More on DXA

A study published in the current issue (February 2009) of the Journal of Bone and Mineral Research points out that change in bone mineral density is an important predictor of future fractures, particularly in patients not on specific therapy for osteoporosis. A change in BMD appears to be a stronger predictor of future fractures than the baseline BMD.

Let’s examine why this should be the case.

BMD is calculated as the ratio of the bone mineral content/the area of the skeleton measured. When performed correctly the area of the skeleton examined should not be changing over time, and any change in BMD should reflect only a change in BMC. In the lumbar spine, for example, the area is only likely to change if there is a fracture (decreased area without change in BMC will result in a seeming increase in BMD) or there are new osteophytes (increased area without change in BMC will result in a seeming decrease in BMD).

Given that for most patients DXA is not repeated inside 12 months, it is often difficult to reposition the patient in exactly the same position time after time. This is particularly an issue in facilities with more than one trained DXA technologist because each one, no matter how skilled and experienced, will have slight differences in the way they position the patient. It is a bigger problem with serial scans of the proximal femur because of the specific requirements for internal rotation. It is for this reason that the recommendations for monitoring proximal femur BMD emphasize that the skeletal site of interest should be the region identified as “total proximal femur” and not the femoral neck. These positioning differences also contribute to the strong recommendation that, to the extent possible, serial DXA studies be performed at the same performance site as the original DXA study.

How can you, the end user, be certain that positioning errors are kept to a minimum? There is no shortcut, and you should request that all DXA reports reach you with not just a summary statement cover note but with a complete print-out of the DXA results. These print-outs contain very specific information about the area scanned and the BMC such that you can quickly confirm that the serial DXA studies are comparing apples to apples. Within a very short period of time, you will gain confidence that a particular DXA center is indeed paying close attention to repositioning and you can more comfortably rely on the summary data alone. It will take you even less time to determine which centers are not paying enough attention to this.

J Bone Miner Res. 2009;24:361-370. PMID: 18847328.