Screening and Risk Assessment for Osteoporosis in Postmenopausal Women

Guideline Synthesis

Agency for Healthcare Research and Quality (AHRQ)


March 16, 2011

In This Article

Comparison of Recommendations: Measurement of BMD: Modality and Frequency

ACPM (2009)

Screening for osteoporosis should be performed with BMD testing by DXA if available, and not more frequently than every 2 years.

Osteoporosis screening modalities. At present, DXA is the most widely accepted and used method of screening for osteoporosis. It is a clinically proven method of measuring BMD, is non-invasive, takes only 10-15 minutes, and exposes patients to only a small amount of radiation (less than one tenth of the amount of a chest x-ray). A t-score, derived from the DXA measurement, expresses an individual's BMD (in SDs) compared to the mean BMD of a "young normal" adult population of the same gender.

One disadvantage of DXA is that the machines are not portable. They also do not provide any information about bone architecture, which can influence fracture risk independent of BMD. Factors interfering with accuracy of DXA include osteoarthritis, vertebral compression fractures, osteophytes, and vascular calcification.

Other potential screening tests include calcaneal QUS, QCT radiography, and the use of biochemical (urine and serum) markers. There is increased interest in osteoporosis screening using QUS because it is portable, does not expose patients to radiation, and is relatively inexpensive. Sound waves are passed through the calcaneus, and the speed of sound and absorption patterns of various sound wavelengths are measured, which is known as broadband ultrasound attenuation. Given the poor sensitivity of QUS for detecting osteoporosis, it has limited application in evidence-based screening programs for osteoporosis.

It is also possible to use QCT to measure BMD. A benefit of QCT is that it can analyze cortical and trabecular bone, so it is less influenced by the changes caused by degenerative disease, which can interfere with DXA accuracy. However, it is more expensive than DXA, and QCT exposes patients to a marked increase in radiation. The use of QCT as a screening tool for osteoporosis has not yet been extensively researched, and it has not yet been validated in relation to t-scores that predict fracture risk.

Other potential screening tests include serum and urine tests for markers of bone formation and resorption. Markers of bone formation include bone-specific alkaline phosphatase, osteocalcin, and procollagen I carboxy and N-terminal extension peptides. Markers of bone resorption include urinary levels of pyridinolines and deoxypyridinolines, and serum and urine levels of type I collagen telopeptides. The level of these markers may identify changes in bone remodeling within a relatively short time interval (several days to months) before changes in BMD can be detected. These biochemical markers of bone turnover are often used in the research setting but have limited clinical utility. They do not predict bone mass or reliably estimate fracture risk, but they may be helpful in monitoring response to antiresorption therapies in patients with osteoporosis. Therefore, they cannot replace BMD testing and are not useful for population-based screening.

NAMS (2010)


  • When BMD testing is indicated, DXA is the preferred technique. The total hip, femoral neck, and posterioranterior lumbar spine should be measured, using the lowest of the three BMD scores.

  • The routine use of biochemical markers of bone turnover in clinical practice is not generally recommended.

  • For untreated postmenopausal women, repeat DXA testing is not useful until 2 to 5 years have passed.

Bone-testing options. Fracture risk can be estimated by a variety of technologies at numerous skeletal sites. BMD measured by DXA is the only diagnostic technology by which measurements are made at hip, spine, and radius. These are also important sites of osteoporotic fracture.

When BMD testing is indicated, NAMS recommends measuring the total hip, femoral neck, and posterior-anterior lumbar spine, using the lowest of the three BMD scores for diagnosis. In some patients, degenerative or other artifacts at the spine site make measurements unreliable. In such cases, the one-third radius should be measured and used as a second site valid for diagnosis. The spine may be a useful site for BMD measurement in early postmenopausal women because decreases in BMD can be faster at the spine than at the hip.

Although bone tests at peripheral sites (e.g., tibia, finger, calcaneus) can identify women at risk of fracture, they are not useful for the diagnosis of osteoporosis and have limited or no value in the follow-up of patients. Peripheral site measurements may be useful to raise awareness about bone health and have been utilized as a prescreen for DXA testing where DXA availability is limited.

Follow-up BMD testing. In most cases, repeat DXA testing in untreated postmenopausal women is not useful until 2 to 5 years have passed, given the rate of bone loss of 1% to 1.5% per year. Postmenopausal women, after substantial BMD losses in early postmenopause, generally lose about 0.5 T-score units in BMD every 5 years.

For women receiving osteoporosis therapy, BMD monitoring may not provide clinically useful information until after 1 to 2 years of treatment. Stable BMD (within the precision error of the instrument) indicates successful therapy; fracture risk reductions for patients on antiresorptive therapy are similar with stable bone density or with increases in BMD. Marked declines in BMD predict greater fracture risk and should trigger a reevaluation for secondary causes of osteoporosis or treatment nonadherence.

Each DXA testing center should perform precision testing to determine the least significant change that can be detected in their patient population. Statistically insignificant decreases in BMD should be reported as stable bone density within the precision error of the instrument. Statistically significant changes in BMD (equal to or greater than the least significant change) should be reported as such.

Bone turnover markers. Biochemical markers of bone turnover can be measured in serum or urine. They can indicate either osteoclastic bone resorption (breakdown products of type I collagen in bone: N-telopeptides, C-telopeptides, deoxypyridinoline) or osteoblast functioning (bone matrix synthesis: bone-specific alkaline phosphatase, procollagen type I N-terminal propeptide, osteocalcin). Bone turnover markers cannot diagnose osteoporosis and have varying ability to predict fracture risk when studied in groups of patients in clinical trials. They also have varying value in predicting individual patient response to therapy. Nevertheless, these tests may show an individual patient's response to therapy earlier than BMD changes, sometimes within 2 to 3 months as opposed to the 1 to 3 years required with BMD. Most bone turnover markers vary greatly from day to day, are affected by food intake and time of day, and lack assay standardization, limiting their clinical utility. In some cases, persistently elevated bone turnover markers in the face of antiresorptive therapy may alert the clinician to nonadherence to therapy, poor absorption of medication, or other secondary causes of osteoporosis.

The value of bone turnover markers in encouraging adherence to therapy has been debated. Several trials have found no difference in adherence when marker values are communicated to women.


Diagnostic testing

DXA. DXA is the test of choice for measuring BMD. Although various skeletal sites can be assessed by DXA, BMD of the nondominant hip is the best predictor of hip fracture and is an excellent predictor of vertebral or wrist fracture. Loss of vertebral bone is accelerated early in menopause and early in glucocorticoid use. Spine BMD measurements may be helpful in these settings.

BMD measurement by DXA may be spuriously elevated by a number of factors. Vertebral compression fractures typically result in a "smaller" vertebral body with no change in the total amount of calcium, and thus produce an apparent increase in BMD. Vertebral osteophytes, degenerative joint disease, and aortic calcifications can also falsely raise BMD measurements. Hip measurements tend to have fewer artifacts.

Initial screening with DXA is addressed in the previous section of this Synthesis. Recommendations for initial screening are based on the clinical risk factors for osteoporosis and related fractures, outlined in Tables 2, 3, and 4 in the original guideline document. (Repeat DXA screening is addressed as part of the discussion of "Follow Up," after the discussion of treatment).

When possible, use the L1-L4 value to diagnose osteoporosis at the spine. If anatomic abnormalities are present, use any combination of two or three vertebrae (e.g., L2-L4). Guidelines suggest not making a diagnosis of osteoporosis based on the T-score of one vertebral body.

For diagnosing osteoporosis of the hip, the femoral neck or total hip are the preferred sites.

Other diagnostic and monitoring modalities. Other testing modalities are available, but have limitations in routine testing.

Quantitative calcaneal ultrasound devices are both portable and inexpensive, and are often used in informal osteoporosis screening programs, such as health fairs. However, meta-analysis suggests limited value for ultrasound screening. For example, a positive study in an otherwise healthy 65 year-old woman raises the likelihood of DXA confirmed osteoporosis from a population-based pre-test estimate of 22% only to 34%. Conversely, a negative study reduces the likelihood of osteoporosis from 22% to 10%. T-scores provided by ultrasound are not equivalent to DXA T-scores, and should therefore not be used for diagnostic purposes. Instead, patients with abnormally low ultrasound T-scores should be evaluated by DXA for more definitive diagnosis.

Biochemical markers of bone resorption are used in research and may be used clinically to assess the effectiveness of antiresorptive therapy. In the latter setting, a decrease in these markers to premenopausal levels usually occurs after two to three months of therapy. Some data suggest that elevated levels of bone resorption makers in older women are an independent risk factor for fractures. However, bone markers are not a reliable predictor of BMD, and are not a substitute for DXA in women at risk. Generally, their use in the diagnosis of osteoporosis is not recommended.


  • Repeat DXA based on a patient's situation (refer to Tables 5 and 8 in the original guideline document) [IC-D]. Consider not repeating DXA on patients with moderate bone loss who are fracture-free on bisphosphonate therapy [IIC].

  • For most persons, ≥2 years between DXAs provides the most meaningful information [B].

  • Early in glucocorticoid use and/or after transplantation consider repeating DXA in 6-12 months [IB].

Follow-up and when to repeat DXA. When deciding if and when to repeat a DXA scan, consider:

  • The patient's clinical risk factors for progression of bone loss and for fracture

  • The results from prior scans

  • Whether a repeat DXA will change management

  • Whether a repeat DXA result may improve compliance with therapy even if it will not change management

Tables 5 and 8 in the original guideline document summarize these and additional factors that may play a role in ongoing management and follow up.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.