoral neck BMD Intertrochanter BMD Ward’s triangle BMD Trochanter BMD Total femur BMD BTT AD-SOS
Table 4 Logistic Regression Measuring Effect of Densitometry and US Parameters on Fracture Riska
Lumbar spine BMD Femoral neck BMD Intertrochanter BMD Ward’s triangle BMD Total femur BMD BTT
β (SE)
β′ = β* (–SD)
OR (95% CI)
–10.8 (3.66) –11.4 (4.95) –8.3 (3.32) –14.6 (5.6) –11.9 (4.6) –4.7 (1.6)
1.72 1.00 1.24 1.90 1.43 1.08
5.61 (1.78–17.67) 5.52 (1.29–23.5) 3.50 (1.64–9.23) 6.67 (2.45–109.4) 4.17 (1.41–12.3) 2.94 (1.44–5.98)
a
Age and BMI were included in the model as independent variables. TrochanterBMD and AD-SOS did not enter into the model.
absorptiometry (DXA) and QUS parameters (ADSOS and BTT) in osteoporotic patients with or without fracture are reported in Table 3. ORs for bone fracture were calculated by performing a logistic regression including all densitometric and QUS parameters other than age and BMI as independent variables. Table 4 gives the ORs resulting from a simple logistic regression for each densitometric parameter and BTT. Both trochanter BMD and AD-SOS did not enter into the logistic regression model when adjusted for age and body mass index (BMI). In a stepwise logistic regression, still including clinical characteristics, DXA and QUS parameters pointed out that only BTT, lumbar spine BMD and Ward’s tringle BMD entered into the model with ORs of 7.1 (95% confidence interval [CI]: 1.14–62.6), 10.8 (95% CI: 1.2–81.4), and 18.6 (95% CI: 1.02–330.2), respectively. Journal of Clinical Densitometry
When the WHO criteria were applied to this population to identify the T-score value for osteoporosis, we found a T-score of –2.9 for AD-SOS, –1.0 for BTT, and –4.2 for UBPI. Sixty-three percent of fractures were below the AD-SOS cutoff, whereas below the BTT and the UBPI cutoff we found 68% of fractures.
Discussion QUS seems to show a moderate ability to discriminate between healthy and osteoporotic male subjects, when the diagnosis of osteoporosis is based on WHO criteria. This finding is apparently in contrast to that previously reported in many studies carried out in the female population (9,10,20). The different ability of QUS parameters in female and male patients could be explained by the differences in bone structure shown Volume 4, 2001
236 by the two sexes; men with respect to women show a major thickness of cortical bone, which in healthy subjects remains constant throughout adult life (3). This fact could explain the lower ability of AD-SOS to predict bone fracture in men compared with women (20); in fact, the major cortical thickness could negatively influence AD-SOS by reducing the signal amplitude (15), which makes changes in AD-SOS less sensitive to predict fracture. Therefore, the difference in bone structure between the two genders could suggest that different criteria should be applied for men in diagnosing osteoporosis; however, this issue still needs to be resolved. Our data suggest that the decrease in BMD, at the both spine and femur, was clearly associated with an increase in fracture risk. After adjustment for age and BMI, ORs for a decrease of 1 SD of BMD ranged from 3.5 to 6.7, depending on the measurement site. Our data are similar to those reported by Legrand et al. (7), who found a two or threefold increase in fracture risk using lumbar spine BMD or femural parameters, respectively. The fact that our patients were osteoporotic and not osteopenic could explain why our study’s ORs are slightly higher than those found by Legrand et al. (7). QUS parameters showed a modest ability as predictors of fracture in men, apart from BTT, that resulted similar to DXA, with an OR of 2.9. BTT is one of the parameters that characterizes the graphic trace, and it has already been used by Machado et al (21) in monitoring bisphosphonate treatment in patients with hyperparathyroidism, as well as by us to evaluate the reference curve in healthy men (15). BTT was shown to be comparable to DXA parameters in discriminating between patients with and without fracture, and when we considered all parameters in a stepwise logistic regression analysis, we found that BTT was independently and more closely related to fracture than lumbar spine BMD and Ward’s triangles BMD, which entered as the second and third in the statistical model. This interesting result could be justified by the fact that BTT is a parameter mostly related to cortical thickness of the phalanx, which plays a crucial role in male osteoporosis and in the incidence of osteoporotic fracture in men, as reported by Seeman (3). In fact, osteoporotic men lose mainly cortical bone and trabeculae thickness, which is different from women, who show a reduction in trabeculae number and connectivity.
Journal of Clinical Densitometry
Montagnani et al. The result of BTT as a predictor of fracture risk independently of DXA and clinical parameters could suggest a useful combination of BTT with lumbar or femoral BMD to enhance the evaluation of fracture risk. However, the sample size of the present study was too small and there were too few fracture patients to perform such a type of analysis. Moreover, an additional limit of the present study is that vertebral fractures were identified on the basis of patient reports; therefore, these data needed to be confirmed by studies using more stringent methods in establishing osteoporotic fractures. In conclusion, our data showed that in men QUS at the phalanx had a moderate ability to diagnose osteoporosis, defined by WHO criteria. Regarding DXA, our findings were comparable with those reported in the literature. Among QUS parameters, AD-SOS was moderately predictive of fracture in men, in contrast to data reported for women (20), probably because ADSOS measurement in men is limited by the major thickness of male phalanx cortical bone; on the other hand, BTT was shown to be a strong predictor of fracture risk with a great ability to discriminate between patients with and without fracture. Since lumbar spine BMD, Ward’s triangle BMD, and BTT were independently related to fracture risk, we can speculate that a combination of such parameters could increase the prediction of fracture. However, studies with many subjects will be necessary to resolve this issue.
References 1. Eastell R, Boyle IT, Compston J, et al. 1998 Management of male osteoporosis: report of the UK Consensus Group. QJM 91:71–92. 2. Oden A, Dawson A, Dere W, Johenel O, Jonson B, Kanis JA. 1998 Lifetime risk of hip fracture is underestimated. Osteoporos Int 8:599–603. 3. Seeman E. 1999 Osteoporosis in men. Osteoporos Int (Suppl. 2) S97–S110. 4. Francis RM, Peacock M, Marshall DH, Horsman A, Aaron JE. 1989 Spinal osteoporosis in men. Calcif Tissue Int 62:185–188. 5. Melton LJ III, Atkinson EJ, O’Connor MK, O’Fallon WM, Riggs BL. 1998 Bone density and fracture risk in men. J Bone Miner Res 13:1915–1923. 6. De Laet C, van Hout BA, Burger H, Weel AEAM, Hofman A, Pols HAP. 1998 Hip fracture prediction in elderly men and women: validation in the Rotterdam study. J Bone Miner Res 13:1587–1593.
Volume 4, 2001
Bone US at Phalanx in Men with or Without Fracture 7. Legrand E, Chappard D, Pascaretti C, et al. 1999 Bone mineral density and vertebral fractures in men. Osteoporos Int 10:265–270. 8. Legrand E, Chappard D, Pascaretti C, et al. 2000 Trabecular bone microarchitecture, bone mineral density and vertebral fractures in male osteoporosis. J Bone Miner Res 15:13–19. 9. Hans D, Dargent-Molina P, Schott AM, et al. for the EP-DOS prospective study group. 1996 Ultrasonographic heel measurements to predict hip fracture in elderly women: the EPIDOS prospective study. Lancet 348:511–514. 10. Bauer DC, Gluer CC, Cauley JA, et al. 1997 Broadband ultrasound attenuation predicts fractures strongly and independently of densitometry in older women: a prospective study. Study of Osteoporotic Fractures Research Group. Arch Intern Med 157(6):629–634. 11. Cepollaro C, Gonnelli S, Pondrelli C, et al. 1997 The combined use of ultrasound and densitometry in the prediction of vertebral fracture. Br J Radiol 70:691–696. 12. Gluer CC, for The international quantitative ultrasound consensus group. 1997 Quantitative ultrasound techniques for the assessment of osteoporosis: expert agreement on current status. J Bone Miner Res 12:1280–1288. 13. Van Daele PLA, Burger H, Algra D, et al. 1994 Age-associated changes in ultrasound measurements of the calcaneus in men and women: the Rotterdam Study. J Bone Miner Res 9:1751–1757.
Journal of Clinical Densitometry
237 14. Cepollaro C, Agnusdei D, Gonnelli S, et al. 1995 Ultrasonographic assessment of bone in normal Italian males and females. Br J Radiol 68:910–914. 15. Montagnani A, Gonnelli S, Cepollaro C, et al. 2000 Quantitative ultrasound at phalanxes in healthy Italian males. Osteoporos Int 11:499–504. 16. Soballa T, Schlegel J, Cadossi R, Isani R, Heilmann P, Ziegler R, Wuster C. 1998 Osteosonography of the phalanges of men. Med Klin 93:131–136. 17. Pluskiewicz W, Drozdzowska B. 1999 Ultrasound measurements at the calcaneus in men: differences between healthy and fractured persons and the influence of age and anthropometric features on ultrasound parameters. Osteoporos Int 10:47–51. 18. Gonnelli S, Montagnani A, Cepollaro C, et al. 2000 Quantitative ultrasound and bone mineral density in patients with primary hyperparathyroidism before and after surgical treatment. Osteoporos Int 3:255–260. 19. Kanis JA. 1994 Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: synopsis of a WHO report. Osteoporos Int 4:368–381. 20. Wuster C, Albanese C, Boschitsch E, et al. 2000 Phalangeal osteosonogrammetry (PhOS): age related changes, diagnostic sensitivity and discrimination power. J Bone Miner Res. 15:1603–1614. 21. Machado ABC, Ingle BM, Eastell R. 1999 Monitoring alendronate therapy with quantitative ultrasound (QUS) and dual X-ray absorptiometry (DXA). J Bone Miner Res 14:S526.
Volume 4, 2001
Original Article
Usefulness of Bone Quantitative Ultrasound in Management of Osteoporosis in Men Andrea Montagnani, MD, Stefano Gonnelli, MD, Chiara Cepollaro, MD, Mario Mangeri, MD, Roberto Monaco, MD, Luigi Gennari, MD, and Carlo Gennari, MD Institute of Internal Medicine, University of Siena, Siena, Italy
Abstract In order to evaluate the usefulness of QUS at the phalanx in the diagnosis of osteoporosis and in the prediction of fracture risk in males, we studied 182 subjects (age 61.2 ± 9.4 yr), of which 22 had had a previous nontraumatic bone fracture. In all subjects, bone mineral density (BMD) at the lumbar spine and femur was measured by dual X-ray absorptiometry (DXA). Moreover, QUS parameters [amplitude-dependent speed of sound (AD-SOS), fast-wave amplitude, signal dynamic, and bone transmission time (BTT)] were assessed at the phalanxes using the DBM Sonic 1200 (IGEA). According to World Health Organization (WHO) criteria, all the patients were divided into two groups: 62 osteoporotic and 120 nonosteoporotic. All QUS parameters were significantly lower in osteoporotic than in nonosteoporotic patients. Receiver operating characteristic (ROC) analysis showed a moderate ability of AD-SOS, BTT, and ultrasound bone profile index (UBPI) in distinguishing between healthy and osteoporotic men. Among osteoporotic patients, BMD values were lower in patients with fracture than in those without fracture. AD-SOS and BTT were significantly reduced in men with fracture. Furthermore, in a regression analysis, only BTT and DXA parameters were predictive of fracture. Moreover, performing a multivariate regression analysis BTT entered before BMD at the lumbar spine and at Ward’s triangle. In conclusion, our data show that QUS parameters are reduced in osteoporotic males; however, only BTT was comparable to DXA parameters in the prediction of fracture risk in men. Key Words: Male osteoporosis; quantitative ultrasound; fracture.
Introduction
occurs at menopause; however, some epidemiologic studies have pointed out that osteoporosis is becoming an important health problem in men as well, since up to 20% of symptomatic vertebral fractures and 30% of hip fractures occur in the male population (1). The lifetime risk of symptomatic vertebral and femoral fracture for a 50-yr-old Caucasian man has been estimated to be lower than the lifetime risk for an age-matched woman: 2 and 3%, respectively, as opposed to 11 and 14% (2). On the other hand, men are affected by higher mortality and morbidity after hip fracture, as reported by Seeman (3).
Osteoporosis is characterized by a reduction in bone density associated with skeletal fragility and an increased risk of fracture after minimal trauma. Osteoporosis is generally regarded as a female disease related to the decline in estrogen levels that
Received 11/20/00; Revised 11/30/00; Accepted 02/02/01. Address correspondence to Dr. A. Montagnani, Institute of Internal Medicine, University of Siena, Policlinico Le Scotte, Viale Bracci, 53100 Siena, Italy. E-mail: [email protected]
231
232 Fracture risk is determined by bone structure, connectivity, and bone mass, characteristics that are different in the male and female skeleton. In fact, men reach a higher peak of bone mass at puberty in comparison with women, and they show a higher bone cortical width owing to a continuous apposition during puberty and a lesser bone loss with aging (3). Moreover, bone loss is different between the two sexes: women show a reduction in trabecular number with aging, whereas men lose bone by thinning trabeculae (4) while maintaining a normal bone connectivity. These data could explain the fact that, although bone mineral density (BMD) is predictive for fracture in both men and women (5,6), there is a substantial difference in fracture risk between males and females at similar levels of bone mineral content. This could suggest that bone quality and bone architecture, as well as bone mass, play a critical role in the determination of bone fracture (7,8). For some years quantitative ultrasound (QUS) has been used to assess skeletal status. The majority of the experience derives from studies carried out in the female population and suggests that QUS at several skeletal sites can predict fracture risk in postmenopausal women, independently of BMD measurements (9–11). This suggests that QUS can provide information not only on bone density, but also on qualitative aspects of bone, such as elasticity, connectivity, and homogeneity (12). On the other hand, the information on the usefulness of QUS in men is limited to a few studies. Some studies were performed by measuring the heel (13,14) and the phalanx in healthy men (15,16) or in patients with osteoporosis or other metabolic bone disease (17,18). Therefore, studies aimed at evaluating ultrasound (US) as a predictor of fracture risk in males could be of great interest. The aim of the present study was to evaluate the usefulness of QUS parameters at the phalanx in diagnosing osteoporosis in men and in discriminating between male patients with and without bone fracture.
Materials and Methods Subjects Among 500 male subjects ages 50–75 yr, contacted by mailing for an epidemiologic study, 245 agreed to to under a complete study of skeletal status at the Journal of Clinical Densitometry
Montagnani et al. Centre for Diagnosis and Therapy of Osteoporosis at the University of Siena. Each subject gave informed consent to participate in the study. All men were interviewed on their clinical history, and those with diseases (such as immobilization, chronic renal failure, hyperparathyroidism, hyperthyroidism, or Cushing disease) or under medication (such as corticosteroids, thyroid hormones, anticonvulsants, or antiacids) affecting bone metabolism or who were currently being treated for osteoporosis were excluded from the study. Moreover, subjects who consumed >200 g/wk of alcohol and those who smoked more than 10 cigarettes per day were also not considered for the study.
Measurements One hundred eighty-two subjects participated in the study. Among these, 22 had a history of a previous nontraumatic fracture, prevalently at the spine, whereas in the study were not planned an X-ray study of the spine. In all subjects, BMD was measured at the lumbar spine (g/cm2) and at all sites of the femur (femoral neck BMD, trochanter BMD, intertrochanter BMD, Ward’s triangle BMD, total BMD, g/cm2) using the QDR 4500 (Hologic, Waltham, WA). In patients with fracture at the spine, measurement was performed on vertebrae without fracture. Moreover, in the same subjects, QUS parameters were assessed at the phalanxes with the DBM Sonic 1200 (IGEA srl, Italy). This device measures the amplitude-dependent speed of sound ([AD-SOS], m/s) and three parameters characterizing the graphic trace: bone transmission time ([BTT], µs), which is the time lapse between the moment of the first peak and that when the US signal reaches the velocity of 1700 m/s; fast-wave amplitude ([FWA], pixel), which is the amplitude of the first wave of US signal; and signal dynamic (SDy, mV/µs2), which is obtained from the mathematical synthesis of area under the first two peaks. Finally, the software dedicated to the DBM Sonic device performs a mathematical synthesis of the latter three parameters in a unique parameter, called ultrasound bone profile index (UBPI) (15): UBPI = 1/[1 + exp(–0.0018 SDy – 0.056 FWA – 1.1467 BTT + 3.03)]. A study of precision of QUS parameters was performed by measuring five healthy subjects for five consecutive days. Volume 4, 2001
Bone US at Phalanx in Men with or Without Fracture
233
Table 1 Clinical Characteristics and Densitometric and Ultrasonographic Parameters in Healthy and Osteoporotic Patients Normal Subjects n Age (yr) BMI Lumbar spine BMD (g/cm2) Femoral neck BMD (g/cm2) Trochanter BMD (g/cm2) Intertrochanter BMD (g/cm2) Total femur BMD (g/cm2) Ward’s triangle BMD (g/cm2) AD-SOS (m/s) BTT (µs) UBPI a b
120 59.2 ± 6.3 26.4 ± 4.1 1.035 ± 0.182 0.833 ± 0.145 0.687 ± 0.123 1.005 ± 0.150 0.890 ± 0.090 0.580 ± 0.118 2055.6 ± 71.3 1.73 ± 0.19 0.78 ± 0.14
Osteoporotic Patients 62 57.0 ± 11.9 25.2 ± 3.2 0.803 ± 0.17a 0.664 ± 0.15a 0.602 ± 0.10a 0.956 ± 0.16b 0.812 ± 0.13b 0.447 ± 0.13a 1980.9 ± 107.3a 1.6 ± 0.2a 0.62 ± 0.24a
p < 0.001 between groups. p < 0.01.
According to World Health Organization (WHO) criteria (19) men were considered osteoporotic when they had a BMD value lower than –2.5 as T-score at either the lumbar spine or femur. One hundred twenty subjects were included in the healthy group, whereas 62 subjects were defined as osteoporotic.
Statistical Analyses Statistical analyses were conducted using analysis of variance to compare data between different groups. To evaluate the discriminative power of densitometric and QUS parameters, we calculated receiver operating characteristics (ROC) curves, while a logistic regression analysis was performed to calculate the odds ratios (ORs) of densitometric and QUS parameters in predicting fracture in males. All statistical analyses were carried out using SPSS software.
Results Precision Study The coefficients of variation (CVs) of US parameters showed different results: AD-SOS and BTT had a CV of 0.5 and 0.8%, respectively; UBPI showed a precision of 2.2%, whereas both FWA and SDy evidenced Journal of Clinical Densitometry
a poor precision, for which reason the last two parameters were not considered in the study.
Diagnosis of Osteoporosis The mean ± standard deviation (SD) values of BMD at lumbar femoral sites and of QUS parameters at the phalanxes for the healthy and osteoporotic groups are reported in Table 1. All parameters were significantly different between the two groups. The ROC curves relative to healthy and osteoporotic patients are plotted in Fig. 1. QUS parameters at the phalanxes showed similar ability to diagnose osteoporosis, with the areas under fitted curves of 0.63 for AD-SOS, 0.62 for BTT, and 0.61 for UBPI.
Prediction of Fracture Among the osteoporotic patients, 22 subjects had a history of previous low-trauma fracture, and they showed BMD values lower than those of osteoporotic patients without fracture (Table 2). In patients with osteoporosis and bone fracture, AD-SOS and BTT were significantly (p < 0.05 and p < 0.001, respectively) lower than in patients without fracture, whereas UBPI did not show a significant difference between the two groups (Table 2). The areas (±standard errors [SEs]) under ROC curves of dual X-ray Volume 4, 2001
234
Montagnani et al.
Fig. 1. ROC curve analysis for QUS parameters in diagnosis of osteoporosis in men. The discriminatory ability of AD-SOS, BTT, and UBPI is expressed by the area under ROC curves (ROC AUC).
Table 2 Clinical Characteristics and Densitometric and Ultrasonographic Parameters in Osteoporotic Patients with and Without Low Traumatic Fracture Osteoporotic patients with fracture n Age (yr) BMI Lumbar spine BMD (g/cm2) Femoral neck BMD (g/cm2) Trochanter BMD (g/cm2) Intertrochanter BMD (g/cm2) Total femur BMD (g/cm2) Ward’s triangle BMD (g/cm2) AD-SOS (m/s) BTT (µs) UBPI
22 58.7 ± 9.7 24.5 ± 2.9 0.690 ± 0.13 0.590 ± 0.07 0.551 ± 0.08 0.858 ± 0.10 0.729 ± 0.07 0.358 ± 0.05 1930.3 ± 108.9 1.49 ± 0.19 0.55 ± 0.18
Osteoporotic patients without fracture 40 59.5 ± 10.1 25.0 ± 3.5 0.864 ± 0.16a 0.705 ± 0.16b 0.629 ± 0.11b 0.989 ± 0.17c 0.852 ± 0.13c 0.495 ± 0.14c 1995.7 ± 109.7b 1.72 ± 0.20a 0.60 ± 0.25
a
p < 0.001 between groups. p < 0.05. c p < 0.01. b
Journal of Clinical Densitometry
Volume 4, 2001
Bone US at Phalanx in Men with or Without Fracture
235
Table 3 Values of Area Under ROC Curves for QUS Parameters in Discriminating Between Fractured and Unfractured Male Osteoporotic Patients Area under the curve
SE
0.70 0.60 0.74 0.70 0.59 0.71 0.79 0.62
0.084 0.093 0.081 0.085 0.094 0.084 0.062 0.075
Lumbar spine BMD Femoral neck BMD Intertrochanter BMD Ward’s triangle BMD Trochanter BMD Total femur BMD BTT AD-SOS
Table 4 Logistic Regression Measuring Effect of Densitometry and US Parameters on Fracture Riska
Lumbar spine BMD Femoral neck BMD Intertrochanter BMD Ward’s triangle BMD Total femur BMD BTT
β (SE)
β′ = β* (–SD)
OR (95% CI)
–10.8 (3.66) –11.4 (4.95) –8.3 (3.32) –14.6 (5.6) –11.9 (4.6) –4.7 (1.6)
1.72 1.00 1.24 1.90 1.43 1.08
5.61 (1.78–17.67) 5.52 (1.29–23.5) 3.50 (1.64–9.23) 6.67 (2.45–109.4) 4.17 (1.41–12.3) 2.94 (1.44–5.98)
a
Age and BMI were included in the model as independent variables. TrochanterBMD and AD-SOS did not enter into the model.
absorptiometry (DXA) and QUS parameters (ADSOS and BTT) in osteoporotic patients with or without fracture are reported in Table 3. ORs for bone fracture were calculated by performing a logistic regression including all densitometric and QUS parameters other than age and BMI as independent variables. Table 4 gives the ORs resulting from a simple logistic regression for each densitometric parameter and BTT. Both trochanter BMD and AD-SOS did not enter into the logistic regression model when adjusted for age and body mass index (BMI). In a stepwise logistic regression, still including clinical characteristics, DXA and QUS parameters pointed out that only BTT, lumbar spine BMD and Ward’s tringle BMD entered into the model with ORs of 7.1 (95% confidence interval [CI]: 1.14–62.6), 10.8 (95% CI: 1.2–81.4), and 18.6 (95% CI: 1.02–330.2), respectively. Journal of Clinical Densitometry
When the WHO criteria were applied to this population to identify the T-score value for osteoporosis, we found a T-score of –2.9 for AD-SOS, –1.0 for BTT, and –4.2 for UBPI. Sixty-three percent of fractures were below the AD-SOS cutoff, whereas below the BTT and the UBPI cutoff we found 68% of fractures.
Discussion QUS seems to show a moderate ability to discriminate between healthy and osteoporotic male subjects, when the diagnosis of osteoporosis is based on WHO criteria. This finding is apparently in contrast to that previously reported in many studies carried out in the female population (9,10,20). The different ability of QUS parameters in female and male patients could be explained by the differences in bone structure shown Volume 4, 2001
236 by the two sexes; men with respect to women show a major thickness of cortical bone, which in healthy subjects remains constant throughout adult life (3). This fact could explain the lower ability of AD-SOS to predict bone fracture in men compared with women (20); in fact, the major cortical thickness could negatively influence AD-SOS by reducing the signal amplitude (15), which makes changes in AD-SOS less sensitive to predict fracture. Therefore, the difference in bone structure between the two genders could suggest that different criteria should be applied for men in diagnosing osteoporosis; however, this issue still needs to be resolved. Our data suggest that the decrease in BMD, at the both spine and femur, was clearly associated with an increase in fracture risk. After adjustment for age and BMI, ORs for a decrease of 1 SD of BMD ranged from 3.5 to 6.7, depending on the measurement site. Our data are similar to those reported by Legrand et al. (7), who found a two or threefold increase in fracture risk using lumbar spine BMD or femural parameters, respectively. The fact that our patients were osteoporotic and not osteopenic could explain why our study’s ORs are slightly higher than those found by Legrand et al. (7). QUS parameters showed a modest ability as predictors of fracture in men, apart from BTT, that resulted similar to DXA, with an OR of 2.9. BTT is one of the parameters that characterizes the graphic trace, and it has already been used by Machado et al (21) in monitoring bisphosphonate treatment in patients with hyperparathyroidism, as well as by us to evaluate the reference curve in healthy men (15). BTT was shown to be comparable to DXA parameters in discriminating between patients with and without fracture, and when we considered all parameters in a stepwise logistic regression analysis, we found that BTT was independently and more closely related to fracture than lumbar spine BMD and Ward’s triangles BMD, which entered as the second and third in the statistical model. This interesting result could be justified by the fact that BTT is a parameter mostly related to cortical thickness of the phalanx, which plays a crucial role in male osteoporosis and in the incidence of osteoporotic fracture in men, as reported by Seeman (3). In fact, osteoporotic men lose mainly cortical bone and trabeculae thickness, which is different from women, who show a reduction in trabeculae number and connectivity.
Journal of Clinical Densitometry
Montagnani et al. The result of BTT as a predictor of fracture risk independently of DXA and clinical parameters could suggest a useful combination of BTT with lumbar or femoral BMD to enhance the evaluation of fracture risk. However, the sample size of the present study was too small and there were too few fracture patients to perform such a type of analysis. Moreover, an additional limit of the present study is that vertebral fractures were identified on the basis of patient reports; therefore, these data needed to be confirmed by studies using more stringent methods in establishing osteoporotic fractures. In conclusion, our data showed that in men QUS at the phalanx had a moderate ability to diagnose osteoporosis, defined by WHO criteria. Regarding DXA, our findings were comparable with those reported in the literature. Among QUS parameters, AD-SOS was moderately predictive of fracture in men, in contrast to data reported for women (20), probably because ADSOS measurement in men is limited by the major thickness of male phalanx cortical bone; on the other hand, BTT was shown to be a strong predictor of fracture risk with a great ability to discriminate between patients with and without fracture. Since lumbar spine BMD, Ward’s triangle BMD, and BTT were independently related to fracture risk, we can speculate that a combination of such parameters could increase the prediction of fracture. However, studies with many subjects will be necessary to resolve this issue.
References 1. Eastell R, Boyle IT, Compston J, et al. 1998 Management of male osteoporosis: report of the UK Consensus Group. QJM 91:71–92. 2. Oden A, Dawson A, Dere W, Johenel O, Jonson B, Kanis JA. 1998 Lifetime risk of hip fracture is underestimated. Osteoporos Int 8:599–603. 3. Seeman E. 1999 Osteoporosis in men. Osteoporos Int (Suppl. 2) S97–S110. 4. Francis RM, Peacock M, Marshall DH, Horsman A, Aaron JE. 1989 Spinal osteoporosis in men. Calcif Tissue Int 62:185–188. 5. Melton LJ III, Atkinson EJ, O’Connor MK, O’Fallon WM, Riggs BL. 1998 Bone density and fracture risk in men. J Bone Miner Res 13:1915–1923. 6. De Laet C, van Hout BA, Burger H, Weel AEAM, Hofman A, Pols HAP. 1998 Hip fracture prediction in elderly men and women: validation in the Rotterdam study. J Bone Miner Res 13:1587–1593.
Volume 4, 2001
Bone US at Phalanx in Men with or Without Fracture 7. Legrand E, Chappard D, Pascaretti C, et al. 1999 Bone mineral density and vertebral fractures in men. Osteoporos Int 10:265–270. 8. Legrand E, Chappard D, Pascaretti C, et al. 2000 Trabecular bone microarchitecture, bone mineral density and vertebral fractures in male osteoporosis. J Bone Miner Res 15:13–19. 9. Hans D, Dargent-Molina P, Schott AM, et al. for the EP-DOS prospective study group. 1996 Ultrasonographic heel measurements to predict hip fracture in elderly women: the EPIDOS prospective study. Lancet 348:511–514. 10. Bauer DC, Gluer CC, Cauley JA, et al. 1997 Broadband ultrasound attenuation predicts fractures strongly and independently of densitometry in older women: a prospective study. Study of Osteoporotic Fractures Research Group. Arch Intern Med 157(6):629–634. 11. Cepollaro C, Gonnelli S, Pondrelli C, et al. 1997 The combined use of ultrasound and densitometry in the prediction of vertebral fracture. Br J Radiol 70:691–696. 12. Gluer CC, for The international quantitative ultrasound consensus group. 1997 Quantitative ultrasound techniques for the assessment of osteoporosis: expert agreement on current status. J Bone Miner Res 12:1280–1288. 13. Van Daele PLA, Burger H, Algra D, et al. 1994 Age-associated changes in ultrasound measurements of the calcaneus in men and women: the Rotterdam Study. J Bone Miner Res 9:1751–1757.
Journal of Clinical Densitometry
237 14. Cepollaro C, Agnusdei D, Gonnelli S, et al. 1995 Ultrasonographic assessment of bone in normal Italian males and females. Br J Radiol 68:910–914. 15. Montagnani A, Gonnelli S, Cepollaro C, et al. 2000 Quantitative ultrasound at phalanxes in healthy Italian males. Osteoporos Int 11:499–504. 16. Soballa T, Schlegel J, Cadossi R, Isani R, Heilmann P, Ziegler R, Wuster C. 1998 Osteosonography of the phalanges of men. Med Klin 93:131–136. 17. Pluskiewicz W, Drozdzowska B. 1999 Ultrasound measurements at the calcaneus in men: differences between healthy and fractured persons and the influence of age and anthropometric features on ultrasound parameters. Osteoporos Int 10:47–51. 18. Gonnelli S, Montagnani A, Cepollaro C, et al. 2000 Quantitative ultrasound and bone mineral density in patients with primary hyperparathyroidism before and after surgical treatment. Osteoporos Int 3:255–260. 19. Kanis JA. 1994 Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: synopsis of a WHO report. Osteoporos Int 4:368–381. 20. Wuster C, Albanese C, Boschitsch E, et al. 2000 Phalangeal osteosonogrammetry (PhOS): age related changes, diagnostic sensitivity and discrimination power. J Bone Miner Res. 15:1603–1614. 21. Machado ABC, Ingle BM, Eastell R. 1999 Monitoring alendronate therapy with quantitative ultrasound (QUS) and dual X-ray absorptiometry (DXA). J Bone Miner Res 14:S526.
Volume 4, 2001