HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lentle, B. C. Articles by Prior, J. C. Search for Related Content PubMed PubMed Citation Articles by Lentle, B. C. Articles by Prior, J. C.

Osteoporosis: What a Clinician Expects to Learn from a Patient’s Bone Density Examination¹

¹ From the Departments of Radiology and Medicine, University of British Columbia, Children’s and Women’s Health Centre of BC, and Vancouver Hospital and Health Sciences Centre, Canada. Received February 12, 2002; revision requested March 2; revision received August 20; accepted August 21. Address correspondence to B.C.L., 7997 Turgoose Terr, Saanichton, British Columbia, Canada V8M 1V4 (e-mail: blentle@shaw.ca).

	ABSTRACT

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
Osteoporosis has lately become recognized as an important disease on two accounts. On one hand, demographic change has resulted in a greatly increased and increasing burden of morbidity and mortality due to osteoporotic fracturing. On the other hand, lifestyle changes and preventive measures have become recognized as important factors in prevention of both osteoporosis and osteoporotic fractures, while several effective drug treatments have recently become available to treat osteoporosis by increasing bone density and reducing fracture incidence. Because bone density is, with age, the best predictor of fracture risk, its measurement has become central to the care of those potentially at risk. When a clinician refers a person for a bone density examination, the clinician should be concerned less with an "imaging diagnosis" than with the requirement that the laboratory has procedures in place for rigorous quality assurance and precision measurements, as well as for education of the staff involved. Implementation of these measures and an understanding of their clinical relevance in diagnosis and follow-up, as well as communication with clinicians in this context, are more important than any diagnostic insight that might be provided by "interpreting" a bone density study.

© RSNA, 2003

Index terms: Bones, absorptiometry, 30.1299, 40.1299 • Bones, diseases, 30.562, 40.562 • Osteoporosis, 30.562, 40.562 • Subtraction, dual-energy, 30.1299, 40.1299

	INTRODUCTION

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
Osteoporosis is a word used in two ways—as a description of a characteristic of some bone and as a diagnosis to identify a disease. When used descriptively, the word means, literally, "porous bones," and there are many causes of such increased porosity. Glucocorticoids (steroids) and other medications, Cushing disease, prolonged inactivity, and nutritional deficiencies are a few of the causes of secondary osteoporosis (1). When used without qualification, the term osteoporosis means a diagnosis of the disease: osteoporosis defined in terms of bone density as 2.5 SDs (T scores) below peak bone mass (2). The word osteopenia, often used in the past by radiologists as a descriptive term, has been preempted by a World Health Organization (WHO) Working Group to define a particular level of bone density short of overt osteoporosis and is now best kept for use in this specific sense (see below) (2).

The measurement of bone density does not discriminate among the causes of altered density. However, the measurement is most often undertaken in the context of hypogonadal osteoporosis (ie, that occurring as a result of amenorrhea or after menopause in women or because of inadequate serum testosterone concentration in men), and that clinical disorder is discussed in this article.

The recent past has seen a change in the perception of osteoporosis. In large degree, this is due to recently developed drugs that increase bone density and reduce fracture incidence in individuals at risk for osteoporotic fractures (3,4). Osteoporosis is thus no longer regarded as an inevitable result of aging but as a preventable and treatable disorder. At the same time, osteoporosis is no longer diagnosed only when symptomatic fractures result. Indeed, as many as 60% of vertebral fractures occur without symptoms. Unless a person has a fracture, osteoporosis usually causes no symptoms at all. So a clinically meaningful evaluation that would facilitate prevention and therapy must depend on a definition that does not require that bones fracture. It is in the context of estimation of the risk of fracture that bone densitometry serves an important clinical role. This risk is age dependent, in that a given degree of bone loss results in a different cumulative lifetime fracture risk when a person aged 40 years is compared with one aged 75, for example.

A more imaginative definition of osteoporosis has been developed that allows recognition of the disease before fractures result—namely, "a chronic progressive disease characterized by low bone mass and microarchitectural deterioration of bone tissue, which leads to bone fragility and a consequent increase in fracture risk" (5,6).

Of the two variables identified, bone mass is the only one that can now be readily measured in life (7–11). That is the role of bone densitometry or other bone measurement methods. There is, however, considerable research interest in the examination of bone microarchitecture, bone fatigue damage, and cancellous bone connectivity. Potentially useful tools for application in clinical practice will likely emerge from such investigations (11).

Formerly, osteoporosis was diagnosed only when a low-trauma fracture resulted. A working group of the WHO (2) has more recently further defined osteoporosis in menopausal white women, for epidemiologic purposes, on the basis of bone mineral density (BMD) measurements, as follows: (a) normal: a value for BMD or bone mineral content (BMC) within 1 SD (score of 1T) of the young-adult reference mean; (b) low bone mass (osteopenia): a value for BMD or BMC lower than 1 SD (<1T) below the young-adult mean but not as low as 2.5 SDs (2.5T) below this value; (c) osteoporosis: a value for BMD or BMC 2.5 SDs or more (2.5T) below the young-adult mean; (d) severe (established) osteoporosis: a value for BMD or BMC 2.5 SDs or more below the young-adult mean in the presence of one or more fragility fractures.

Notably, these definitions use bone density to categorize those at risk. While never intended as a series of intervention thresholds, they have become nearly universally used in the investigation and treatment of osteoporosis. As a result, the definitions have become the focus for communicating densitometric findings to a referring physician. However, it is important to realize that the definitions given were specifically intended to apply to menopausal white women. They do not necessarily apply to all sites measurable or all techniques that might be used (2).

The use of these definitions to reflect the risk of fracture in men or nonwhite persons or even in women before menopause, is at best an extrapolation, unsupported by firm evidence. Equally, use of the definitions is quite invalid in children or those who have not achieved or who will not, by virtue of skeletal disease (eg, types of osteogenesis imperfecta, Hadju-Cheney disease), ever achieve peak bone mass.

Bone density has a Gaussian, or normal bell-shaped, distribution in both women and men at any given age. However, the definition of osteoporosis cited earlier (5,6) makes reference to "low bone mass" while ignoring the fact that bone mass is a continuous variable. Therefore, any classification of osteoporosis based on bone density alone—such as that described by the WHO Working Group (2)—must be to some degree arbitrary. The purpose of a bone density measurement cannot be to "diagnose" osteoporosis as one might diagnose cancer by using a mammogram—an all-or-nothing finding—rather, the purpose is to contribute to the assessment of fracture risk, which is itself a continuous variable.

Ideally, knowledge of a person’s bone density should be but one part, if a very important one, of a global osteoporosis risk assessment otherwise derived from such factors as age, family history of kyphosis or low-trauma fracturing, or personal history of low-trauma fracture (considered a fragility fracture if trauma is due to a fall that is equivalent to or less than that from a standing height). A history of a late menarche or irregular menstrual cycle in women, of alcohol abuse in men, or of smoking in either sex is also part of the assessment of osteoporosis risk. It is the scale of this risk that will guide advice about fall avoidance, lifestyle modification, or treatment interventions. These risks of fracture need to be communicated, therefore, to a referring physician.

It should be noted that osteoporosis, while more common among women, is increasingly being recognized as a disease in men as well. For example, if baseline vertebral deformities, determined by using quantitative criteria, are assumed to represent osteoporotic fractures, then, in a Canadian population-based sample of persons older than 50 years (12), the prevalence in men (21.5%) is similar to that in women (23.5%).

	TOOLS FOR BONE MEASUREMENT

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
It is possible to measure bone in both the central and the peripheral skeleton by using an array of methods (7–11). These methods are summarized in the following sections.

Quantitative Computed Tomography
A software package available for computed tomography (CT) devices allows the conversion of relative Hounsfield units, determined in the center of a vertebral body, to an absolute measurement of bone density of calcium in milligrams per cubic centimeter. This is usually accomplished by including standards containing known amounts of hydroxyapatite or other test material in the field of view. By using measurements of the Hounsfield units of these standards, a calibration curve is constructed and the absolute volumetric density of the vertebra being examined is calculated. A further increase in accuracy may be achieved by measuring at two x-ray energies and thus adjusting, on the basis of differential absorption, for the amount of soft tissue in the tissue volume examined (7). Care is necessary in setting the cursor to avoid partial-volume averaging artifacts if endplate or disk is included in the volume interrogated.

Because of the expense, radiation dose, and other calls on (ie, demands for other examinations with) large-aperture CT machines, very-small-aperture alternatives for peripheral quantitative CT have been built and are available commercially. These machines may be used to measure skeletal sites in the extremities (eg, distal radius, calcaneus) in the same way.

Single-Photon Absorptiometry
Although little used in North America at present, single-photon absorptiometry was used to advance bone measurement from the early days of measurement of bone size on radiographs of the hand or crude determinations of optical density from similar images. It is an effective technique for measurements of bone in the distal radius and ulna. Much of our current knowledge of osteoporosis and fracture risk was gained with this method (13). However, since the morbidity and mortality resulting from osteoporosis arise primarily from fractures of the central skeleton, the challenge of measuring bone in the proximal femur and spine prompted developments in technology.

Dual-Photon Absorptiometry and Dual X-ray Absorptiometry
Measurement of the BMC of spine and proximal femur (or any part or all of the skeleton) requires measurement of the relative attenuation of two differing photon energies to permit a correction for soft-tissue attenuation. This allows an assay of the calcium content in deeper structures, although the technique only provides an areal density of calcium (in grams per square centimeter) (7), not a true volumetric density such as may be achieved with quantitative CT (11). Standard dual x-ray absorptiometric (DXA) spinal measurements are performed in the posteroanterior projection, and both spine and proximal femur are measured most commonly (Figs 1–4).

View larger version (104K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Posteroanterior DXA image of L1 through L4 vertebrae. The image is not used for "diagnosis" but to monitor patient positioning, region-of-interest placement, and potential presence of artifact due, for example, to discogenic sclerosis.

View larger version (71K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Lateral DXA image obtained for spinal morphometry or vertebral deformity assessment. Recognition of prevalent fractures is a powerful predictor of further fracturing (24). Note anterior wedge deformity of what is probably T11 in a patient with T score indicative only of osteopenia.

View larger version (71K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Lateral DXA image reveals evidence of multiple mixed vertebral endplate and anterior compression fracturing, although the entire spine is not depicted.

View larger version (158K): [in this window] [in a new window] [Download PPT slide]   Figure 4. On posteroanterior DXA image, the lesser trochanter is not seen in its entirety, indicating proper internal rotation of the thigh for analysis of the femoral neck at its greatest long-axis measurement. Regions of interest (ROIs) in the proximal femur are applied as defaults by analytic software. Like those in lumbar spine, ROIs may be overridden by the operator, but this should be an unusual operation. It must also be recognized that ROIs are not identical between manufacturers, making comparisons imprecise, particularly as regards the proximal femur. Outer rectangle defines global ROI (often described as "total hip"), narrow transverse rectangle orthogonal to the axis of symmetry constitutes the femoral neck, while the Ward "triangle" is the area of least density in a smaller area within a search area centered on the midline at the inferior edge of the femoral neck rectangle. The longer line is constructed along the axis of symmetry of the femoral neck; it is used in defining the ROI and again may be subject to manual override. A shorter subsidiary line separates "trochanteric region" from "intertrochanteric region"; it uses the Ward triangle as a starting point and is drawn to the base of the greater trochanter. Among these ROIs, femoral neck or "total hip" are best used for risk estimation, while "total hip" is preferred for follow-up because precision is greater given the larger sample size. There is no evidentiary basis to support use of the Ward triangle for risk estimation, and many clinicians might prefer to see this measurement eliminated.

The methods used to provide the two photon energies involve either a pulsed change in peak voltage or the insertion of a cerium or samarium filter into the primary beam. Different manufacturers use one method or the other. The filter reduces "soft" radiation but also adds characteristic radiation to produce the two photon energies, the second determined by the peak voltage across the tube.

Quantitative Ultrasonography of Bone
The conceptual basis for using ultrasound transmission to interrogate bone is still poorly understood. Nevertheless, at a practical level both the speed of ultrasound through bone and the attenuation of the signal have been found to correlate well with DXA measurements of bone density (14–17) and, at least in some populations, to predict fracture risk (18). However, although there has been a large increase in the use of quantitative ultrasonography (US) recently, the data produced are machine specific and subject to the use of dissimilar databases by different manufacturers. At the same time, anthropometric (19), clinical (eg, ankle edema) (20), and other variables are potential causes of error. There is also evidence that the choice of right or left heel for measurement can influence the result unduly (21).

Radiographic Absorptiometry
Radiographic absorptiometry involves digitization and computed analysis of hand radiographs acquired with inclusion of a standardized wedge used to calibrate bone density. Accuracy and precision are reported to be excellent (22), but outcome studies are lacking. More recently, a machine has been marketed that automates this analysis, eliminating the need to send images for central analysis. The advent of computed radiography may facilitate use of this method in centers without access to other technologies. However, the evidentiary basis for its adoption is modest (22,23).

Lateral Radiographs of Thoracic and Lumbar Spine
It is easy to overlook conventional radiographs in cases of osteoporosis, given the other technologies available. However, recognition of prevalent fractures implies a risk of further fracturing of the spine or hip. Fracture risk in an individual with an "asymptomatic vertebral deformity" is at least as great as that represented by 1 SD (-1T) of bone mass below peak (24). In appropriate patients, therefore, spinal radiographs may be helpful.

Surprisingly, there is still debate about what constitutes a vertebral fracture, since there are no good prospective studies of variations in vertebral morphology. Nevertheless, the identification of vertebral fractures is very important in risk estimation. As noted, the WHO Working Group defined severe osteoporosis as a bone density score of -2.5T or less and a prevalent low-trauma fracture. Indeed, the epidemiologic data suggest that one or more low-trauma fractures should be regarded as of osteoporotic provenance irrespective of bone density (25). The radiologist is thus in a strong position to guide the clinician in evaluating fracture risk.

Combined DXA and Morphometry
Some densitometers are now constructed to allow imaging of the spine in the lateral projection (26). Thus, measurements of vertebral morphometry may be made in lateral projections of the segments from T4 to L4 inclusively (26) (Fig 3). (It is usually thought that fractures above T4 are unlikely to be of osteoporotic origin, while the shape of L5 is subject to great morphologic variation and is difficult to assess quantitatively.) This may be accomplished by constructing the DXA machine as a rotating C arm or by moving the patient into a lateral decubitus position. Lateral spinal morphometry is combined with posteroanterior or lateral lumbar densitometry. Lateral densitometry may be used to assess BMC or BMD in the body of the vertebra and thus eliminate the contribution from mineral in the predominantly cortical bone in the posterior vertebral complex. This fact is offset by potential contributions from mineral in ribs or iliac crests, however. Also, there is a loss of precision due to the greater soft-tissue thickness encountered when measuring BMC or BMD in the coronal plane as compared with measurement in the sagittal plane.

The clinician needs to know the site at which bone is assayed and should be advised that any follow-up required is best performed with the same machine and in the same season. Differences exist between the measurements made with machines of different manufacturers at the same site(s), and even smaller differences exist between any two machines made by the same manufacturer. Such small differences become important when it is realized that physiologic and therapeutic changes in mineralization may amount to no more than 0.5% per year. Likewise, there are winter losses and summer gains in BMD, especially in more northerly latitudes, of about 1%.

	THE ROLE OF BONE DENSITOMETRY

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
The power of bone density measurements thus lies in the fact that, after age and the finding of a prevalent low-trauma fracture, such measurements are the most effective predictors of fracture risk (13,27–31). Moreover, in a comparative analysis, Grampp et al (32) found that most bone measurement methods, central and peripheral, are effective for tracking age-related changes in bone or for correlating with the presence of prevalent fractures. While bone densitometry is, to some degree, site specific (in that density measured, for example, in the spine best reflects the risk of spinal fracturing), it is nevertheless true that any site measured does provide an index of global risk. This is an important concept to relate to the referring clinician.

DXA is also particularly effective in that its greater precision allows more effective use in follow-up (33) than do peripheral techniques. Thus, other applications of DXA are in the sequential examination of patients being treated for osteoporosis or of those at increased risk by virtue of factors such as medication or late perimenopause that adversely affect bone. At present, DXA is the most widely used technique for performing bone measurement in North America.

There is a trend toward increasing complexity of DXA machines. On one hand, the use of fan-beam rather than pencil-beam technology permits more rapid scanning, albeit at the cost of a small increase in radiation exposure. On the other hand, some machines now have the capacity for both lateral densitometry and morphometry, as discussed above (25).

	CONTRAINDICATIONS TO DXA

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
Referring physicians need to be aware that the presence of barium or recent administration of radionuclides (within 10 half-lives of the injected tracer) may yield inaccurate BMD measurements (34). Barium in the transverse colon will cause considerable artifact in measurements of lumbar spine BMD but is usually apparent on the images. Severe fractural deformity or degenerative changes at the site to be measured are also potential causes of artifact (35–38). Radiopaque implants in the measurement area, most commonly at the hip, as well as a patient’s inability to remain motionless for the duration of the examination, are further sources of inaccuracy. Additional causes of an inaccurate DXA study are an inability to straighten the lumbar spine or internally rotate the thigh. In cases of both marked obesity and extremely low BMD, the operation of machine software may be less effective, although use of a slower scanning speed is recommended (see below). Such constraints need to be conveyed to the referring physician, together with advice on alternative strategies for risk estimation (eg, peripheral DXA, quantitative CT).

	SERIAL STUDIES AND PRECISION

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
As the number of people being treated for osteoporosis increases, so do the number of repeat examinations performed to assess treatment response. In such a context, it is important to be able to distinguish true or least significant change from random variation. Even when the machine used is the same, there are small variations in measurement from examination to examination due to machine-related factors (eg, changes in region-of-interest placement) and larger variations related to the positioning of the patient by the technologist (39). Thus, in densitometry (with any method of bone measurement), it is important to distinguish accuracy from precision. Accuracy is an expression of how well the results obtained with the machine reflect the true BMC; precision is a reflection of the reproducibility of repeat measurements. Limiting the conduct of DXA examinations and quality assurance measurements to a few skilled and dedicated technologists, when possible, is an important consideration in the operation of a densitometry service.

It is also important to recognize that reporting of T scores in patients undergoing treatment to build bone mass is not entirely valid and may mislead about the scale of fracture risk. The data from several trials (3,4,40) indicate that drugs such as the bisphosphonates and selective estrogen receptor modulators produce a drop in fracture rates in excess of that expected on the basis of any increase in BMD that results from treatment. The slopes of the curves plotted to relate fracture incidence to BMD may, therefore, not be identical for density decreases and treatment-related increases. The relationship between serum cholesterol level and vascular disease before and after treatment with the statins is similarly complex.

Within those constraints, precision is a measure of reproducibility, and in vivo precision is determined by means of repeat examinations with repositioning of the subject between examinations. Precision is distinct from accuracy, which is a measure of how well the machine reflects reality—in the case of DXA, the BMC or areal density of the bone assayed. The in vivo (short-term) precision of a DXA study of the lumbar spine has been quoted as 0.5%–1.5%, but these numbers are potentially misleading. Precision expressed as a percentage varies in absolute measurements (ie, in grams per square centimeter) through the observed range of BMDs. Therefore, it is better quoted in absolute rather than relative terms. For example, authors of studies on changes in bone density commonly reported percentages. Equally, the effectiveness of medications is often compared in terms of the percentage increase in bone density. But if one study included women with osteopenia (eg, mean spinal BMD of 1.00 g · cm^-2) who received medication A and another study included women with osteoporosis (eg, mean spinal BMD of 0.50 g · cm^-2) who received medication B, then a 3% mean increase in both groups would represent 0.03 g · cm^-2 in the first study but only 0.015 g · cm^-2 in the second.

A full treatment of the statistical basis for precision measurements is beyond the scope of this article. Suffice it to say that anyone working in the field should consult the protocols for determining precision, which are readily available (41). He or she should measure precision in his or her own laboratory, with each of the technologists performing repeat DXA examinations with repositioning of the patient, then calculate either the BMC or BMD value or the percentage on a sliding scale that constitutes the least significant change at 95% (or other) confidence limits. A rigorous approach to the understanding of precision and its application to the determination of change is probably the most important contribution to be made to improving the practice of bone densitometry and providing key information to referring physicians. A report of a follow-up DXA study must provide the absolute change (in grams per square centimeter), the percentage change, and a statement as to whether this meets the criterion for real or clinically important change (exceeding the value for least significant change).

The radiologist should also be aware of the expected rates of change in age- and sex-specific populations. Thus, if bone loss is expected (eg, in a menopausal woman starting steroid treatment), stabilization of bone may reflect a treatment response as much as an increase in BMD. For an elderly person whose BMD would be expected to decrease over time, prevention of that decrease is probably important in its own right.

Short-term in vivo DXA precision errors for posteroanterior lumbar spine (L1 through L4) and total proximal femoral studies are quoted to be about 1% and 1%–2% of normal densities, respectively (2), but these numbers relate to normal bone. Given the limited precision alluded to and the fact that rates of change in bone density are small (eg, 0.5% per year with aging), it is responsible at follow-up to recommend that examinations be repeated at most every 2nd year. Exceptions include cases where patients who are starting glucocorticoid (steroid) therapy, who have undergone orchidectomy, or who have undergone premenopausal ovariectomy (42–44). Several investigators (45,46) have found that the most sensitive site at which to examine change in BMD is the lumbar spine (on posteroanterior images). In addition, because of small seasonal variations in BMD, successive DXA examinations should ideally be performed in the same season. Most large databases, as well as results of reported trials (eg, 3,4) of osteoporosis medications, have not systematically corrected for such artifacts.

All of these constraints may need to be alluded to in reports of bone density examinations. Referring physicians, particularly those who have a particular interest in osteoporosis, will be familiar with the rigor required in the establishment of precision and, hence, in the determination of a significant change in BMD. They may be reassured by periodic updates about the measurements from a given laboratory by means of either a brief bulletin or face-to-face meetings outside of the constraints of communicating individual patient data.

Confounding Variables
BMD may be increased artifactually in posteroanterior measurements of the lumbar spine; reasons for this include the presence of osteophytes, aortic calcifications, degenerative facet joint disease with hypertrophy and reactive bone, and degenerative disk disease (35–38). These are important considerations, especially in elderly patients, since DXA measures the calcific density projected in the area of interest defined by the software (or selected by the operator). The area may thus sometimes include such "nonstructural" mineralization. Orwoll et al (35), in particular, demonstrated that in men the consequence of osteophytes can lead to misleading DXA results. Thus, radiographic correlation may be helpful in the assessment of artifact.

Similarly, since fractures reduce the projected area of a vertebra, they commonly result in an apparent increase in density, as shown by Ryan et al (38). It is helpful to examine individual segments and eliminate measurements out of step with adjacent segments. (Moving distally from L1 to L4, there is usually a small and steady increase in density from vertebra to vertebra.) However, as Ryan et al pointed out, fractures themselves cannot be reliably identified from the usual posteroanterior DXA image.

However, a major factor in the accuracy and precision of densitometers is the education and application of the technologists concerned (39) and, as noted, performance of the procedure by a few dedicated operators. Pedagogic programs such as those provided by the International Society of Clinical Densitometry are particularly relevant in this context.

The images accompanying a DXA report have a note specifically indicating that the images are not to be used for diagnostic purposes. Nevertheless, recognition of and reference to positioning errors or artifacts in the report and the choice of regions of interest (vertebrae analyzed) can help the clinician, as well as improve the quality of a DXA study. At the same time, however, it must be recognized that the fewer the vertebrae analyzed, the greater the potential for errors in accuracy and precision, such that any diagnostic inference based on one or two vertebral segments is not advised. For similar reasons, it is preferable to use segments L1 through L4 unless there is good reason to exclude L1.

Most standard DXA densitometers also allow accurate measurement of the radius or calcaneus by using regions of interest like those used in single-photon absorptiometry and single-energy absorptiometry measurements, as well as in user-defined subregions.

Referring physicians need to be made aware of the limitations of BMD measurements in the obese or if measurements are complicated by substantial weight gain or loss. BMD measured with DXA correlates positively with total body fat mass. Also, DXA measurement of the spine and BMD will be underestimated if the amount of fat overlying the spinal column is greater than that on either side of the lumbar spine (47,48). It has been reported (47) that a mean fat thickness difference of 2 cm results in a BMD error of 9%–10%. For some patients, this error was as large as 16%. A change in bone marrow fat of 50% will change BMD by 5%–6%. If we assume a change in yellow marrow from 40% at young-adult age to 60% at age 80 years, each year might result in a 1% error. Results of studies in cadavers (48) support these observations. Authors of more recent studies (48) of DXA measurement accuracy found similar errors for BMC, as well as for areal measurements, but smaller errors for BMD. The errors introduced by soft tissue and intramedullary fat limit the application of DXA to patients who have an abdominal anteroposterior diameter of 15–30 cm (49). Weight change does, therefore, have the potential to reduce the precision of BMD measurements.

Database
BMD measurements in and of themselves would be of little intrinsic value in the estimation of osteoporosis risk (although changes in density might be informative). Their value lies in the benchmarking of measurement with a comparison population. That, after all, is the basis for T scores. At present for the total proximal femur, this is almost uniformly the measurement made as part of the third generation of the U.S. National Health and Nutrition Epidemiological Study, or NHANES III (50). Nevertheless, it has been suggested that in an ideal world the appropriate database to use should be local and certainly take into account ethnic differences. Thus, recently completed baseline measurements from the Canadian Multicentre Osteoporosis Study allow use of vertebral and femoral neck peak BMD data that are specific to urban areas of Canada (51).

Components of a Report of Bone Measurement
Lenchik et al (52) have described a model reporting format for DXA, and Stock et al (53) have noted that effective reporting of densitometric findings favorably influences physician management of osteoporosis by family doctors. However, a recent survey found that a comment about fracture risk in particular was not included by as many as 30% of the reporting centers responding (54). A description of a sample reporting format is suggested in the Appendix.

The data that each manufacturer has programmed densitometers to provide with respect to each patient probably contains more information than most consumers need—knowledgeable specialists aside. However, the document will usually reflect BMD to three decimal places and, on occasion, T scores to two decimal places. This information should not to be communicated to the referring physician in the report. Such numbers far exceed the performance characteristics of the machine or (in the case of T scores) the conceptual basis of the WHO classification of the disease. Therefore, report BMD to at most two decimal places and T scores to one. To do otherwise is to misrepresent the quantitative limitations on the accuracy and precision of the technology or the arbitrariness of disease classifications in osteoporosis.

The printed output of bone densitometers includes the T score in addition to a low-resolution image. This convenience may result in the limitations of the densitometer sometimes being overlooked. The threshold T scores for diagnosis of osteopenia (-1.0T) and osteoporosis or severe osteoporosis (-2.5T) were arrived at by a WHO Working Group attempting to define osteoporosis for epidemiologic purposes in menopausal women of predominantly white extraction. As noted above, it would be inappropriate to report, without at least a caveat, T scores in persons not at an age at which peak bone mass might have been achieved or in nonwhite persons. Equally, while T scores are corrected for a database in men, it is not yet certain if the negative T scores derived have the same predictive power for fracturing as in women. For example, hip fractures are less common in women of Asian extraction than in those of white extraction, even for a given BMD value.

Purists should also note that T scores as used in osteoporosis risk assessment are slightly different from T scores as they are usually understood and applied in epidemiologic studies. The wider issue of the validity of the WHO Working Group classification (2) will be the subject of future debate.

Machine Used
Although densitometers all measure bone mass by using dual-energy x rays, which permits differential measurements of bone and soft-tissue attenuation, the machines are far from identical. The production of two energies may be achieved by rapidly alternating the voltage applied to the tube or by inserting a filter (cerium or samarium) to both reduce "soft" rays and add characteristic radiation from the filter material to the tube output. In addition, the assumptions built into the analytic software are different for each manufacturer, such that although results from a given make of densitometer are consistent, substantial systematic differences exist for measurements made between different machines fabricated by different manufacturers. Thus, as noted above, it is wise to indicate the machine used and to add a caveat to the effect that follow-up examinations are best performed with the same machine or, at the very least, one made by the same manufacturer. However, a standardized BMD has been developed in an attempt to address this issue, but it should be used only if absolutely necessary (50,55).

	DISCUSSION

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
There are enormous problems to be resolved in the evaluation of osteoporosis. Demographic changes, especially in aging populations, make it imperative that these problems be addressed. It has been estimated that in the United States, the economic burden of osteoporosis amounted to $13.8 billion in 1995, of which $8.6 billion was spent for inpatient care; $3.9 billion, for nursing home care; and $1.3 billion, for outpatient services (56). The relationship of peripheral to central bone measurements and the respective values of each are slowly becoming understood, but technologic diversification sometimes occurs more quickly than the technologies can be evaluated. Quantitative transmission US can be used to measure some aspect of bone that probably is not very different from bone density and may become more widely used (15). At present, however, quantitative US cannot be recommended for follow-up examinations because its precision appears to be substantially less than that of DXA. Also, at present, there is no database reflecting peak bone mass data nor a widely accepted phantom for standardization of values. Artifacts peculiar to quantitative US also need to be recognized (18–21). Moreover, the variety of machines available, the measurements that can be made with them, and the assumptions built into the analytic software are manufacturer specific.

Reports from the National Osteoporosis Foundation (31) and the Osteoporosis Society of Canada (57) point out that osteoporosis is defined in practice by an intermediate outcome (the BMD), not a health outcome (eg, fracture). The National Osteoporosis Foundation report (31) notes that "there is a strong association between BMD and the likelihood of fracture but ... other factors influence fracture risk as well. In this way, osteoporosis is similar to hypertension, diabetes mellitus, hypercholesterolemia, or atherosclerosis." Furthermore, the National Osteoporosis report (31) continues:

Because bone mass declines steadily after young adulthood, the prevalence of osteoporosis increases with age. For some women who are identified as osteoporotic, however, the risk of a fracture during their remaining lifetime is sufficiently low that treatment would not be appropriate (other than supplemental vitamin D and calcium). Conversely, many women who do not have osteoporosis might have other risk factors and circumstances that would justify treatment (such as when starting high dose prednisone therapy). Thus, although the definition of osteoporosis is based on BMD, it is not advisable to measure bone density in everyone. The decision to obtain a DXA examination should be based on each individual’s risk factors ([t]able 1) and the treatment being considered. In addition, testing is not indicated unless the results might influence a treatment decision.

The risk factors mentioned in table 1 of this report (31) are summarized in the Table.

	THE FUTURE

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
Ultimately, the "holy grail" of osteoporosis diagnosis will probably be a capacity to examine not only bone density but also the architecture of cancellous bone, the connectivity between trabeculae, and the existence of fatigue damage and its repair. The "microarchitectural deterioration" of bone referred to in the definition quoted above is hardly a precise concept (5,6). It is probably best understood, given our present knowledge, as decreases in the number and diameter of and connections between the trabeculae in cancellous bone, as well as the accumulation and incomplete repair of fatigue damage (59). These properties lend themselves to radiologic investigation in the broadest sense, so the opportunities for research are many.

For practical purposes, in the more immediate future we believe it is desirable that fracture risk estimation based on the patient’s history and physical, radiologic, and biochemical examination results be synthesized into a comprehensive index of risk. The data from which to deduce the precise magnitude of that risk, in either absolute or relative terms, are or are becoming available. These data serve as a guide to advising patients about lifestyle modification, hormone use at menopause, and treatment interventions when these are appropriate (5,6,9,29–31).

	CONCLUSION

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
The interpretation of individual DXA studies is not difficult. However, the responsibility of a physician overseeing a densitometry service lies more in familiarity with the conceptual context as it relates to the role of densitometry in and the management of osteoporosis, as described in this article. In addition, the physician must ensure that the densitometry service has procedures in place for rigorous quality assurance and precision measurements, as well as for education of the staff involved. Finally, interpretations must be carried out with these facts in mind. These represent the proper expectations of a referring physician far more than do any diagnostic insights that might be provided by "interpretation" of a BMD study.

	APPENDIX

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 
Figure A1 shows an example of a format for reporting DXA results. It may also be useful to append information such as the classification of osteoporosis described (2).

View larger version (46K): [in this window] [in a new window] [Download PPT slide]   Figure A1. DXA report template. NHANES III = third generation of the U.S. National Health and Nutrition Epidemiological Study.

	FOOTNOTES

 
Abbreviations: BMC = bone mineral content, BMD = bone mineral density, DXA = dual-energy x-ray absorptiometry, WHO = World Health Organization

	REFERENCES

TOP ABSTRACT INTRODUCTION TOOLS FOR BONE MEASUREMENT THE ROLE OF BONE... CONTRAINDICATIONS TO DXA SERIAL STUDIES AND PRECISION DISCUSSION THE FUTURE CONCLUSION APPENDIX REFERENCES

 

1. Khosla S, Melton LJ, III. Secondary osteoporosis. In: Riggs BL, Melton LJ, III, eds. Osteoporosis: etiology, diagnosis, and management. 2nd ed. Philadelphia, Pa: Lippincott-Raven, 1995; 183-204.
2. World Health Organization. Assessment of osteoporotic fracture risk and its role in screening for postmenopausal osteoporosis WHO Technical Report Series, no. 843. Geneva, Switzerland: World Health Organization, 1994; 5.
3. Harris ST, Watts N, Genant HK, et al. Effects of risedronate treatment on vertebral and nonvertebral fractures in women with postmenopausal osteoporosis: a randomized controlled trial. JAMA 1999; 282:1344-1352.[Abstract/Free Full Text]
4. Black DM, Thompson DE, Bauer DC, et al. Fracture risk reduction with alendronate in women with osteoporosis: the Fracture Intervention Trial. J Clin Endocrinol Metab 2000; 85:4118-4124.[Abstract/Free Full Text]
5. Miller PD, Bonnick SL, Rosen CJ, et al. Clinical utility of bone mass measurements in adults: consensus of an international panel. Semin Arthritis Rheum 1996; 25:361-372.[CrossRef][Medline]
6. Miller PD, Bonnick SL, Rosen CJ, et al. Consensus of an international panel on the clinical utility of bone mass measurements in the detection of low bone mass in the adult population. Calcif Tissue Int 1996; 58:207-214.[Medline]
7. Blake GM, Fogelman I. Technical principles of dual energy x-ray absorptiometry. Semin Nucl Med 1997; 27:210-228.[CrossRef][Medline]
8. Baran DT, Faulkner KG, Genant HK, Miller PD, Pacifici R. Diagnosis and management of osteoporosis: guidelines for the utilization of bone densitometry. Calcif Tissue Int 1997; 61:433-440.[CrossRef][Medline]
9. Compston JE, Cooper C, Kanis JA. Bone densitometry in clinical practice. BMJ 1995; 310:1507-1510.[Abstract/Free Full Text]
10. Blake GM, Gluer CC, Fogelman I. Bone densitometry: current status and future prospects. Br J Radiol 1997; S177-S186.
11. Genant HK, Engelke K, Fuerst T, et al. Noninvasive assessment of bone mineral and structure: state of the art. J Bone Miner Res 1996; 11:707-730.[Medline]
12. Jackson SA, Tenenhouse A, Robertson L. Vertebral fracture definition from population-based data: preliminary results from the Canadian Multicentre Osteoporosis Study (CaMos). Osteoporos Int 2000; 11:680-687.[CrossRef][Medline]
13. Duppe HR, Gardsell P, Nilsson B, Johnell O. A single bone density measurement can predict fractures over 25 years. Calcif Tissue Int 1997; 60:171-174.[CrossRef][Medline]
14. Njeh CF, Boivin CM, Langton CM. The role of ultrasound in the assessment of osteoporosis: a review. Osteoporos Int 1997; 7:7-22.[CrossRef][Medline]
15. Gregg EW, Ecriska AM, Salamone LM, et al. The epidemiology of quantitative ultrasound: a review of the relationships with bone mass, osteoporosis and fracture risk. Osteoporos Int 1997; 7:89-99.[CrossRef][Medline]
16. Rang C, Speller R. Comparison of ultrasound and dual energy x-ray absorptiometry measurements in the calcaneus. Br J Radiol 1998; 71:861-867.[Abstract]
17. Salamone LM, Ecrall EA, Harris S, Dawson-Hughes B. Comparison of broadband ultrasound attenuation to single x-ray absorptiometry measurements at the calcaneus in postmenopausal women. Calcif Tissue Int 1994; 54:87-90.[CrossRef][Medline]
18. Hans D, Dargent-Molina P, Schott AM. Ultrasonographic heel measurements to predict hip fracture in elderly women: the EPIDOS prospective study. Lancet 1996; 348:511-514.[CrossRef][Medline]
19. Hans D, Schott AM, Arlot ME, Sornay E, Delmas PD, Meunier PJ. Influence of anthropomorphic parameters on ultrasound measurements of os calcis. Osteoporos Int 1995; 5:371-376.[CrossRef][Medline]
20. Johansen A, Stone MD. The effect of ankle edema on bone ultrasound assessment at the heel. Osteoporos Int 1997; 7:44-47.[CrossRef][Medline]
21. Howard GM, Nguyen T, Pocock NA, Kelly PJ, Eisman JA. Influence of handedness on calcaneal ultrasound: implications for assessment of osteoporosis and study design. Osteoporos Int 1997; 7:190-194.[CrossRef][Medline]
22. Yang SO, Hagiwara S, Engelke K, et al. Radiographic absorptiometry for bone mineral measurement of the phalanges: precision and accuracy study. Radiology 1994; 192:857-859.[Abstract/Free Full Text]
23. Kleerekoper M, Nelson DA, Flynn MJ, et al. Comparison of radiographic absorptiometry with dual-energy x-ray absorptiometry and quantitative computed tomography in older white and black women. J Bone Miner Res 1994; 9:1745-1749.[Medline]
24. Ross PD, Davis JW, Epstein RS, Wasnich RD. Pre-existing fractures and bone mass predict vertebral fracture incidence in women. Ann Intern Med 1991; 114:19-23.[CrossRef]
25. Ross PD, Davis JW, Wasnich RD, Vogel JM. A critical review of bone mass and the risk of fractures in osteoporosis. Calcif Tissue Int 1990; 46:1149-1161.
26. Blake GM, Rea JA, Fogelman I. Vertebral morphometry studies using dual-energy x-ray absorptiometry. Semin Nucl Med 1997; 27:276-90.[CrossRef][Medline]
27. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996; 312:1254-1259.[Abstract/Free Full Text]
28. Cummings SR, Black DM, Nevitt MC, et al. Bone density at various sites for prediction of fractures. Lancet 1993; 341:72-75.[CrossRef][Medline]
29. Miller PD, Harris ST. Clinical applications of bone densitometry. In: Genant HK, Guglielmi G, Jergas M, eds. Bone densitometry in osteoporosis. Berlin, Germany: Springer-Verlag, 1998; 477-487.
30. Kanis JA. Assessment of bone mass in osteoporosis. Osteoporosis Cambridge, Mass: Blackwell Science, 1994; 114-147.
31. Eddy DM, Johnston CC, Cummings SR, et al. Osteoporosis: review of the evidence for prevention, diagnosis and treatment and cost-effectiveness analysis. Osteoporos Int 1998; 8(suppl 4):S1-S88.
32. Grampp S, Genant HK, Mathur A, et al. Comparisons of noninvasive bone mineral measurements in assessing age related loss, fracture discrimination, and diagnostic classification. J Bone Miner Res 1997; 12:697-711.[CrossRef][Medline]
33. Mazess R, Chestnut C, III, McClung M, Genant H. Enhanced precision with dual-energy x-ray absorptiometry. Calcif Tissue Int 1992; 51:14-17.[CrossRef][Medline]
34. Rosenthall L. Estimation of the effect of a preinjection of Tc-99m MDP on lumbar spine bone mineral density determinations. Clin Nucl Med 1992; 17:195-197.[Medline]
35. Orwoll ES, Oviatt SK, Mann T. The impact of osteophytic and vascular calcifications on vertebral mineral density measurements in men. J Clin Endocrinol Metab 1990; 70:1202-1207.[Abstract]
36. Jaovisidha S, Sartoris DJ, Martin EM, De Maeseneer M, Szollar SM, Deftos LJ. Influence of spondylopathy on bone densitometry using dual energy x-ray absorptiometry. Calcif Tissue Int 1997; 60:424-429.[CrossRef][Medline]
37. Rand T, Seidl G, Kainberger F, et al. Impact of spinal degenerative changes on the evaluation of bone mineral density with dual energy x-ray absorptiometry (DXA). Calcif Tissue Int 1997; 60:430-433.[CrossRef][Medline]
38. Ryan PJ, Evans P, Blake GM, Fogelman I. The effect of vertebral collapse on spinal bone mineral density. Bone Miner 1992; 18:267-272.[CrossRef][Medline]
39. Staron RB, Greenspan R, Miller TT, Bilezekian JP, Shane E. Computerized bone densitometric analysis: operator-dependent errors. Radiology 1999; 211:467-470.[Abstract/Free Full Text]
40. Ettinger B, Black DM, Mitlak BH, et al. Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene. JAMA 1999; 282:637-645.[Abstract/Free Full Text]
41. Glüer CC, Blake G, Lu Y, Blunt BA, Jergas M, Genant HK. Accurate assessment of precision errors: how to measure the reproducibility errors of bone densitometry techniques. Osteoporos Int 1995; 5:262-270.[CrossRef][Medline]
42. He YF, Ross PD, Davis JW, Epstein RS, Vogel JM, Wasnich RD. When should bone mass measurements be repeated? Calcif Tissue Int 1994; 55:243-248.[CrossRef][Medline]
43. Prior JC, Vigna YM, Wark JD, et al. Premenopausal ovariectomy-related bone loss: a randomized, double-blind one year trial of conjugated estrogen or medroxyprogesterone acetate. J Bone Miner Res 1997; 12:1851-1863.[CrossRef][Medline]
44. Prior JC. Perimenopause: the complex endocrinology of the menopausal transition. Endocr Rev 1998; 19:397-428.[Abstract/Free Full Text]
45. Jones CD, Laval-Jeantet AM, Laval-Jeantet MH, Genant HK. Importance of measurement of spongious vertebral bone mineral density in the assessment of osteoporosis. Bone 1987; 8:201-206.[Medline]
46. Blake GM, Herd RJM, Fogelman I. A longitudinal study of supine lateral DXA of the lumbar spine: a comparison with posteroanterior spine, hip and total body DXA. Osteoporos Int 1996; 5:462-470.
47. Reid IR, Ames R, Evans MC, et al. Determinants of total body regional bone mineral density in normal postmenopausal women: a key role for fat mass. J Clin Endocrinol Metab 1992; 75:45-51.[Abstract]
48. Bolotin HH. A new perspective on the causal influence of soft tissue composition on DXA-measured in vivo bone mineral density. J Bone Miner Res 1998; 13:1739-1746.[CrossRef][Medline]
49. Wahner HW, Fogelman I. Clinical bone density London, England: Dunitz, 1994; 88-120.
50. Simmons A, Simpson DE, O’Doherty MJ, Barrington S, Coakley AJ. The effects of standardization and reference values on patient classification for spine and femur. Osteoporos Int 1997; 7:200-206.[CrossRef][Medline]
51. Tenenhouse A, Joseph L, Krieger N, et al. Estimation of prevalence of low bone density in Canadian women and men using a population-specific DEXA reference standard: the Canadian Multicentre Osteoporosis Study (CaMos). Osteoporos Int 2000; 11:897-904.[CrossRef][Medline]
52. Lenchik L, Rochmis P, Sartoris DJ. Optimized interpretation and reporting of dual x-ray absorptiometry (DXA) scans. AJR Am J Roentgenol 1998; 171:1509-1520.[Free Full Text]
53. Stock JL, Waud CE, Coderre JA, et al. Clinical reporting to primary care physicians leads to increased use and understanding of bone densitometry and affects the management of osteoporosis: a randomized trial. Ann Intern Med 1998; 128:996-999.[Abstract/Free Full Text]
54. Fuleihan GE, Stock JL, McClung MR, Saifi G. A national random survey of bone mineral density reporting in the United States. J Clin Densitom 2002; 5:3-9.[CrossRef][Medline]
55. Hanson J. Standardization of femur BMD: International Committee for Standards in Bone Measurement. J Bone Miner Res 1997; 12:1316-1317.[CrossRef][Medline]
56. Ray NF, Chan JK, Thamer M, Melton LJ. Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: report from the National Osteoporosis Foundation. J Bone Miner Res 1997; 12:24-35.[CrossRef][Medline]
57. Sturtridge W, Lentle B, Hanley DA. Prevention and management of osteoporosis: consensus statements from the Scientific Advisory Board of the Osteoporosis Society of Canada. 2. The use of bone density measurement in the diagnosis and management of osteoporosis. CMAJ 1996; 55:924-929.
58. Prior JC, Vigna YM, Barr SI, Rexworthy C, Lentle BC. Cyclic medroxyprogesterone treatment increases bone density: a controlled trial in active women with menstrual cycle disturbances. Am J Med 1994; 96:521-530.[CrossRef][Medline]
59. Burr DR, Forwood MR, Fytirie DP, Martin RB, Schaffler MB, Turner CH. Bone microdamage and skeletal fragility in osteoporotic and stress fractures. J Bone Miner Res 1997; 12:6-15.[CrossRef][Medline]

This article has been cited by other articles:

				J. F. Griffith, D. K. W. Yeung, G. E. Antonio, S. Y. S. Wong, T. C. Y. Kwok, J. Woo, and P. C. Leung Vertebral Marrow Fat Content and Diffusion and Perfusion Indexes in Women with Varying Bone Density: MR Evaluation Radiology, December 1, 2006; 241(3): 831 - 838. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Submit a response