Diagnostic Dilemmas in the Use of BMD Testing

CASE 1: 40-Year-Old Woman With Lupus

The first patient case we will discuss concerns a 40-year-old woman who has been receiving steroid treatment for lupus for 20 years. Because of her increased risk for osteoporosis, she has been routinely given BMD testing by DXA at three-year intervals. Her current BMD test results indicate T-scores and Z-scores of -0.6 at the spine and -1.0 at the hip. She has experienced a wrist fracture after a fall and an atraumatic pelvis fracture occurred more recently.

Does this patient have osteoporosis?

According to WHO criteria, the patient’s T-scores do not classify her as having osteoporosis. It must be noted that WHO criteria are not intended for application to patients with any secondary cause of osteoporosis. However, her low-trauma fracture indicates she has clinical osteoporosis and her glucocorticoid steroid use indicates that she is indeed at risk for future fracture. This type of drug is destructive to bone. It decreases the rate of bone formation, increase the rate of bone breakdown, and cause death of bone cells, thus rapidly weakening its microarchitecture and increasing the risk for fracture. Regardless, 10% of postmenopausal women who suffer an osteoporotic fracture have normal BMD T-scores. Thus this patient appears to have osteoporosis with a false negative test result.

How can the clinician confirm a diagnosis of osteoporosis?

Densitometry testing alone will not always help. A history of fragility fracture is the most reliable predictor of the risk of future fracture, and can be used to make a diagnosis of osteoporosis. Clinical assessment here is more important than the patient’s T-score. We do not know the patient’s peak bone density, so one cannot tell if this hip T-score/ Z-score score is different from her baseline (after steroids) or preexisting (before steroids). If the loss is acute, it is more likely to result in perforation of trabecular structures and fragility that is out of proportion to the bone density.

The patient is treated for osteoporosis with a bisphosphonate. Her bone density increases over the years and no further fractures are noted. A diagnosis of osteoporosis can be made in the presence of a fragility fracture despite the presence of normal BMD in this instance.

CASE 2: 28-Year-Old Woman

The second patient we will discuss is a 28-year-old woman athlete and avid jogger of ten years’ standing. Her work as a radiographer for a medical group made it possible for her to receive a bone densitometry test, and so she had it done. Her T-scores on DXA were -2.5 at the spine and -0.1 at the hip. But the Z-scores were -0.5 and 0 respectively.

How should the clinician proceed?

The clinician runs a blood count, biochemical profile and serum 25-vitamin D. The results are normal. The patient’s physical exam indicates a 2" loss in height since her last yearly physical. The patient has had no menstrual irregularity or amenorrhea and has no known risk factors for osteoporosis.

Does this patient have a high risk for fragility fracture, or osteoporosis, as her bone density scores suggest?

According to the WHO standards, one may say yes, but one must recognize that these standards apply to postmenopausal women and are not directly applicable to premenopausal at other ages. This is particularly relevant for young women around the age of peak mass, the age whence T-scores are derived. Currently Z-scores are recommended for diagnosis of ‘low BMD’ in premenopausal women and men < 50 years of age. It is likely she has no increased risk of fragility fracture now, but has a lower peak bone mass than expected and healthy bone. However even if her relative risk is increased, because of her low BMD, her absolute risk of fracture is very low as age is almost as important for predicting risk as low BMD. Thus the risk of fracture for any given BMD is much lower for an otherwise similar woman in her 20s, than a 70+ year old woman. A complete history reveals no risk factors for osteoporosis: she is premenopausal, has had no fragility fractures, has no family history of osteoporosis, and has no underlying disease condition that would predispose her to bone loss. It is essential to keep in mind that osteoporosis is not just abnormal bone density. It is a state of chronic and progressive bone loss that is destructive to skeletal microarchitecture, leading to increased fracture risk. From her history of an active lifestyle and no fractures, it is apparent that the patient’s skeleton is healthy. Pharmacologic treatment for osteoporosis would be inappropriate. However, appropriate dietary intake of calcium and vitamin D are very important. These measures alone can stabilize her bone and help prevent loss due to aging.

What is to account for the patient’s height loss?

More than likely the patient’s height was measured incorrectly on her last physical. The clinician checks medical records for several years past and finds that this is the case. The patient has not lost height.

What is the cause of the patient’s low T- / Z-score?

The patient’s low bone mass may be the result of heredity, poor childhood nutrition, or childhood diseases that compromise bone growth. Studies show genotype accounts for 70-80% of peak bone mass, The T-score represents the patients BMD compared to a healthy population at peak bone mass in units of standard deviation (SD). Thus a T-score of -2.5 indicates her BMD at that site is 2.5 SD below mean peak bone mass. . It does not necessarily signify disease of ongoing bone loss and imminent fracture risk as it would in a postmenopausal woman. Her Z-score represents her BMD compared to a healthy age-matched population, similarly in SD. Thus one would expect her T-score and Z-score to be the same or similar in this age group irrespective of her BMD. However because of variations in definition, DXA manufacturer software and skeletal site, discrepancies occur when they should not and can confuse clinicians. The differences between her T-score and Z-score thus reflect calculation problems with these values due to lack of standardization not a ‘bone disease’ such as osteoporosis. It is important when faced with a situation like this to understand that there may be errors in calculation of these values when faced with results that do not make sense. Clinicians need to check which ‘T-score’ and ‘Z-score’ they are using on DXA machine.

Is the patient at increased risk for fracture?

Not at present. However, bone-density related problems may crop up in her earlier than in someone with higher peak bone mass. When she reaches menopause, the patient should be aware that existing low bone mass may predispose her to increased fracture risk with hormone-related bone loss.

How should the clinician manage the patient?

To confirm that the patient is actually not losing bone, the patient’s rate of bone turnover is assessed using biochemical markers like urine n-telopeptide or serum bone specific alakaline phosphatase. The results are normal. The patient is counseled to be aware of her low bone mass and the propensity it gives her to osteoporosis after menopause, when a repeat BMD would be appropriate. (Some clinicians may even check the BMD in 5–10 years.) She is counseled to maintain a normal body weight and to avoid exercising to the point of amenorrhea.

CASE 3: 40-Year-Old Woman

The third patient we will discuss is a 40-year-old executive and recreational jogger. She visits her healthcare provider after experiencing persistent hip pain that is interfering with professional and recreational activities. She has regular menstrual cycles and no risk factors for osteoporosis.

X-rays reveal a fracture of the femoral neck. DXA measurement indicates a T-score of -1.9 at the hip and spine. The Z-score is -2.7. (Note that at age 40, the patient’s expected Z-score would be higher than her T-score because she is past peak bone mass.)

Should this patient be treated for osteoporosis?

Not until other possible causes of low bone mass are ruled out. The patient’s personal and family histories do not point to increased risk for osteoporosis. There may very well be some other factor involved.

What is the clinician’s next step?

The patient’s clinical history makes her physician suspect a secondary disorder. Laboratory tests show low calcium, phosphorus and 25-vitamin D and an elevated total serum alkaline phosphatase. the serum bone-specific alkaline phosphatase was also high and indicated that bone was the source of the changes in total serum alkaline phosphatase

What is the cause of the patient’s low bone mass?

Consultation with a specialist reveals the patient has secondary hyperparathyroidism and very low urinary calcium. This was suspicious for a malabsorption problem causing osteomalacia, not primary osteoporosis. The completed workup shows occult celiac disease (see Clinical Updates on secondary causes of osteoporosis).

What is the recommended treatment for this patient?

The patient is treated with calcium and vitamin D supplementation and put on a gluten-free diet to control her celiac disease. At one year, her bone density is tested by DXA and is increased by 12%, indicating a reduced fracture risk.

Appropriate and Inappropriate Uses of Bone Density Testing

The development of DXA catapulted diagnosis and treatment of osteoporosis to the forefront of clinical medicine and public awareness. Before widespread use of this technology, diagnosis of osteoporosis and assessment of fracture risk was usually made only after fracture of the spine or hip, or after a patient had routine spinal X-rays taken for other reasons. In both circumstances, the disease had been silently progressing for many years. Use of DXA made it possible to diagnose osteoporosis before fractures occurred, in time to prevent fracture, deformity, and the manifold negative outcomes of osteoporosis. Like all diagnostic tests, DXA can have true positives (patient has a positive test and the disease), false positives (patient has a positive test but does not have the disease, (see patient 2), true negatives (patient has a negative test and does not have the disease) and false negatives (patient has a negative test but has the disease, see patient 1). Along with its benefits, however, use of DXA brought a few unexpected problems when applied outside the postmenopausal population for which it originally developed.

DXA uses the attenuation of x-rays through bone to measure bone mineral content at a skeletal site. This type of measurement is areal, providing a two-dimensional representation of bone. DXA does not yield a volumetric, or three-dimensional, picture of bone. However, bone strength is directly related to its three-dimensional properties: its microarchitecture, the number and strength of trabeculae (cross-beams inside cancellous bone) constituting the inner structure of bone. The thinner or fewer the trabeculae, the weaker the bone and the higher the risk for fracture. Disorders that affect bone mineral, such as osteomalacia may have a more profound effect on DXA than osteoporosis (see patient 3). Because DXA does not measure the volumetric characteristics of bone but measures the presence of bone mineral in a given area, it may not provide an accurate picture of bone health in every individual.^[1]

Defining Osteoporosis on the Basis of BMD T-Score

To aid in the diagnosis and treatment of this disorder, the WHO defined osteoporosis in terms of BMD T-score values obtained from central DXA. These definitions are based on the relationship between BMD and fracture risk in postmenopausal white women over age 65. Because the relationship between BMD and fractures has not been established in other patient groups, such as young women, other ethnic groups, men, or patients with other forms of bone loss such as steroid-induced osteoporosis, the WHO definitions do not necessarily apply to all patients. Application of the T-score cutoffs may lead to misdiagnosis. In addition, the WHO stated these criteria should not be applied in the presence of vitamin D deficiency. A T-score diagnosis simplifies interpretation of results but as illustrated is imperfect.

BMD is expressed as a relationship to two norms: the expected BMD for individuals of the same age and gender (Z-score), and for young normal adults of the same gender (T-score). The difference between the patient’s score and the norm is expressed in standard deviations (SD) above or below the mean. (Usually, 1 SD equals 10 to 20% of the bone density value.) T-scores decline in parallel to the steady drop in bone mass that occurs with aging (see figure).

In the above example, the BMD of a 57¬year-old woman is expressed in comparison to young-normal average BMD (T-score) and age-matched average BMD (Z-score). Thus, this same woman has a Z-score of -1 and a T-score of -2.5.

The original intent of densitometry was to identify postmenopausal female patients at risk for osteoporotic fracture before they fractured. In the relevant population of postmenopausal women, it is reasonable to assume that BMD more than two standard deviations below the young normal mean points to microarchitectural bone destruction resulting from osteoporosis and a higher risk for low-impact fractures in the majority of persons. Studies show similar findings when applied to men over 50 years of age.

In practice, application of standard T-score rankings may not be predictive of fracture when used for premenopausal women, younger men, or patients with other conditions that cause demineralization. Consequently, misinterpretation can and does occur in such patients when diagnosis, investigation and management are made on the basis of T-score alone.

Using the Z-Score

The second value derived from BMD measurement is the Z-score, or age-matched comparison. This measurement is also potentially a very useful one. The Z-score compares a patient’s BMD to his or her same-sex, age-matched group; therefore, a BMD equal to the mean BMD in the patient’s peer group carries a Z-score of 0.

Z-scores that are positive signify that an individual’s bone mass is better than average in the same-aged group. More important, negative Z-scores indicate that an individual’s bone mass is lower than that which is expected based on the referenced same age population. In children this may reflect an underlying bone disease, but often individual variations such as precocious or delayed growth spurts which may not be pathologic. In adults this suggests something may be causing bone to deteriorate faster than attributable to age and primary osteoporosis alone. However in young adults the value should be essentially the same as the T-score. A variety of secondary conditions can cause bone loss. Although most experts believe a low Z-score may indicate a disease process that would go unrecognized unless more information is sought, in reality secondary causes of low bone loss are more common in persons with low BMD relative to their peers, whether this value is in grams/centimeter², T-score or Z-score. Many of the differences we see today (see patient 2) are as much to do with lack of standardization of these values, as much as a specific disease process. Well-designed studies show that secondary causes of low bone mass or osteoporosis are found in approximately 40% of persons with normal BMD but such disorders are more prevalent in persons with lower BMD.

For patients in whom a specific secondary cause of osteoporosis is being considered, relevant blood and urine studies (such as urine calcium, thyrotropin, protein electrophoresis, cortisol or antibodies associated with gluten-sensitive enteropathy) should be obtained. This should also be done in all patients prior to initiating therapy and in anyone where there is concern for abnormal bone loss or osteoporosis. Such disorders can have specific therapies, may not be improved by drugs currently prescribed for treatment of osteoporosis, or indeed may mean such therapies are contraindicated.

Occasionally, we will see bone density results from young individuals, both men and women, which are markedly abnormal. The question arises as to whether these tests actually indicate osteoporosis, or smaller but architecturally normal bone, or some other disorder such as demineralization. A low BMD measurement in a young person does not have the same significance as a similar measurement in a postmenopausal woman or elderly man. Young people are less likely to fracture than older people at any given T-score value or absolute bone density. This is true even at T-scores equal to those found in older patients with documented osteoporosis.

Age is the deciding factor in interpreting a BMD measurement. The younger the patient, the lower the risk for fracture, even with T-scores that might be considered abnormal.^[2] This discrepancy may result from differences in the quality of young bone compared with older bone. The architectural qualities of young bone may make it both stronger and less susceptible to fracture than older bone, despite T-score values that seem to indicate otherwise. Other factors such as balance, muscle strength, and propensity to fall may contribute to differences in fracture risk between young and old bones. To date no biopsy studies looking at architecture of normal bone confirms clinical observations in this regard.

What is Normal BMD?

What is to be made of abnormal BMD in patients who are not postmenopausal women? Adherence to principals of good clinical practice (that is, obtaining a clinical history, family history, basic lab tests, physical examination) will often assist in identifying confounding variables that make the T-score meaningful. Once causes of secondary osteoporosis have been ruled out, what next? Should the patient be diagnosed with osteoporosis and treated? Does low bone density in a young person in itself constitute osteoporosis?

Low but normal bone mass is frequently identified in younger premenopausal women through densitometry measurements taken for nonclinically significant reasons. Such a low BMD reading probably does not signify destroyed bone tissue, loss of bone mass, and does not indicate osteoporosis. What it indicates is inherently lower-than-average BMD that is probably genetically determined in this particular individual.

Studies have looked for a “bone density gene” to identify healthy people who have less bone than the norm. Twin studies indicate that bone density is an inheritable factor, being more statistically robust in monozygotic twins (80-90%) than in dizygotic twins (70-80%). Studies show premenopausal healthy women of mothers with osteoporosis have lower bone density than those of mothers without osteoporosis. Racial differences and anthropomorphic and lifestyle factors have a great impact on bone development as well.^[3] Assessment of genetic effects on peak bone mass is critical to the clinician’s determination of whether a low T-score indicates a pathological disease process or not.^[4]clinically, this is often inferred from the patient’s and family’s phenotype. All family members have a small stature and skeleton.

On a purely statistical basis, some healthy individuals can be expected to “fall outside” of the normal distribution and appear abnormal, since by definition approximately 95% of people will be within 2 SD of the mean value. In such individuals, daily activity levels and low rates of fracture argue against belief that their bones are truly osteoporotic.

Causes of Lower-Than-Average Peak Bone Mass

In addition to genetics, why might a healthy person have low bone mass? Studies show that diet and hormonal status around puberty determine peak adult bone mass. Inadequate bone mass acquisition in adolescence leads to low adult skeletal mass. Consequently, the lower the peak adult bone mass, the less reserve there is to offset age-related bone loss later on.

Recent studies show that low bone mass often occurs in premenopausal and perimenopausal women.^[5] In about 66% of the observed cases, an identifiable cause of this bone deficiency was found—most often related to an aberration in estrogen metabolism and/or menstrual function. These women were healthy and did not sustain the kind of fractures normally associated with osteoporosis. However, the other 34% of patients with low bone mass had no identifiable reason for their condition. Such patients present a dilemma as to what should be done for them, if anything. In some cases, there is likely a genetic component to their low BMD, as evidenced by the fact that over 70% of the menopausal women in this particular study had a family history of osteoporosis. Most women in this age group tend to have stable bone mass that is not deteriorating over time and is clearly not osteoporotic in the usual sense of the term.

Does Low T-Score = Osteoporosis?

A patient’s low T-score would seem to suggest that the patient has significant bone loss. In young patients, this suggestion should be viewed with skepticism. It assumes that the patient began with normal bone density that was lost over time, and this may not be the case. Without baseline bone density values against which to compare them, current BMD test results cannot be used as the basis for diagnosing ongoing bone loss.

Clinicians may assume that low bone mass in such patients represents loss due to the aging process. This also is generally not the case, since the amount of bone loss that occurs on a yearly basis due to aging is very small and can’t account for the marked changes associated with T-scores of -2.5 or less. Most likely, such patients started out with lower-than average bone mass, and this is the cause of current low T-scores.

Efforts to develop a critical prediction rule to identify premenopausal women with low bone mass have been advocated. One recent proposal has been to correlate low peak bone mass to body weight, age at menarche, and physical activity.

Studies have attempted to identify factors associated with low Z-scores.^[6,7] Data suggest that low body mass index together with early age of menarche is predictive of lower bone mass in the mid 40s.^[8]

In a studied cohort of 20- and 30-year-old women, low body weight, late menarche, and physical inactivity predicted about 92% of the cases of low peak bone mass. This finding did not point to osteoporosis, but did indicate lower-than-optimal adult BMD and, therefore, the need for healthy lifestyle measures to preserve existing bone.

Bone densitometry measured by DXA is still the best indicator of actual BMD. It appears to be a stronger predictor than body weight, height, body mass index, or even calcium intake.^[9]

Studies have attempted to determine the predictive value of fractures in young patients.^[10] Data indicate that fractures between age 20 and 50 confer an increased risk of osteoporotic fracture after age 50, suggesting that individuals who fracture at a young age are reasonable candidates for BMD measurement and more-frequent follow-up than other individuals.

Summary

In the end, what is the utility of BMD measurement in women who fall outside of the narrow cohort of white, postmenopausal women over 65? Until databases of BMD measurements for different populations are developed for the purpose of comparison, BMD measurement for younger women and men is an imperfect diagnostic tool, to be used with reservation and clinical skepticism. DXA should not be performed on normal premenopausal women, those who are normally menstruating, not on corticosteroids, and not suffering fragility fractures.

When looking at BMD T-scores of premenopausal women, clinicians need to keep in mind that they are asking the technology to do something it wasn’t designed to do. They need to recognize that an abnormal T-score does not necessarily point to osteoporosis and increased fracture risk in all circumstances. Healthy bone can yield a low T-score, especially in young individuals, both men and women, who have achieved a lower-than-average peak bone mass. What’s needed to assess a patient’s “normal” bone mass is a thorough evaluation of the patient’s clinical history, family history, and lifestyle practices.

We do not mean to imply that premenopausal women and men cannot get osteoporosis. In most cases when this does occur, osteoporosis turns out to be of a secondary nature, the cause of which must be identified and treated. In such cases, a suitable work-up often leads to an identifiable problem. Although the ISCD recommends using Z-scores for diagnosis of low bone mass in premenopausal women and men < 50 years, and many authors advocate a variety of Z-score cut-offs for deciding who to evaluate for secondary causes of osteoporosis, the reality is that Z-scores have significant variability due to calculation differences. This makes interpretation difficult and raises questions about their true validity. The best available data suggests that secondary causes of low bone mass and osteoporosis are very common in all populations but increasingly prevalent in persons with lower BMD.

1. Schussheim DH, Rubin MR, Shane E. Discrepant areal and volumetric bone density measurements in a young woman with ideopathic low bone mineral density. Endocr Pract. 2003; 9:36–9.
2. Yang RS, Tsai KS, Liu TK, Chieng PU, Dorey FJ. Risk of underestimation of the bone mineral density of proximal femur. Bone. 1997;20(3):295–300.
3. Finkelstein JS, Lee ML, Sowers M, Ettinger B, Neer RM, Kelsey JL, Cauley JA, Huang MH, Greendale GA. Ethnic variation in bone density in premenopausal and early perimenopausal women: effects of anthropometric and lifestyle factors. J Clin Endocrinol Metab. 2002;87(7):3057–67.
4. Bainbridge KE, Sowers MF, Crutchfield M, Lin X, Jannausch M, Harlow SD. Natural history of bone loss over 6 years among premenopausal and early postmenopausal women. Am J Epidemiol. 2002;156(5):410–7.
5. Moreira Kulak CA, Schussheim DH, McMahon DJ, et al. Osteoporosis and low bone mass in premenopausal and perimenopausal women. Endocrine Practice. 2000;6(4):296–304.
6. Rollins D, Imrhan V, Czajka-Narins DM, Nichols DL. Lower bone mass detected at femoral neck and lumbar spine in lower-weight vs normal-weight small-boned women. J Am Diet Assoc. 2003;103(6):742–4.
7. Hawker GA, Jamal SA, Ridout R, Chase C. A clinical prediction rule to identify premenopausal women with low bone mass. Osteoporos Int. 2002; 13(5):400–6.
8. Blum M, Harris SS, Must A, Phillips SM, Rand WM, Dawson-Hughes B. Weight and body mass index at menarche are associated with pre-menopausal bone mass. Osteoporos Int. 2001;12(7):588–94.
9. Picard D, Imbach A, Couturier M, Lepage R, Picard M. Familial resemblance of bone mineral density between females 18 years and older and their mothers. Can J Public Health. 2001; 92(5):353–8.
10. Wu F, Mason B, Horne A, Ames R, Clearwater J, Liu M, Evans MC, Gamble GD, Reid IR. Fractures between the ages of 20 and 50 years increase women’s risk of subsequent fractures. Arch Intern Med. 2002;162 (1):33–6.
11. Carey JJ, Delaney MF, Love TE, Richmond BJ, Cromer BA, Miller PD, Manilla-McIntosh ML, Lewis SA, Thomas CL, Licata AA. 2007 DXA-generated Z-scores and T-scores may differ substantially and significantly in young adults. J Clin Densitom. 10:351-358.
12. Carey JJ, Delaney MF, Love TE, Cromer BA, Miller PD, Richmond BJ, Manilla-McIntosh ML, Lewis SA, Thomas CL, Licata AA. 2009 Dual-energy X-ray absorptiometry diagnostic discordance between Z-scores and T-scores in young adults. J Clin Densitom.12:11-16.
13. Camacho PM, Agrawal M, Norton I, Diaz I. 2006 Using DXA Z-Scores to predict the presence of secondary causes of osteoporosis. Proc 28th Meeting American Society for Bone and Mineral Research, Philadelphia, PA. M092.

Continuing Education

CE Credit

After participating in this activity, the reader has the option of taking a post-test to qualify for continuing education credit for this activity. It is estimated it will take 1.0 hour(s) to complete the reading and take the post-test. Continuing education credit will be available for two years from the date of publication.

The National Osteoporosis Foundation is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians.  The National Osteoporosis Foundation designates this educational activity for a maximum of 1.0 AMA PRA Category 1 Credit(s)TM. Physicians should only claim credit commensurate with the extent of their participation in the activity.

The National Osteoporosis Foundation is accredited as a provider of continuing nursing education by the American Nurses Credentialing Center’s Commission on Accreditation.

The National Osteoporosis Foundation designates this educational activity for a maximum of 1.0 continuing nursing education credit(s).

Other healthcare providers will also be able to receive a certificate of completion; nurse practitioners and physician assistants may request an AMA PRA Category 1 Credit(s)™ certificate.

If you wish to receive continuing education credit, please click here to access the Continuing Education Application. Print this form and complete offline. Complete the registration form then take the 10-question post test and circle your answers on the CE Reporting Form. Then complete the activity evaluation found at the bottom of the CE Reporting Form. You must complete the Post-Test and CE Reporting Form in order to receive credit.

There is a $10.00 fee for this continuing education activity. Payment may be made by check, VISA or MasterCard. We cannot accept any other forms of payment. Please complete the payment information on the CE Reporting Form.

Once the Continuing Education Application is complete, return it with your payment to NOF at the address indicated on the form. Your continuing education certificate will be emailed to you within 3 weeks of receipt.

Software Requirement: Adobe Acrobat Reader is required to read the text for this continuing education activity.