A population-based comparison of quantitative dual-energy X-ray absorptiometry with dual-photon absorptiometry of the spine and hip

Abstract

Dual photon absorptiometry (DPA) is currently the most widely used method for noninvasive bone mineral density (BMD) measurement of the axial skeleton. Dual energy X-ray absorptiometry (DEXA) is a recently developed technique that uses an X-ray tube as a photon source; it has demonstrated several significant advantages over DPA in preliminary studies. We report here a quantitative comparison of the DEXA and DPA technologies using a Hologic DEXA (Hologic QDR model 1000, Waltham, MA) scanner and a Lunar DPA (Lunar Radiation DP3, gandolineum-153 source) scanner at both the proximal femur and lumbar spine sites using bone density measurements from a population-based sample of older white men and women who had complete DEXA and DPA measurements of the hip (n = 217) or the spine (n = 176). To examine the relationship of BMD measured by the DPA scanner to BMD measured on the DEXA scanner, normal least squares linear regression was used to regress the DPA BMD on the DEXA BMD for each site. DEXA values were consistently lower than DPA values, with an average difference of 16%. The squared multiple correlation (R2) values were at or above 0.95 for almost all sites, with Ward's triangle having the lowest value (0.89). The slope for all sites was similar, ranging from 0.94 to 1.1. Research and clinical centers that wish to change to DEXA technology because of its shorter examination time and greater precision can therefore compare DEXA with DPA values using representative conversion factors.