(TBBM, fat, and nonbone lean tissue); and (3) the evaluation of regional body composition.
DXA instruments have been commercially available since 1987. Advantages of DXA compared to its predecessor, dual-photon absorptiometry (DPA), include reduced radiation exposure, stability of the energy source, faster data collection, and increased precision of measurements. The three major manufacturers of DXA instruments in the United States are Hologic (Waltham, Mass.), Lunar (Madison, Wis.), and Norland (Fort Atkinson, Wis.). Although the first generation of DXA instruments from these manufacturers was limited to the measurement of BMC and BMD of selected regions of the skeleton, all of the manufacturers now produce machines that can assess both total body and regional bone mineral and soft tissue masses. The principles of operation, methods of calibration, and other features of the various models of DXA instruments produced by each manufacturer have been reviewed by Lohman (1996).
The short-and long-term reliability and precision of BMD measured by DXA have been shown to be excellent. Table 6-1 summarizes the findings of just a few of the studies that have assessed the short-term coefficient of variation (CV) in measuring BMD by DXA. It is apparent that there is good precision across manufacturers, instrument models, data collection modes, and measurement sites. Long-term precision of DXA, which is typically evaluated by measuring BMD of a spine phantom (a mechanical model for predicting irradiation dosage deep in the body), has been shown to be excellent, with CVs less than 1.0 percent (Lilley et al., 1991; Orwoll et al., 1993). The highly reproducible measurement of total body bone mineral content (Jensen et al., 1994; Mazess et al., 1990; Pritchard et al., 1993; Slosman et al., 1992; Snead et al., 1993; Svendsen et al., 1993) makes DXA particularly suitable for use in multicompartment models of body composition.
The validity of an instrument depends on how well it measures what it is intended to measure. In the case of the measurement of BMD by DXA, validity remains questionable because there is a wide disparity among BMD measures