• bone ultrasound;
  • fractures;
  • osteoporosis;
  • bone densitometry;
  • meta-analysis


  1. Top of page
  2. Abstract
  7. Acknowledgements

The relationship between bone QUS and fracture risk was estimated in a systematic review of data from 14 prospective studies of 47,300 individuals and 2350 incident fractures. In older women, low QUS values were associated with overall fracture risk, low-trauma fractures, and with hip, forearm, and humerus fractures separately.

Introduction: Bone quantitative ultrasound (QUS) has emerged as a promising technique to evaluate bone status. The aim of this study was to determine the association between measurements of QUS with the risk of fracture.

Materials and Methods: A meta-analysis of prospective cohort studies published between 1985 and June 2005 with a baseline measurement of QUS and subsequent follow-up for fractures was carried out. Fourteen separate study populations, consisting of about 47,300 individuals (85.4% women), with about 124,000 person-years of observation and over 2350 fractures, including 653 hip, 529 forearm, and 386 humeral fractures, were analyzed. The main outcome measure was the estimated relative risk of fracture for a decrease in bone QUS parameters of 1 SD below sex- and age-adjusted mean in women.

Results: Eleven studies evaluated QUS at the heel, with patella and phalanx (two studies each) and distal radius (one study) being scarcely used. There was not significant heterogeneity among the studies included in the review. Relative risk estimates (95% CI) for overall fractures were 1.55 (1.35–1.78) for each SD decrease in broadband ultrasound attenuation (BUA), 1.63 (1.37–1.93) for speed of sound (SOS), and 1.74 (1.39–2.17) for QUS index/stiffness index (QUI/SI). Risk estimates were similar or slightly higher for hip fractures and low-energy trauma fractures. Humeral and forearm/wrist fractures were also related with lower QUS values.

Conclusions: Measurements of bone QUS are significantly associated with nonspinal fracture risk in older women in a similar degree to DXA. QUS may be a valid alternative to evaluate fracture risk in situations where DXA is not accessible.


  1. Top of page
  2. Abstract
  7. Acknowledgements

The identification of individuals at high risk of osteoporotic fractures, particularly among the elderly, is of pivotal importance to implement effective preventive measures. Fragility fractures occur because a combination of low bone mass and deterioration in the quality properties of the bone, including macro- and microarchitectural changes in the trabecular and cortical bone, and an age-related increase in the risk of falls. Therefore, the current approach to the assessment of fracture risk is based on the evaluation of several risk factors, including age, the existence of previous fractures, a family history of fractures, and several clinical factors related with the bone status and propensity to fall.(1–3) However, the measurement of BMD by DXA is still considered one of the most important predictors of fracture risk in both sexes,(4) probably because it is a measurable factor, and it is one that can be influenced by therapeutic manipulations. However, axial DXA equipment are expensive, nonportable, involve radiation exposure, and are usually restricted to tertiary care hospitals or specialized clinics because dedicated trained personnel is required. Therefore, in some geographical areas, there are clear limitations in the accessibility to this technique.(5) Thus, the introduction of new alternative methods to evaluate the status of the bone would be of interest.

Quantitative bone ultrasound (QUS) is emerging as a low-cost screening technique that is able to identify women at risk for the osteoporosis and that may be used by general practitioners in primary care and in ambulatory settings.(6) QUS parameters permit analysis of some physical properties of bone tissue, which, in turn, are important determinants of bone stiffness, load failure, and fracture risk by providing additional information to bone mass alone.(7) In fact, a number of cross-sectional studies have shown that the values of the parameters measured with QUS, principally at the calcaneus, are lower in women with a history of osteoporotic fracture, regardless of the BMD values determined by DXA.(8–15) A recent review has also concluded that broadband ultrasound attenuation (BUA) at the heel is an independent risk indicator for hip fractures, but the information was limited to two cohorts.(4) Whereas QUS is increasingly incorporated to clinical practice, no firm conclusions have yet been drawn regarding its potential use in fracture risk assessment. For this reason, the present systematic review of the literature for all prospective studies aims to analyze the association of QUS measurements with the risk of fracture.


  1. Top of page
  2. Abstract
  7. Acknowledgements

Search strategy and study selection

We identified literature and primary studies for this report by searching computer databases of the medical literature (Medline and EMBASE) from 1985 to June 2005. We based our search on the following keywords: “bone and bones,” “bone density,” “quantitative ultrasound,” and “fractures” combined with the different measurement sites in humans. No language limitations were imposed. We identified 493 articles in this way. A hand search in these references was performed according to the prespecified eligibility criteria for inclusion in the present meta-analysis: (1) prospective studies with a baseline measurement with a QUS technique reported in absolute units, (2) fractures have to be the main outcome and must have occurred after bone QUS measurement had been taken, and (3) studies gave a relative risk (RR) estimate (or to allow its estimation) for fractures. With these criteria, 14 studies were identified and included (Fig. 1). Two prospective studies were not used in this review. The study by Porter et al.(16) did not include any RR estimate. The NORA Study reported only T score values of one nonspecified QUS parameter, and absolute values of bone ultrasound parameters were not included.(17) In one instance (EPIDOS Study), there was more than one published report of the same study population at different time-points and fracture locations.(18–21) In this case, we used the most updated published data.

thumbnail image

Figure Figure 1. Flow chart of study selection and inclusion process.

Download figure to PowerPoint

Data were independently extracted by two investigators (FM and MDR) and compared. Methodological quality of the studies was assessed according to STROBE (Strengthening the Reporting of Observational studies in Epidemiology) statement.(22) The following information was summarized: country where the study was performed, study period, sex, age, type of population (general population, ambulatory, or hospital-based), length of follow-up, anatomical measurement site, QUS equipment, QUS parameters with SDs in the study populations, incident fracture sites, fracture rates, type of incident fracture (low energy or any type), RR or related measures (hazard ratio, OR) and its 95% CI, model used to yield RRs, and, if relevant, variables adjusted for in multivariate analyses. Fracture assessment methods were reviewed. Information on all fractures was used regardless of type of trauma. In addition, fractures associated with low trauma (fragility fractures) were analyzed in those series where this information was included. Finally, hip, forearm/wrist, and humerus fractures were considered separately.

Outcome variables and statistical analysis

We chose RR for fracture associated with a decrease in QUS parameters of 1 SD adjusted by age, because it is independent of QUS devices and population characteristics. RR estimates were pooled using as weight their inverse of variance under a random effects model.(23) This model was chosen as simulations have shown that it works better than the fixed effects model when the number of primary studies is <20.(24,25) Heterogeneity was assessed in all estimates applying the Q statistic and using a cut-off level of p = 0.1 to establish its presence.(23) The degree of inconsistency was assessed estimating I2, which describes the percentage of total variation across studies that is caused by heterogeneity rather than to chance.(26,27) All the analyses were stratified by the type of parameter measured by QUS (bone ultrasound attenuation, BUA; speed of sound, SOS; quantitative ultrasound index/stiffness index, QUI/SI). The presence of publication bias was ascertained by a Christmas tree, representing the natural logarithm of RR (x-axis) against its precision [1/SE(ln RR)] in the y-axis, and by the procedures of Egger et al.(28) and Macaskill et al.(29) that were applied for the QUS parameters, and in fragility and all fractures types (all sites combined). The Stata 8-SE package (College Station, TX, USA) was used for statistical analyses.


  1. Top of page
  2. Abstract
  7. Acknowledgements

Description of the study population

Table 1 gives details of the 14 study cohorts, consisting of ∼47,300 individuals, 124,100 person-years of observation, and ∼2350 fractures, including 653 hip, 528 forearm/wrist, and 386 humerus fractures, that were suitable for inclusion in our analysis.(19–21,30–43) Only 71 incident vertebral fractures or deformities were collected. All fractures were verified by health care providers except for one study(32) where fractures were collected by patient self-report.

Table Table 1.. Summary Details of Prospective Studies of Quantitative Ultrasound and Fracture Risk That Were Included in Meta-Analysis
Thumbnail image of

The studies were carried out in six European Union member States (n = 10), the United States (n = 3), and Switzerland (n = 1; Table 1). In four studies, the mean age at study baseline was <60 years (33.3% of the total population included in the meta-analysis), whereas the majority of studies (57%) assessed populations ⩾70 years of age (59.7% of the population included in the meta-analysis). Only three studies included a male population with a total of ∼6670 individuals (14.1% of the total population included in the meta-analysis); however, most of them (96.7%) were included in a single study(42) that was the only one that provides stratified results by sex; therefore, a separate meta-analysis of the male population was not feasible. Moreover, because it has been shown than average BUA and SOS in men is higher than in women,(44) we focused the analysis in women only. Two studies that provided results pooling both sexes(35,36) were excluded to keep the grouping more homogeneous. The average length of follow-up of the different primary studies ranged from 1.0 to 5.5 years. Sample size was also variable: there are five studies (35.7%) with >5000 people, whereas seven (50%) have <1000 participants. The most frequent measurement site was the calcaneus (11 studies: 78.6%), whereas the patella and the phalanx were used in only two studies each,(30,32,37,40) and the distal radius in a single, small study(37) (Table 1), which precluded a meta-analysis of the results of noncalcaneus measurement sites.

Heterogeneity and publication bias

We did not observe any heterogeneity for the fractures outcome (either all fractures or low-trauma fractures) for any of the three QUS parameters analyzed (BUA, SOS, and QUI/SI; p > 0.25). The degree of inconsistency was low: I2 values were usually <20% and all of the 95% CIs included the null value. A Christmas tree for the association between BUA and fracture risk (any site) did not reveal any relevant asymmetry (Fig. 2). Similarly, regression methods to detect publication bias, applied to this association and others, did not suggest that publication bias was present.

thumbnail image

Figure Figure 2. Christmas tree for the association between BUA and fracture risk (any site).

Download figure to PowerPoint

Relationship between QUS values and fracture

Lower values of the QUS parameters were associated with a significant increase of any subsequent fracture at any site (Table 2; Fig. 3). The age-adjusted RR for any fracture per each SD decrease ranged between 1.55 and 1.74 depending on the QUS parameter analyzed. The results were also significant for hip (Table 3; Fig. 4), forearm/wrist, and humeral fractures (Table 2). The age-adjusted relationships between QUS parameters and low-energy fractures are shown in Table 3. The RRs for this type of fracture were similar or slightly higher than for any fracture regardless the level of trauma. Thus, significant RRs of 1.74 (1.38–2.21), 1.73 (1.38–2.17), and 1.66 (1.39–1.99) for each SD decrease in BUA, SOS, and QUI/SI, respectively, were found (Table 3).

Table Table 2.. RR and 95% CI of Fractures, Regardless of the Energy Impact, for 1 SD Decrease in the QUS Parameters in Women
Thumbnail image of
Table Table 3.. RR and 95% CI of Low-Energy Trauma Fractures for 1 SD Decrease in the QUS Parameters in Women
Thumbnail image of
thumbnail image

Figure Figure 3. Relative risk estimates per 1 SD decrease in QUS parameters in women measured at any site, for fracture, regardless the level of trauma. (A) BUA. (B) SOS. (C) QUI/SI. Size of the RR node reflects the weight of each study in the combined estimate.

Download figure to PowerPoint

thumbnail image

Figure Figure 4. Relative risk estimates per 1 SD decrease in BUA for hip fractures in women, regardless the level of trauma. Size of the RR node reflects the weight of each study in the combined estimate.

Download figure to PowerPoint

We finally analyzed the magnitude of the association of overall fractures risk in those studies that, in addition to QUS, included a BMD measurement by DXA. In the five studies where these data were available, which comprised nearly 14,000 patients, the relationship between BMD measured by DXA and fractures was comparable with BUA [RRBMD = 1.60 (1.22–2.05) versus RRBUA = 1.50 (1.26–1.77)]. Data for SOS were similar, but based on a smaller number of cases (∼7000) [RRSOS = 1.77 (1.17–2.68) versus RRBMD 1.74 (1.50–2.02); Table 4].

Table Table 4.. Relationship of QUS and DXA Measurements With Any Fracture
Thumbnail image of


  1. Top of page
  2. Abstract
  7. Acknowledgements

Meta-analysis of 14 prospective studies, which included nearly 43,700 individuals, showed the relationship between low QUS values and increased fracture risk independently of age and other covariates. Depending on the QUS parameter, the type of fracture, and the ultrasound measurement site, the estimated relative risk ranged between 1.23 and 1.94 for each SD decrease in QUS measurements, being statistically significant in all circumstances.

QUS methods have been introduced in clinical practice for the assessment of skeletal status in the last 15 years.(15) Its low cost and portability make QUS an attractive technology for assessing risk of fracture in larger populations than may be suitable for axial DXA. Moreover, the lack of exposure to ionizing radiation enhances the patient's acceptance of the test, and the operation of the equipment because less demanding regulations and space. Bone ultrasound parameters might measure different physical properties of bone to provide complementary information about fracture risk beyond BMD.(6) However, their potential clinical applications have not been fully validated, and several approaches have been proposed.(6,45,46) These include the diagnosis of osteoporosis, monitoring of skeletal changes caused by therapeutic interventions or disease progression, fracture risk assessment, or as a prescreening tool for identifying subjects who would be referred to further diagnostic work-up using DXA or biochemical markers of bone turnover. Whereas some of these strategies are still in early stages of research, the present systematic review of prospective fracture studies clearly support the concept that QUS, mainly at the heel, is a valuable tool to detect an increased risk of fracture. These results also confirm the findings of numerous cross-sectional studies that have shown that QUS values are generally lower in patients with fragility fractures.(8–14)

The association with an increased risk of fracture we have found (RR of 1.55–1.74 for each 1 SD reduction in QUS parameters for any fracture regardless of trauma and 1.66–1.74 for any low-trauma fracture) shows that the performance of QUS devices is similar to peripheral or axial DXA for these same categories of fractures reported in meta-analysis and large series(4,17,47,48) or in our own analysis (Table 4). However, it was inferior to BMD measurement at the femoral neck with DXA as a risk factor of hip fractures, as reported in a recent review of 12 cohorts, where the risk ratio increased by 2.94 in women for each SD decrease in hip BMD.(4)

It should not be concluded from this work that QUS may substitute DXA. Axial DXA is still considered the “gold standard” method for osteoporosis diagnosis or fracture assessment because its superior precision and capability to accurate measure BMD at axial sites and because more clinical experience and longer follow-up data have been accumulated in the last decades. Moreover, the current World Health Organization diagnosis criteria of osteoporosis, and the main treatment trials in osteoporosis have selected the patient population based on DXA values, making this technique the most validated to support therapy interventions with drugs. However, QUS may constitute a valid alternative to evaluate fracture risk in those circumstances and geographies where DXA equipment are not easily accessible. As with DXA, one important factor that should be taken into consideration when using and interpreting QUS results, is that the different equipment are not comparable, and that the results obtained on one device cannot de directly translated to other different QUS devices, especially if different QUS parameters are measured, as with forearm or finger equipment. The available devices show a big technological diversity and appropriate standardization approaches are eagerly needed. Moreover, the precision of QUS measurements is normally lower than with DXA, with %CV ∼ 2–3%, and varies depending on the parameter measured, the site, and the device used.(49)

In any case, it should be stressed that the assessment of fracture risk can not rely solely on the basis of bone properties, measured by DXA, QUS, or any other method, but in the integration of the skeletal and clinical risk factors in an individual for a given period of time, to express the fracture risk in absolute terms.(50,51) In fact, low QUS values may be considered like an additional independent risk factor for fractures, more than a diagnostic tool to establish whether the patient may be qualified as osteoporotic, osteopenic, or normal. Bone QUS may be also used as a prescreening tool for identifying subjects at highest risk who should be subjected to further risk assessment with bone densitometry, biochemical markers of bone turnover, or alternative techniques to evaluate the bone status.(46)

The two principal variables that are measured with QUS are the SOS passing through the bone, and the attenuation of the ultrasound wave (BUA). These variables are related to several material properties of the tissue as BMD, trabecular orientation, the proportion of trabecular and cortical bone, the composition of the organic and inorganic components, and fatigue damage of the bone. The velocity of an ultrasound wave depends mainly on the material properties of the tissues through which it is propagating, specifically of the BMD and the Young's modulus (a measure of resistance to deformation). The main mechanism of attenuation or energy lost of the ultrasound wave when propagated through the bone tissue is believed to be scattering and absorption.(52) Most of the manufacturers have developed derived parameters from BUA and SOS, such as the so-called “stiffness index” (SI) or the “quantitative ultrasound index” (QUI). These parameters are not directly related to biomechanical properties of bone, but they improve the standardized CV of SOS or BUA alone and compensate for heel temperature variations.(53) Moreover, QUI is more strongly correlated to actual heel BMD obtained with DXA. These derived parameters have been more recently introduced, and the number of prospective studies that have tested them is lower. However, our results show that the decrease in QUI/stiffness measurements shows the better relationship with increased fracture risk for overall fractures, regardless the level of trauma, and hip, forearm/wrist and humerus fractures analyzed separately. The relative risk for hip fracture associated with 1 SD decrease in heel QUI/stiffness in our analysis was 1.94 (1.46–2.59). This value represents a stronger association than the risk of hip fracture observed with 1 SD decrease in spinal BMD measured by DXA according to the results reported in a meta-analysis of prospective studies involving bone densitometry [RR = 1.6 (95% CI = 1.2–2.2)].(47) Our results also show the use of these parameters to evaluate the risk of forearm/wrist and humerus fractures, two common fracture sites associated to low bone mass,(54) and very prevalent in the elderly population. In contrast, we found very limited data in prospective studies to evaluate the capacity of QUS to assess vertebral fracture risk. Thus, the bulk evidence of the relationship of QUS and vertebral fracture risk remains supported by retrospective studies,(10,55–58) which are less accurate and with limited external validity.

The results of our systematic review are based in the assessment of nearly 47,300 individuals and ∼124,100 person-years of prospective observation. These numbers are in the same order of magnitude than meta-analysis of BMD measured with DXA and fracture prediction, where 90,000(47) and 168,366(4) person-years data have been reported. However, some limitations for its external validity should be noted. First, given the very few published studies and the small populations included in prospective cohorts, we were not able to perform a meta-analysis of the fracture risk associated to QUS measured at nonheel sites. Measurements of QUS at the patella and phalanx are restricted to a single variable, namely velocity of sound or its variations, and although these equipment are able to discriminate between fractured and nonfractured patients in most,(10,11,57,59) but not all,(58,60,61) cross-sectional studies, the magnitude of the relationship is normally lower than with BUA, and there is a need for long-term prospective studies with larger patient populations to evaluate the capability of these equipment to assess fracture risk. Second, given the low number of studies in men, the results of this meta-analysis apply only to women. In the largest prospective study in men, Khaw et al.(42) showed that each SD decrease in BUA and SOS values significantly increased the risk of future fracture by 87% and 65%, respectively, in 6485 men, but this analysis was based on 33 fractures only. Thus, until further evidence is available in men, the applicability of our results to male populations should be questioned. Moreover, although the race of the patients was usually not specified, given the geographies where most of these prospective studies were carried out, the results should be interpreted with caution for black women. It has been recently shown that the relationship between BMD and incident nonspinal fractures is significantly lower in black women from the SOF study,(62) and it is uncertain whether this may also be true for QUS. Finally, the results are mainly applicable to elderly women ⩾60 years of age.

In summary, this systematic review has shown that there is strong evidence from prospective studies that QUS measurements, BUA, and SOS, and their derived parameters (QUI/stiffness), are significantly associated with fracture risk, mainly among older women. The strength of the association of QUS with nonspinal fractures is similar to axial or peripheral DXA, but it is inferior to the association between BMD measured at the hip and hip fractures. Furthermore, lower heel QUS values are associated with forearm/wrist and humerus fractures. QUS can be considered a simpler and valid alternative to DXA to assess future fracture risk at nonspinal sites.


  1. Top of page
  2. Abstract
  7. Acknowledgements

This study was supported by an unrestrictive research grant from the Medical Research Department, Eli Lilly and Company, Madrid, Spain.


  1. Top of page
  2. Abstract
  7. Acknowledgements
  • 1
    Kanis JA 2002 Diagnosis of osteoporosis and assessment of fracture risk. Lancet 359: 19291936.
  • 2
    Black DM, Steinbuch M, Palermo L, Dargent-Molina P, Lindsay R, Hoseyni MS, Johnell O 2001 An assessment tool for predicting fracture risk in postmenopausal women. Osteoporos Int 12: 519528.
  • 3
    Eddy DM, Johnston CC, Cummings SR, Dawson-Hughes B, Lindsay R, Melton LJ III, Slemenda CH 1998 Osteoporosis: Review of the evidence for prevention, diagnosis and treatment and cost-effectiveness analysis. Osteoporos Int 8(Suppl 4): S1S88.
  • 4
    Johnell O, Kanis JA, Oden A, Johansson H, De Laet C, Delmas P, Eisman JA, Fujiwara S, Kroger H, Mellstrom D, Meunier PJ, Melton LJ III, O'Neill T, Pols H, Reeve J, Silman A, Tenenhouse A 2005 Predictive value of BMD for hip and other fractures. J Bone Miner Res 20: 11851194.
  • 5
    Osteoporosis in Europe 2005 Indicators of progress. Available online at Accessed February 4, 2006.
  • 6
    Glüer 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: 12801288.
  • 7
    Njeh CF, Fuerst T, Diessel E, Genant HK 2001 Is quantitative ultrasound dependent on bone structure? A reflection. Osteoporos Int 12: 115.
  • 8
    Hernández JL, Marín F, González-Macías J, Díez-Pérez A, Vila J, Giménez S, Galán B, Arenas MS, Suárez F, Gayola L, Guillén G, Sagredo T, Belenguer R, Morón A, Arriaza E 2004 Discriminative capacity of calcaneal quantitative ultrasound and fracture risk factors in postmenopausal women with osteoporotic fractures. Calcif Tissue Int 74: 357365.
  • 9
    Nguyen TV, Center JR, Eisman JA 2004 Bone mineral density-independent association of quantitative ultrasound measurements and fracture risk in women. Osteoporos Int 15: 942947.
  • 10
    Glüer CC, Eastell R, Reid DM, Felsenberg D, Roux C, Barkmann R, Timm W, Blenk T, Armbrecht G, Stewart A, Clowes J, Thomasius FE, Kolta S 2004 Association of five quantitative ultrasound devices and bone densitometry with osteoporotic vertebral fractures in a population-based sample: The OPUS Study. J Bone Miner Res 19: 782793.
    Direct Link:
  • 11
    Krieg MA, Cornuz J, Ruffieux C, Sandini L, Büche D, Dambacher MA, Hartl F, Hauselmann HJ, Kraenzlin M, Lippuner K, Neff M, Pancaldi P, Rizzioli R, Tanzi F, Theiler R, Tyndall A, Wimpfheimer K, Burckhardt P 2003 Comparison of three bone ultrasounds for the discrimination of subjects with and without osteoporotic fractures among 7562 elderly women. J Bone Miner Res 18: 12611266.
  • 12
    López-Rodríguez F, Mezquita-Raya P, de Dios Luna J, Escobar-Jiménez F, Muñoz-Torres M 2003 Performance of quantitative ultrasound in the discrimination of prevalent osteoporotic fractures in a bone metabolic unit. Bone 32: 571578.
  • 13
    Frost ML, Blake GM, Fogelman I 2001 Does the combination of quantitative ultrasound and dual-energyX-ray absorptiometry improve fracture discrimination? Osteoporos Int 12: 471477.
  • 14
    Frost ML, Blake GM, Fogelman I 2002 A comparison of fracture discrimination using calcaneal quantitative ultrasound and dual X-ray absorptiometry in women with a history of fracture at sites other than the spine and hip. Calcif Tissue Int 71: 207211.
  • 15
    Bauer DC 1999 Clinical applications of quantitative ultrasound. NjehCF, HansD, FuerstT, GlüerCC, GenantHK (eds.) Quantitative Ultrasound: Assessment of Osteoporosis and Bone Status. Martin Dunitz, London, UK, 284297.
  • 16
    Porter RW, Miller CG, Grainger D, Palmer SB 1990 Prediction of hip fracture in elderly women: A prospective study. BMJ 301: 638641.
  • 17
    Miller PD, Siris ES, Barrett-Connor E, Faulkner KG, Wehren LE, Abbott TA, Chen YT, Berger ML, Santora AC, Sherwood LM 2002 Prediction of fracture risk in postmenopausal white women with peripheral bone densitometry: Evidence from the National Osteoporosis Risk Assessment. J Bone Miner Res 17: 22222230.
  • 18
    Hans D, Dargent-Molina P, Schott AM, Sebert JL, Cormier C, Kotzki PO, Delmas PD, Pouilles JM, Breart G, Meunier PJ 1996 Ultrasonographic heel measurements to predict hip fracture in elderly women: The EPIDOS prospective study. Lancet 348: 511514.
  • 19
    Hans D, Schott AM, Duboeuf F, Durosier C, Meunier PJ, on behalf of the EPIDOS Group. 2004 Does follow-up duration influence the ultrasound and DXA prediction of hip fracture? The EPIDOS prospective study. Bone 35: 357363.
  • 20
    Schott AM, Kassaï Koupaï B, Hans D, Dargent-Molina P, Ecochard R, Bauer DC, Breart G, Meunier PJ 2004 Should age influence the choice of quantitative bone assessment technique in elderly women? The EPIDOS study. Osteoporos Int 15: 196203.
  • 21
    Lee SH, Dargent-Molina P, Bréart G, for the EPIDOS Group. 2002 Risk factors for fractures of the proximal humerus: Results from the EPIDOS prospective study. J Bone Miner Res 17: 817825.
  • 22
    Altman D, Egger M, Pocock S, Vandenbroucke JP 2005 STROBE Statement: STrengthening the Reporting of OBservational studies in Epidemiology. Version 3Available online at Accessed October 18, 2005.
  • 23
    DerSimonian R, Laird N 1986 Meta-analysis in clinical trials. Control Clin Trials 7: 177188.
  • 24
    Villar J, Mackey ME, Carroli G, Donner A 2001 Meta-analyses in systematic reviews of randomized controlled trials in perinatal medicine: Comparison of fixed and random effects models. Stat Med 20: 36353647.
  • 25
    Brockwell SE, Gordon IR 2001 A comparison of statistical methods for meta-analysis. Stat Med 20: 825840.
  • 26
    Higgins JPT, Thompson SG, Deeks JJ, Altman DG 2003 Measuring inconsistency in meta-analyses. BMJ 327: 557560.
  • 27
    Higgins JPT, Thompson SG 2002 Quantifying heterogeneity in a meta-analysis. Stat Med 21: 15391558.
  • 28
    Egger M, Smith GD, Schneider M, Minder C 1997 Bias in meta-analysis detected by a simple, graphical test. BMJ 315: 629634.
  • 29
    Macaskill P, Walter SD, Irwig L 2001 A comparison of methods to detect publication bias in meta-analysis. Stat Med 20: 641654.
  • 30
    Heaney RP, Avioli LV, Chesnut CH III, Lappe J, Recker RR, Brandenburger GH 1995 Ultrasound velocity through bone predicts incident vertebral deformity. J Bone Miner Res 10: 341345.
  • 31
    Bauer DC, Glüer CC, Cauley JA, Vogt TM, Ensrud KE, Genant HK, Black DM 1997 Broadband ultrasound attenuation predicts fractures strongly and independently of densitometry in older women. A prospective study. Arch Intern Med 157: 629634.
  • 32
    Mele R, Masci G, Ventura V, de Aloysio D, Bicocchi M, Cadossi R 1997 Three-year longitudinal study with quantitative ultrasound at the hand phalanx in a female population. Osteoporos Int 7: 550557.
  • 33
    Thompson PW, Taylor J, Oliver R, Fisher A 1998 Quantitative ultrasound (QUS) of the heel predicts wrist and osteoporosis-related fractures in women age 45-75 years. J Clin Densitom 1: 219225.
  • 34
    Huang C, Ross PD, Yates AJ, Walker RE, Imose K, Emi K, Wasnich RD 1998 Prediction of fracture risk by radiographic absorptiometry and quantitative ultrasound: A prospective study. Calcif Tissue Int 63: 380384.
  • 35
    Pluijm SMF, Graafmans WC, Bouter LM, Lips P 1999 Ultrasound measurements for the prediction of osteoporotic fractures in elderly people. Osteoporos Int 9: 550556.
  • 36
    Stewart A, Walker LG, Porter R, Reid DM, Primrose WR 1999 Predicting a second hip fracture. The potential role of dual X-ray absorptiometry, ultrasound and other risk factors in targeting of preventive therapy. J Clin Densitom 2: 363370.
  • 37
    Gnudi S, Ripamondi C, Malavolta N 2000 Quantitative ultrasound and bone densitometry to evaluate the risk of nonspine fractures: A prospective study. Osteoporos Int 11: 518523.
  • 38
    McGrother CW, Donaldson MMK, Clayton D, Abrams KR, Clarke M 2002 Evaluation of a hip fracture risk score for assessing elderly women: The Melton Osteoporotic Fracture (MOF) study. Osteoporos Int 13: 8996.
  • 39
    Huopio J, Kröger H, Honkanen R, Jurvelin J, Saarikoski S, Alhava E 2004 Calcaneal ultrasound predicts early postmenopausal fractures as well as axial BMD. A prospective study of 422 women. Osteoporos Int 15: 190195.
  • 40
    Krieg MA, Cornuz J, Ruffieux C, Burckhardt P 2004 Rôle des ultrasons osseoux pour la prédiction du risque de fracture de la hanche chez la femme âgée de 70 ans ou plus: Résults de l'étude SEMOF et comparaison avec les données de la littérature. Rev Med Suisse Rom 124: 5962.
  • 41
    Krieg MA, Cornuz J, Ruffieux C, Burckhardt P 2003 Prediction of non vertebral fracture risk by heel ultrasound in a population of 7,585 swiss women aged 70 to 80 years. Calcif Tissue Int 72: O43.
  • 42
    Khaw KT, Reeve J, Luben R, Bingham S, Welch A, Wareham N, Oakes S, Day N 2004 Prediction of total and hip fracture risk in men and women by quantitative ultrasound of the calcaneus: EPIC-Norfolk prospective population study. Lancet 363: 197202.
  • 43
    Díez-Pérez A, González-Macías J, Vila J, Gómez-Martín F, Guerrero T, Rodríguez-Roca G, Montero JJ, Fortea MA, Nieto I, Arqueros P, Marín F 2005 Prediction of non-vertebral fracture risk in 5201 postmenopausal women by quantitative ultrasound of the calcaneus: 36-month final results from the ECOSAP prospective study. J Bone Miner Res 20: S1; S67.
  • 44
    Welch A, Camus J, Dalzell N, Oakes S, Reeve J, Khaw KT 2004 Broadband ultrasound attenuation (BUA) of the heel bone and its correlates in men and women in the EPIC-Norfolk cohort: A cross-sectional population-based study. Osteoporos Int 15: 217225.
  • 45
    Glüer CC, Hans D 1999 How to use ultrasound for risk assessment: A need for defining strategies. Osteoporos Int 9: 193195.
  • 46
    Marín F, López-Bastida J, Díez-Pérez A, Sacristán JA 2004 Bone mineral density referral for dual-energy absorptiometry using quantitative ultrasound as a prescreening tool in postmenopausal women from the general population: A cost-effectiveness analysis. Calcif Tissue Int 74: 277283.
  • 47
    Marshall D, Johnell O, Wedel H 1996 Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 312: 12541259.
  • 48
    Barr RJ, Adebajo A, Fraser WD, Halsey JP, Kelsey C, Stewart A, Reid DM 2005 Can peripheral DXA measurements be used to predict fractures in elderly women living in the community? Osteoporos Int 16: 11771183.
  • 49
    Cheng S, Hans D, Genant HK 1999 Calcaneal quantitative ultrasound systems: Gel-coupled. NjehCF, HansD, FuerstT, GlüerCC, GenantHK (eds.) Quantitative Ultrasound: Assessment of Osteoporosis and Bone Status. Martin Dunitz, London, UK, 125144.
  • 50
    Kanis JA, Black D, Cooper C, Dargent P, Dawson-Hughes B, De Laet Delmas P, Eisman J, Johnell O, Jonsson B, Melton L, Oden A, Papapoulos S, Pols H, Rizzioli R, Silman A, Tenenhouse A 2002 A new approach to the development of assessment guidelines for osteoporosis. Osteoporos Int 13: 527536.
  • 51
    Kanis JA, Borgstrom F, De Laet C, Johansson H, Johnell O, Jonsson B, Oden A, Zethraeus N, Pfleger B, Khaltaev N 2005 Assessment of fracture risk. Osteoporos Int 16: 581589.
  • 52
    Glüer CC, Wu CY, Jergas M, Goldstein SA, Genant HK 1994 Three quantitative ultrasound parameters reflect bone structure. Calcif Tissue Int 55: 4652.
  • 53
    Morris R, Mazess RB, Trempe J, Hanson JM 1998 Stiffness compensates for effects in ultrasonometry. J Clin Densitom 1: 104.
  • 54
    Seeley DJ, Browner WS, Genant HK, Nevitt MC, Scott JC, Cummings SR 1991 Which fractures are associated with low appendicular bone mass in elderly women? Ann Intern Med 115: 837842.
  • 55
    Bauer DC, Glüer CC, Genant HK, Stone K for the Fracture Intervention Trial Research Group 1995 Quantitative ultrasound and vertebral fracture in postmenopausal women. J Bone Miner Res 10: 353358.
  • 56
    Stegman MR, Davies KM, Heaney RP, Recker RR, Lappe JM 1996 The association of patellar ultrasound transmissions and forearm densitometry with vertebral fracture, number and severity: The Saunders County Bone Quality Study. Osteoporos Int 6: 130135.
  • 57
    Hartl F, Tyndall A, Kraenzlin M, Bachmeier C, Gückel C, Senn U, Hans D, Theiler R 2002 Discriminatory ability of quantitative ultrasound parameters and bone mineral density in a population-based sample of postmenopausal women with vertebral fractures: Results of the Basel Osteoporosis Study. J Bone Miner Res 17: 321330.
  • 58
    Scheneider J, Bundschuh B, Späth C, Landkammer C, Müller H, Sommer U, Gotz M, Nawroth P, Heilmann P, Kasperk C 2004 Discrimination of patients with and without vertebral fractures as measured by ultrasound and DXA osteodensitometry. Calcif Tissue Int 74: 246254.
  • 59
    Wünster C, Albanese C, de Aloysio D, Duboeuf F, Gambacciani M, Gonnelli S, Glüer CC, Hans D, Joly J, Reginster JY, de Terlizzi F, Cadossi R 2000 Phalangeal osteosonogrammetry study: Age-related changes, diagnostic sensitivity, and discrimination power. J Bone Miner Res 15: 16031614.
  • 60
    Gerdhem P, Magnusson H, Karlsson MK, Akesson K 2002 Ultrasound of the phalanges is not related to a previous fracture. J Clin Densitom 5: 159166.
  • 61
    Ekman A, Michaëlsson K, Petrén-Mallmin M, Ljunghall S, Mallmin H 2001 DXA of the hip and heel ultrasound but not densitometry of the fingers can discriminate female hip fracture patients from controls: A comparison between four different methods. Osteoporos Int 12: 185191.
  • 62
    Cauley JA, Lui LY, Ensrud KE, Zmuda JM, Stone KL, Hochberg MC, Cummings SR 2005 Bone mineral density and the risk of incident nonspinal fractures in black and white women. JAMA 293: 21022108.