Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health, vol. 16, no. 3, 341e346, 2013 Ó Copyright 2013 by The International Society for Clinical Densitometry 1094-6950/16:341e346/$36.00 http://dx.doi.org/10.1016/j.jocd.2012.07.006
Original Article
Three-Year Prospective Study on Fracture Risk in Postmenopausal Women by Quantitative Ultrasound at the Phalanges Giuseppe Guglielmi,*,1,2 Maurizio Rossini,3 Mario Guido Nicolosi,4 Alessandro Ragno,5 Giovanni Lentini,6 and Francesca de Terlizzi7 1
Department of Radiology, University of Foggia, Foggia, Italy; 2Department of Radiology, Scientific Institute ‘‘Casa Sollievo della Sofferenza’’ Hospital, Foggia, Italy; 3Department of Rheumatology, Azienda Ospedaliera Universitaria Integrata di Verona, Verona, Italy; 4Department of Gynecology and Obstetrics, Divisione Ospedaliera B, S. Anna Hospital, Torino, Italy; 5Department of Internal Medicine, Ospedale ‘‘Regina Apostolorum,’’ Albano Laziale, Italy; 6Department of Gynecology and Obstetrics, Buccheri La Ferla FBF Hospital, Palermo, Italy; and 7IGEA SpA, Clinical Biophysics Laboratory, Modena, Italy
Abstract The purpose of this study was the calculation of fracture risk in a prospective study on postmenopausal women by quantitative ultrasound (QUS) at the phalanges. A total of 2341 postmenopausal women were recruited in 5 centers in Italy during 2006 and 2007 for QUS measurement during a screening program for osteoporosis. Two ultrasound parameters were collected: amplitude-dependent speed of sound (AD-SoS) and ultrasound bone profile index (UBPI). Women were then recontacted in 2010 and were asked about fracture occurrence during the period since previous QUS measurement. Data about new fracture occurred in this period, site and cause of fracture were requested. Two thousand two hundred eleven women were successfully recontacted. Mean age of the recruited women was 60.9 10.0 yr, mean age at menopause was 49.3 4.4 yr, mean body mass index (BMI) was 26.5 4.6 kg/m2. A total number of 108 new major osteoporotic fractures occurred during the 3-yr period, of which 23 are hip fractures, 51 are vertebral fractures. Relative risk (RR) per standard deviation (SD) decrease for major fractures was 1.77 (confidence interval [CI]: 1.59e1.97) for AD-SoS and 2.06 (CI: 1.78e2.37) for UBPI. When corrected for age, BMI, age at menopause, the RRs are still significant and equal to 1.44 (CI: 1.26e1.65) for AD-SoS and 1.67 (CI: 1.39e2.00) for UBPI. RR for vertebral fractures was 1.63 (CI: 1.41e1.88) for AD-SoS and 1.73 (CI: 1.44e2.08) for UBPI. RR for hip fractures was 1.92 (CI: 1.55e2.37) for AD-SoS and 2.68 (CI: 1.86e3.86) for UBPI. Ultrasound parameters AD-SoS and UBPI are able to significantly predict future major fractures in a prospective cohort of more than 2000 postmenopausal women. Key Words: Bone quality; fracture risk; osteoporosis; quantitative ultrasound; radiation-free technique.
postmenopausal women older than 50 yr is 17%, and after this period, the risk increases exponentially (2). However, when osteoporosis is diagnosed, treatments prevent hip fractures by up to 60% (3,4) and nonvertebral fractures up to 50% (5). Although it is possible to observe in the last years a progressive increase in the awareness of the disease, the large majority of the population at potential risk of osteoporosis has never been diagnosed (6). For this purpose, it is critically important to have simple and noninvasive techniques to identify women at higher risk of osteoporotic fracture.
Introduction It is known that about 50% of postmenopausal white women will suffer from an osteoporotic fracture during their life (1). Furthermore, life-time risk of hip fracture in
Received 04/20/12; Revised 07/11/12; Accepted 07/18/12. *Address correspondence to: Giuseppe Guglielmi, MD, Department of Radiology, University of Foggia, Viale L. Pinto 1, 71100 Foggia, Italy. E-mail:
[email protected]
341
342 Quantitative ultrasound (QUS) methods have been developed for the indirect assessment of bone quality and skeletal status on the basis of a variety of experiences, suggesting that ultrasound parameters provide information not only about bone density but also on architecture and elasticity (7e9). Unlike dual-energy X-ray absorptiometry (DXA), QUS is a radiation-free technique, relatively inexpensive and easily transportable. Thus, QUS method has been proposed as a cost-effective diagnostic choice for osteoporosis screening and risk assessment (10). In particular, phalangeal QUS technique has provided large evidence of efficacy in the discrimination of osteoporotic fractured subjects in crosssectional studies (11e15), whereas contradictory results have been reported in prospective studies. Mele et al (16) found QUS to be a significant independent predictor of osteoporotic low-trauma fractures in a small cohort of Italian postmenopausal women, whereas Krieg et al (17) found opposite results in a large elderly Swiss population even if it has to be outlined that Krieg et al reported some difficulties with the use and calibration of the first-generation device so that a large part of the measurements had to be excluded from the analysis. A new version of the DBM Sonic device is now available, the DBM Sonic Bone Profiler (IGEA SpA, Modena, Italy), which has been automated and it is not operator dependent. The aim of this prospective study was to determine the efficacy of the DBM Sonic Bone Profiler in the prediction of future major osteoporotic fractures in a large Italian postmenopausal population enrolled in a multicenter observational study. Furthermore, the influence of the use of antiosteoporosis drugs as well as supplementation with calcium and/or vitamin D (vitD) has been evaluated in addition to age, age at menopause, and body mass index (BMI).
Materials and Methods Patients In this study, we investigated the osteoporotic fracture prediction of phalangeal QUS in a cohort of postmenopausal women recruited in 5 different centers, 2 located in the northern part of Italy (Torino and Verona), 1 in the center (Roma) and 2 in the south (Palermo and S. Giovanni Rotondo), thus representative of the entire Italian population. A total of 2341 postmenopausal women aged between 45 and 94 yr have been measured with phalangeal QUS technique as they were included in a screening program for osteoporosis. They have been randomly selected among the lists of some general practitioners referring to the menopausal centers. All participants gave their written informed consent, and the study was approved by the local ethics committee and was conducted in accordance with good clinical practice rules. In all centers, the device used for QUS measurement was the Sonic Bone Profiler (IGEA SpA, Modena, Italy). Information including age, BMI, age at menopause, complete medical and pharmacologic anamnesis were recorded by the clinicians during the entire period of the observation.
Guglielmi et al. Exclusion criteria included the conditions likely to influence bone metabolism such as rheumatoid arthritis, impaired renal function, present or previous hyperparathyroidism or hypothyroidism, chronic use of corticosteroids and anticonvulsants. After about 3 yr since visit and QUS measurement, women were successfully recontacted by phone call or for routine visit, and information regarding the continuing persistence of inclusion/exclusion criteria, new fracture occurrence, and osteoporosis treatments in the last 3 yr were recorded.
QUS Measurement One single ultrasound measurement was performed in all subjects at the proximal phalanx metaphysis of the last 4 fingers of the hand using a DBM Sonic Bone Profiler device (IGEA). Measurements were performed by a trained physician on the nondominant hand of the study subjects. Each measurement lasts about 5 min per patient. Two probes mounted on a precision caliper, one acting as a generator of the signal, the other as a receiver, are positioned on the lateral and medial surface of each finger. The emitting probe generates an ultrasound signal with a frequency of 1.25 MHz, the receiving probe receives the ultrasound pulse crossing the phalanx. The coupling of the probes with the skin is mediated by standard ultrasound gel. The device calculates the speed of sound (SoS in meters per second [m/s]) through the phalanx by measuring the width of the finger (including soft tissues) divided by the time of flight, defined as the time from emitted pulse to received signal, considering the signal that reaches a predetermined minimum amplitude value (2 mV) for the first time; thus, the assessed ultrasound velocity is amplitude-dependent SOS (AD-SoS). AD-SoS is directly proportional to true bone density and mechanical resistance of the bone; thus, in osteoporotic subjects, AD-SoS is low, whereas in healthy subjects, AD-SoS is high. We also evaluated the ultrasound bone profile index (UBPI), a combination of 3 ultrasound parameters calculated by signal analysis technique on the ultrasound signal received once it is transmitted through the phalanx: fast-wave amplitude is the amplitude of the first ultrasound pulse received at the receiving probe once the ultrasound pulse has propagated throughout the phalanx, bone transmission time is the time needed for ultrasound wave to propagate through the bone tissue alone, and signal dynamic is the sharpness of the first 2 ultrasound pulses received at the receiving probe and is calculated as the second derivative of the amplitude by time. UBPI calculation is performed by the device and has been optimized to provide a determination of the probability of fracture in the individual subject, as reported by W€uster et al (12).
Quality Assurance Additional quality controls (QCs) were done every time the QUS devices were used according to the manufacturer’s guidelines to verify the stability of the respective systems. In particular, the calibration of the QUS device was performed with a standard Plexiglas phantom provided by the manufacturer, and all the devices resulted to be
Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health
Volume 16, 2013
Fracture Risk in Postmenopausal Women by QUS cross-calibrated. Once every 2 yr, a maintenance control was foreseen by the manufacturer for overall QC, amplitude calibration was performed by manufacturer personnel with a particular ‘‘amplitude phantom’’ at IGEA laboratories.
Statistical Analysis Data analysis was performed with SPSS 13.0 software (SPSS Inc., Chicago IL). The results are expressed as mean standard deviation (SD). Comparison of the data between groups and the presence of vertebral fracture was determined using an unpaired 2-tailed Student’s t-test. Multiple logistic regression analysis was used to estimate the relative risk (RR) per SD decrease for osteoporotic fracture and the corresponding 95% confidence interval (CI). RRs were adjusted for the resulting significant anthropometric pharmacologic variablesdage, BMI, use of osteotrophic drugs, supplementation with calcium and vitDdto calculate the risk of fracture for 1 SD decrease of each considered variable of QUS parameters. The area under a receiver operating characteristic (ROC) curve was used to determine the ability of different variables in the discrimination of subjects with osteoporotic fractures from nonfractured subjects. The area under the curve (AUC) and the standard errors were obtained, and p values were calculated using the method of Hanley and McNeil (18). The values of p ! 0.05 were considered significant.
Results QC data of both devices did not show any shift or drift during the entire study period. The devices used in the present study were therefore characterized as stable. The precision of the measurement of AD-SoS was below 0.9% in each of the 5 recruitment centers. A total of 2341 women were initially recruited but only 2211 (94%) could be successfully recontacted after a period of 3 yr and continuing to fulfill inclusion/exclusion criteria. Dropout women did not differ significantly from the original group (data not shown). Table 1 Demographics of the Population Demographic parameters
Mean
SD
Minimum Maximum
Age (yr) 60.9 10.0 45 Menopausal age (yr) 49.3 4.4 20 BMI (kg/m2) 26.5 4.6 15.8 AD-SoS (m/s) 1993 115 1600 AD-SoS T-score 1.87 1.64 7.49 UBPI (U) 0.58 0.25 0.05 UBPI T-score 1.81 1.78 5.93
94 61 51.3 2260 1.94 1.00 1.21
Abbr: AD-SOS, amplitude-dependent speed of sound; BMI, body mass index; SD, standard deviation; UBPI, ultrasound bone profile index.
343 Demographics of the population are reported in Table 1. After 3 yr from QUS measurement, 142 new fractures occurred, of which a total number of 108 new low-trauma major osteoporotic fractures (hip, clinical vertebral, wrist, shoulder) were identified. Among these, the most common are hip fractures (n 5 23) and clinical vertebral fractures (n 5 51). Unfractured women were significantly younger than fractured ones ( p ! 0.0001 for both fractured groups) and had similar age at menopause ( p 5 0.23) in the women with hip fracture, whereas clinical vertebral fracture group had significantly lower age at menopause ( p ! 0.0001). AD-SoS and UBPI T-scores were significantly lower ( p ! 0.0001) in both fractured groups with respect to unfractured group (Table 2). Upon direct request, 183 women reported to have been on treatment during the last 3 yr or to be currently on treatment with any osteotrophic drug (calcitonin, estrogen, raloxifene, bisphosphonates, strontium ranelate, anabolic agents, teriparatide), and 205 women reported to follow a supplementation treatment with calcium and/or vitD. Women on treatment with osteotrophic drugs were older than women untreated (67.9 8.8 yr vs 60.2 9.8 yr; p ! 0.0001) and had significantly lower AD-SoS and UBPI Tscores ( 3.5 1.7 SD vs 1.7 1.5 SD for AD-SoS; 3.6 1.2 SD vs 1.6 1.7 SD for UBPI; p ! 0.0001 for both). In the untreated group, 81 women sustained a fracture (4%), whereas in the treated group, 54 women sustained a fracture in the study period (30%), p ! 0.0001. Women who were on supplementation with calcium and/or vitD were older than other women (65.8 8.7 yr vs 60.4 10.0 yr; p ! 0.0001) and had significantly lower AD-SoS and UBPI T-scores ( 3.1 1.7 SD vs 1.7 1.6 SD for AD-SoS; 3.2 1.5 SD vs 1.7 1.8 SD; p ! 0.0001 for both). RR per SD decrease for major fractures was 1.77 (CI: 1.59e1.97) for AD-SoS and 2.06 (CI: 1.78e2.37) for UBPI. Data of RRs after correction for relevant covariates are reported in Table 3. Combination of AD-SoS and UBPI in the multilinear logistic model for the prediction of osteoporotic fractures does not increase significantly the likelihood of the model itself (data not shown). RRs per SD decrease for hip fractures were 1.92 (CI: 1.55e2.37) for AD-SoS and 2.68 (CI: 1.86e3.86) for UBPI. RRs for clinical vertebral fractures were 1.63 (CI: 1.41e1.88) for AD-SoS and 1.73 (CI: 1.44e2.08) for UBPI. Analyses by means of ROC curve are displayed in Figs. 1 and 2 and Table 4.
Discussion In this prospective cohort of more than 2000 postmenopausal women, ultrasound parameters AD-SoS and UBPI are able to significantly predict major fractures and, in particular, clinical vertebral and hip fractures. The incidence of hip fractures in the study group was 3.5 per 1000 women per year, which is lower with respect to that obtained in the Study of Osteoporotic Fractures (SOF) by
Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health
Volume 16, 2013
344
Guglielmi et al. Table 2 Data Correlation of Fractured and Unfractured Subjects Unfractured (N 5 2103)
Parameters Age (yr) Age at menopause (yr) BMI (kg/m2) Anti-osteoporosis drug (%) vitD supplementation (%) AD-SoS T-score UBPI T-score
Hip fractures (n 5 23)
60.3 9.7 49.4 4.3 26.5 4.6 129 (6.1) 168 (8.0) 1.76 1.59 1.69 1.75
Clinical vertebral fractures (n 5 51)
74.9 9.0 48.0 5.0 27.5 4.4 9 (39.1) 9 (39.1) 4.15 1.59 4.19 1.11
70.4 9.3 46.1 5.5 26.8 4.9 40 (78.4) 8 (15.7) 4.33 1.71 4.13 1.04
No fracture vs hip fracture
No fracture vs vertebral fracture
Hip fracture vs vertebral fracture
!0.0001 0.23 0.30 !0.0001 !0.0001 !0.0001 !0.0001
!0.0001 !0.0001 0.65 !0.0001 0.08 !0.0001 !0.0001
0.03 0.17 0.56 0.002 0.038 0.67 0.82
Abbr: AD-SoS, amplitude-dependent speed of sound; BMI, body mass index; UBPI, ultrasound bone profile index; vitD, vitamin D.
for AD-SoS and was 1.70 with a 95% CI of 1.57e1.84, which is very similar to the RR value obtained in the present study. The SEMOF study could not evidence that QUS at phalanges predict hip fractures, the finding can be explained by considering that in the Swiss study device calibration was not homogenous among centers leading to the exclusion of a large number of subjects from the analysis; furthermore, in that study, the first-generation device was used that did not include an automatic procedure to control the QUS probe positioning and measurement procedure (17). In our study, the RRs for osteoporotic vertebral fractures and hip fractures for AD-SoS and UBPI are similar to those obtained in the SEMOF study for calcaneal QUS SOS measurements in the prediction of femoral neck fractures (17). The EPIDOS prospective study on osteoporotic fracture prediction with heel QUS reported RR per 1 SD decrease, adjusted for age, weight, and center of 2.0 for broadband ultrasound attenuation (BUA) and 1.7 for SOS, whereas RR for femoral neck BMD measured by DXA was 1.9 (21). In the SOF study, RR was 2.0 for heel QUS BUA and 2.6 for femoral neck BMD (22). Results obtained in all these prospective studies conducted with QUS devices, mainly at the heel (21e24), are difficult to directly compare with the present study, especially due to the age difference among the populations investigated. The RRs obtained for BUA are higher than
Bauer et al (4.4 per 1000 women-year) and in the SEMOF study by Krieg et al (3.9 per 1000 women-year), but in these 2 studies, the average age of the population was sensibly high (76 yr in the SOF study and 75 yr in the Swiss Evaluation of Methods of Measuring Osteoporotic Fracture Risk [SEMOF] study). We could compare the incidence of hip fracture in Italy by age classes, having the reference data provided by Piscitelli et al (19) for the year 2005. In this report, in the age range from 45 to 64 yr, the incidence is 0.6 per 1000 women per year, and in our population, we observed an incidence of 0.7 in the same age range, which is very similar to that reported by Piscitelli et al. Our results are in agreement with the previous prospective study on QUS measurement at the phalanges, published by Mele et al, in which the RR for nonvertebral osteoporotic low-trauma fractures for AD-SoS was estimated as 1.5 (CI: 1.1e1.7) (16). Our results confirm the ability of the QUS measurement to predict all major low-trauma fractures; in particular, in the present study, both AD-SoS and UBPI demonstrated to be able to predict hip fracture, as well as clinical vertebral fractures. Similar results were presented by Kanis et al (20) in a study of large dimensions for the evaluation of the 10-yr risk of clinical vertebral fractures in the European population; in this study, the gradient of risk was calculated
Table 3 Relative Risks for Osteoporotic Vertebral and Hip Fractures for AD-SoS and UBPI After Correction for Relevant Covariates
Unadjusted
Adjusted for age, BMI, age at menopause
Adjusted for age, BMI, age at menopause, center, therapy, supplementation
Parameters
RR
95% CI
RR
95% CI
RR
95% CI
AD-SoS UBPI
1.77 2.06
1.59e1.97 1.78e2.37
1.44 1.67
1.26e1.65 1.39e2.00
1.24 1.39
1.06e1.44 1.14e1.70
Abbr: AD-SoS, amplitude-dependent speed of sound; BMI, body mass index; CI, confidence interval; RR, relative risk; UBPI, ultrasound bone profile index. Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health
Volume 16, 2013
Fracture Risk in Postmenopausal Women by QUS
345
Fig. 1. ROC curve for amplitude-dependent speed of sound in the discrimination of major osteoporotic fractures. ROC, receiver operating characteristic.
Fig. 2. ROC curve for ultrasound bone profile index in the discrimination of major osteoporotic fractures. ROC, receiver operating characteristic.
the RR obtained for AD-SoS and UBPI, whereas RRs obtained for SOS are quite similar. Femoral neck BMD in the SOF study showed higher RR, whereas femoral neck BMD in the EPIDOS study showed a similar RR to that of ADSoS and UBPI. It is to be outlined that RR calculated in the previous studies did not take into consideration the effect of the age at menopause, which has been observed in the literature as an independent risk factor for osteoporosis (25) and, most importantly, the effect of therapies and supplementations, which tends to cover the true relationship between fracture prediction and QUS parameters. A recent meta-analysis published by Moayyeri et al (26) reported that RR of osteoporotic fracture per standard decrease of QUS parameters measured at the calcaneus, calculated on 21 prospective studies, was in the range between 1.69 and 2.26, which is similar to the range obtained in the present study for QUS parameters measured at the phalanges from 1.77 to 2.06. In fact, we added in our analysis the investigation about the effect of pharmacologic treatments (calcitonin, estrogen, raloxifene, bisphosphonates, strontium ranelate, anabolic agents, teriparatide) and of calcium/vitD supplementation. To our knowledge, this is the first prospective study on QUS measurement in which the effect of drugs and supplementation in the occurrence of new osteoporotic fractures is taken into consideration. AD-SoS and UBPI remain significant predictors of future osteoporotic fractures even taking into account the effect of anti-osteoporosis treatments or calcium/vitD supplementation.
Analysis by the mean of ROC curve confirms the good ability of AD-SoS and UBPI to discriminate between subjects who underwent major osteoporotic fractures, as well as hip and clinical vertebral fractures, with high values of AUC from 0.813 to 0.867. It has been observed in several studies that AD-SoS and UBPI investigate different properties of the bone (12,27). Table 4 Area Under the ROC Curve for AD-SoS and UBPI in Discriminating Major Osteoporotic Fractures, Hip Fractures, and Clinical Vertebral Fractures Fractures and QUS parameters Major fractures AD-SoS UBPI Hip fractures AD-SoS UBPI Vertebral fractures AD-SoS UBPI
AUC SD
95% CI
p Value
0.823 0.014 0.798e0.853 !0.0001 0.813 0.018 0.778e0.847 !0.0001 0.853 0.029 0.797e0.910 !0.0001 0.867 0.031 0.806e0.927 !0.0001 0.853 0.032 0.790e0.916 !0.0001 0.862 0.028 0.808e0.916 !0.0001
Abbr: AD-SoS, amplitude-dependent speed of sound; AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic; SD, standard deviation; UBPI, ultrasound bone profile index.
Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health
Volume 16, 2013
346 The combination of AD-SoS and UBPI in all the statistical models applied does not increase the fracture prediction ability of the QUS measurement at the phalanges, as the different characteristics of the bone investigated by the 2 parameters are all strictly related to the fracture occurrence and thus probably intercorrelated. This study has several limitations. The selection of participants was performed among women already following a surveillance program, thus probably at higher risk of osteoporosis with respect to the Italian postmenopausal population. Women were interviewed at about 3 yr since the QUS measurement by phone call or follow-up visit, some of them could not be reached or decided not to answer to the questions. No radiologic documentation of fractures occurred was requested and all information regarding fractures, pharmacologic treatments, calcium/vitD supplementation were recorded on the basis of patients’ answers. Sometimes, the information regarding the name of the drug or dose was not clear; for this reason, we decided not to take into consideration the effect of different osteotrophic drugs but to consider only the data regarding the use of drugs to treat osteoporosis. No information regarding smoke was collected in this cohort of women, as it is known that the percentage of smoking persons among Italian postmenopausal women is low. In the determination of RR for AD-SoS and UBPI in the logistic multivariate analysis for vertebral fractures only and for femoral fractures only, we could not calculate adjusted RRs for covariates because of the too small number of fracture cases and too high difference in the mean age of the groups, with consequent inadequate likelihood of the model to describe the data. In conclusion, our results indicate that AD-SoS and UBPI are significant independent predictors of major osteoporotic fractures as well as hip and clinical vertebral fractures in postmenopausal women.
References 1. Melton LJ III, Chrischilles EA, Cooper C, et al. 1992 How many women have osteoporosis? J Bone Miner Res 7:1005e1010. 2. Gullberg B, Johnell O, Kanis JA. 1997 World-wide projections for hip fracture. Osteoporos Int 7:407e413. 3. Cranney A, Tugwell P, Adachi J, et al. 2002 Meta-analysis of risedronate for the treatment for postmenopausal osteoporosis. Endocr Rev 23:517e523. 4. Hauselmann HJ, Rizzoli RA. 2003 A comprehensive review of treatments for postmenopausal osteoporosis. Osteoporos Int 14:2e12. 5. Cranney A, Wells G, Willan A, et al. 2002 Meta-analysis of alendronate for the treatment for postmenopausal osteoporosis. Endocr Rev 23:508e516. 6. Guglielmi G, Muscarella S, Bazzocchi A. 2011 Integrated imaging approach to osteoporosis: state-of-the-art review and update. Radiographics 31(5):1343e1364. 7. de Terlizzi F, Battista S, Cavani F, et al. 2000 Influence of bone tissue density and elasticity on ultrasound propagation: an in vitro study. J Bone Miner Res 15:2458e2466. 8. Barkmann R, L€usse S, Stampa B, et al. 2000 Assessment of the geometry of human finger phalanges using quantitative ultrasound in vivo. Osteoporos Int 11:745e755.
Guglielmi et al. 9. Guglielmi G, de Terlizzi F, Scalzo G, et al. 2010 Cortical thickness and medullary canal dimensions of the bone phalanx are predicted by quantitative ultrasound parameters. J Clin Densitom 13(2):219e227. 10. Benitez CL, Schneider DL, Barrett-Connor E, Sartoris DJ. 2000 Hand ultrasound for osteoporosis screening in postmenopausal women. Osteoporos Int 11:203e210. 11. Guglielmi G, Cammisa M, De Serio A, et al. 1999 Phalangeal US velocity discriminates between normal and vertebrally fractured subjects. Eur Radiol 9:1632e1637. 12. W€ uster C, Albanese C, de Aloysio D, et al. 2000 Phalangeal osteosonogrammetry study (PhOS): age related changes, diagnostic sensitivity and discrimination power. J Bone Miner Res 15(8):1603e1614. 13. Hartl F, Tyndall A, Kraenzlin M, et al. 2002 Discriminatory ability of quantitative ultrasound parameters and bone mineral density in a population-based sample of postmenopausal women with vertebral fractures: result of the Basel Osteoporosis Study. J Bone Miner Res 17:321e330. 14. Albanese CV, Cepollaro C, de Terlizzi F, et al. 2009 Performance of five phalangeal QUS parameters in the evaluation of gonadal-status, age and vertebral fracture risk compared with DXA. Ultrasound Med Biol 35(4):537e544. 15. Alexandersen P, de Terlizzi F, Tanko LB, et al. 2005 Comparison of quantitative ultrasound of the phalanges with conventional bone densitometry in healthy postmenopausal women. Osteoporos Int 16:1071e1078. 16. Mele R, Masci G, Ventura V, et al. 1997 Three-year longitudinal study with quantitative ultrasound at the hand phalanx in a female population. Osteoporos Int 7:550e557. 17. Krieg MA, Cornuz J, Ruffieux C, et al. 2006 Prediction of hip fracture risk by quantitative ultrasound in more than 7000 Swiss women O or 570 years of age: comparison of three technologically different bone ultrasound devices in the SEMOF study. J Bone Miner Res 21(9):1457e1463. 18. Hanley JA, McNeil BJ. 1983 A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 148(3):839e843. 19. Piscitelli P, Gimigliano F, Gatto S, et al. 2010 Hip fractures in Italy: 2000e2005 extension study. Osteoporos Int 21:1323e1330. 20. Kanis JA, Johnell O, Oden A, et al. 2005 Ten-year probabilities of clinical vertebral fractures according to phalangeal quantitative ultrasonography. Osteoporos Int 16:1065e1070. 21. Hans D, Dargent-Molina P, Schott AM, et al. 1996 Ultrasonographic heel measurements to predict hip fracture in elderly women: the EPIDOS prospective study. Lancet 348:511e514. 22. Bauer DC, Gluer CC, Cauley JA, et al. 1997 Broadband ultrasound attenuation predict fractures strongly and independently of densitometry in older women. A prospective study. Arch Intern Med 157:629e634. 23. Porter RW, Miller CG, Grainger D, Palmer SB. 1990 Prediction of hip fracture in elderly women: a prospective study. BMJ 29(301):638e641. 24. Guessous I, Cornuz J, Ruffieux C, et al. 2008 Osteoporotic fracture risk in elderly women: estimation with quantitative heel US and clinical risk factors. Radiology 248:179e184. 25. Gambacciani M, de Aloysio D, Elia D, et al. 2004 Quantitative ultrasound (QUS) of bone in the management of postmenopausal women. Maturitas 47:139e149. 26. Moayyeri A, Adams JE, Adler RA, et al. 2012 Quantitative ultrasound of the heel and fracture risk assessment: an updated meta-analysis. Osteoporos Int 23:143e153. 27. Agostinelli D, de Terlizzi F. 2007 QUS in monitoring raloxifene and estrogen-progestogens: a 4-year longitudinal study. Ultrasound Med Biol 33(8):1184e1190.
Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health
Volume 16, 2013