When and how should bone mineral density be measured?


  • Marjo Tuppurainen

Osteoporosis (OP) is defined as a skeletal disease in which bone mass is reduced, accompanied by microarchitectural changes in the skeleton leading to fractures (1). OP is related to aging and spontaneous, surgical or pharmacological menopause, but it can also be a secondary phenomenon caused by other diseases or medications, for example malabsorption syndromes, chronic renal failure, hyperthyroidism, hyperparathyroidism, Cushing's disease and prednisolone use over 7.5 mg per day (2). In addition, low body mass index (< 19 kg/m2), low calcium intake, smoking, previous fracture (3) and history of a mother's hip fracture are traditional risk factors for osteoporosis and fractures.

In puberty, sex steroids, growth hormone and insulin-like growth factors, 1,25-dihydroxyvitamin D, muscle mass and calcium intake are important determinants for bone development (4). In addition, hereditability accounts for three-fourths of the variance in bone mass (5). Peak bone mass in the axial skeleton is achieved by the age of 20–25 years. At menopause, bone loss increases and by the age of 80 women have lost nearly half of their bone minerals whereas men only about 20%. An accelerated phase of bone loss occurs during the first 5-year period after menopause, being about 1–2% per year (2,6), and declining thereafter.

Traditional risk factors have poor specificity and sensitivity in prediction of either fractures or bone mineral density (BMD) – approximately 30% of osteoporotic patients have known risk factors. The first sign of OP is often a wrist fracture after a minor trauma at menopause. At older ages, sudden back pain and kyphosis, especially in the thoracic spine and loss of height refers to OP.

Measurement of bone mass by densitometry has become central to the diagnosis of osteoporosis as well as decisions about treatment to prevent fractures. Low bone mineral density (BMD) predicts the risk of fractures similarly as high cholesterol or high blood pressure predicts the risk of a myocardial infarction and stroke (7). In routine clinical use, BMD helps to detect patients with OP. Women suffering from long-term secondary amenorrhoea (> 1 years) or having the menopause under the age of 45 should be sent for BMD measurements. In addition, the presence of secondary diseases and medications predisposing to OP, as well as long-term immobilization, loss of height and radiographic evidence of osteopenia or vertebral deformity are indications for BMD measurement (8).

Several technologies can be used to assess BMD. Single-photon absorptiometry (SPA) and single-energy X-ray absorptiometry (SXA) are techniques for non-invasive measurement of BMD at peripheral sites. Quantitative computed tomography (QCT) of the spine is available on standard computed tomography devices and a smaller device, peripheral QCT, performs QCT of the distal forearm. Quantitative ultrasound (QUS) techniques are also used for peripheral bone quality assessments, most commonly calcaneus. QUS of the heel resembles other peripheral measurements in terms of ability to predict fractures of elderly women (9). However, the diagnosis of OP can not be based on QUS techniques and the effects of bone-sparing treatments cannot be assessed. Dual energy X-ray absorptiometry (DXA) of the spine and hip is currently the ‘golden standard’ for BMD measurement (10).

The diagnostic criteria of OP are based on T-score values in BMD according to WHO (7), which gives the clinical relevance from the occurrence of fractures in postmenopausal white women. A T-score of 0 means that the BMD is equivalent to the mean peak BMD of a population of healthy premenopausal white women aged 20–29 years. T-scores between 1 and 2.5 SD below the average of the reference population are considered as osteopenic, and values below −2.5 SD are defined to be osteoporotic. However, a T-score-based OP definition has met increasing criticism because it can not be used for all densitometric techniques. The advantages of DXA are stable calibration, short scanning time (1–5 min), low precision error (1–2%) and low radiation dose (1/10 of thoracic X-ray). However, central DXA is large, expensive, and is not available everywhere and this assesses great needs for peripheral measurement techniques.

BMD of the distal radius, phalanx and calcaneus, correlates reasonably well (r = 0.6) with BMD of the lumbar spine and hip (11). Thus, a low BMD at any site is likely to represent a global reduction in BMD and peripheral BMD measurement can be used to assess fracture risk at both peripheral and central sites (12). The discordance between BMDs measured at different sites is greater in the early postmenopausal population than in women over the age of 65 (13). It has been suggested that due to general spinal deformities, such as osteophytes, facet sclerosis and aortic calcifications, the most reliable measurement site would be the femoral neck (14).

Diagnostic and intervention criteria are not the same, while a BMD below −1 SD Z-score (BMD of the women of the own age group) could be a practical limit to start treatment (15). Every 1 SD (about 10%) decrease in BMD increases the fracture risk about two- to threefold (16). However, a T-score has a different prognostic significance at different ages. For example, in Swedish women the 10-year probability of sustaining a hip, spine, forearm or shoulder fracture with a T-score of −2.5 SD varies from 8.1% at the age of 50 years to 25.6% at the age of 80 years (17).

Measurements of BMD can be used for the monitoring of changes in BMD which occur with advancing age or during different therapies. In the case of osteopenia, the assessment should be repeated in 3–5 years without therapy. However, the closer the current value is to a threshold where treatment would be started, the sooner the BMD measurement should be repeated (18). Generally, during the treatment there is no need to repeat BMD more often than every year to 2 years. It would be ideal if a patient had serial measurements performed on the same machine and perhaps by the same technician.

BMD measurement using DXA predicts the severity of bone loss and predicts the fracture risk. However, half of the fractures occur in persons with normal a BMD. Thus, the risk of fracture should not be based totally on BMD values, but other risk factors, such as an increased likelihood of falling, fast bone turnover measured by bone markers and advanced age (19) should be taken into account when considering the risk of a fracture.

The recent recommendation of the US Preventive Services Task Force (USPSTF) is that women over 65 years of age and older, should be screened routinely for osteoporosis and in addition, women over the age of 60 at increased risk of osteoporotic fractures should be screened (20). Before this recommendation gets wider acceptance, the diagnosis of OP should be made at least at the time of the first fracture. Thus, each peri- and postmenopausal women with fractures should be treated potentially osteoporotic and a BMD measurement should be considered. Early detection of women at increased fracture risk is a great challenge to every physician because effective therapies to increase bone strength and to prevent fractures are available.

Address for correspondence:
Marjo Tuppurainen
Bone and Cartilage Research Unit
Kuopio University
and Kuopio University Hospital
Department of Obstetrics and Gynaecology
P.O. Box 1627
Fin-70211 Kuopio
e-mail: Marjo.Tuppurainen@kuh.fi