Digital X-ray radiogrammetry better identifies osteoarthritis patients with a low bone mineral density than quantitative ultrasound.

Research paper by Gerhard W GW Goerres, Diana D Frey, Thomas F TF Hany, Burkhardt B Seifert, Hans Jörg HJ Häuselmann, Annina A Studer, Dagmar D Hauser, Nathalie N Zilic, Beat A BA Michel, Didier D Hans, Daniel D Uebelhart

Indexed on: 06 Sep '06Published on: 06 Sep '06Published in: European Radiology


This study assessed the ability of quantitative ultrasound (QUS) and digital X-ray radiogrammetry (DXR) to identify osteopenia and osteoporosis in patients with knee osteoarthritis (OA). One hundred and sixty-one patients with painful knee OA (81 men, 80 women; age 62.6+/-9.2 years, range 40-82 years) were included in this cross-sectional study and underwent dual-energy X-ray absorptiometry (DXA) of both hips and the lumbar spine, QUS of the phalanges and calcanei of both hands and heels, and DXR using radiographs of both hands. Unpaired t-test, Mann-Whitney U test, ROC analysis and Spearman's rank correlation were used for comparisons and correlation of methods. Using DXA as the reference standard, we defined a low bone mineral density (BMD) as a T-score < or =-1.0 at the lumbar spine or proximal femur. In contrast to phalangeal or calcaneal QUS, DXR was able to discriminate patients with a low BMD at the lumbar spine (p<0.0001) or hips (p<0.0001). ROC analysis showed that DXR had an acceptable predictive power in identifying OA patients a low hip BMD (sensitivity 70%, specificity 71%). Therefore, DXR used as a screening tool could help in identifying patients with knee OA for DXA.