Results from this manuscript have been presented at the sixteenth annual meeting of the American Society for Bone and Mineral Research, Kansas City, Missouri, September 9–13, 1994.
Estimates of volumetric bone density from projectional measurements improve the discriminatory capability of dual X-ray absorptiometry†
Article first published online: 3 DEC 2009
Copyright © 1995 ASBMR
Journal of Bone and Mineral Research
Volume 10, Issue 7, pages 1101–1110, July 1995
How to Cite
Jergas, M., Breitenseher, M., Glüer, C.-C., Yu, W. and Genant, H. K. (1995), Estimates of volumetric bone density from projectional measurements improve the discriminatory capability of dual X-ray absorptiometry. J Bone Miner Res, 10: 1101–1110. doi: 10.1002/jbmr.5650100715
- Issue published online: 3 DEC 2009
- Article first published online: 3 DEC 2009
- Manuscript Accepted: 8 MAR 1995
- Manuscript Revised: 15 FEB 1995
- Manuscript Received: 17 OCT 1994
To determine whether estimates of volumetric bone density from projectional scans of the lumbar spine have weaker associations with height and weight and stronger associations with prevalent vertebral fractures than standard projectional bone mineral density (BMD) and bone mineral content (BMC), we obtained posteroanterior (PA) dual X-ray absorptiometry (DXA), lateral supine DXA (Hologic QDR 2000), and quantitative computed tomography (QCT, GE 9800 scanner) in 260 postmenopausal women enrolled in two trials of treatment for osteoporosis. In 223 women, all vertebral levels, i.e., L2–L4 in the DXA scan and L1–L3 in the QCT scan, could be evaluated. Fifty-five women were diagnosed as having at least one mild fracture (age 67.9 ± 6.5 years) and 168 women did not have any fractures (age 62.3 ± 6.9 years). We derived three estimates of “volumetric bone density” from PA DXA (BMAD, BMAD*, and BMD*) and three from paired PA and lateral DXA (WA BMD, WA BMDHol, and eVBMD). While PA BMC and PA BMD were significantly correlated with height (r = 0.49 and r = 0.28) or weight (r = 0.38 and r = 037), QCT and the volumetric bone density estimates from paired PA and lateral scans were not (r = −0.083 to r = 0.050). BMAD, BMAD*, and BMD* correlated with weight but not height. The associations with vertebral fracture were stronger for QCT (odds ratio [OR] = 3.17; 95% confidence interval [CI] = 1.90–5.27), eVBMD (OR = 2.87; CI 1.80–4.57), WA BMDHol (OR = 2.86; CI 1.80–4.55) and WA-BMD (OR = 2.77; CI 1.75–4.39) than for BMAD*/BMD* (OR = 2.03; CI 132–3.12), BMAD (OR = 1.68; CI 1.14–2.48), lateral BMD (OR = 1.88; CI 1.28–2.77), standard PA BMD (OR = 1.47; CI 1.02–2.13) or PA BMC (OR = 1.22; CI 0.86–1.74). The areas under the receiver operating characteristic (ROC) curves for QCT and all estimates of volumetric BMD were significantly higher compared with standard PA BMD and PA BMC. We conclude that, like QCT, estimates of volumetric bone density from paired PA and lateral scans are unaffected by height and weight and are more strongly associated with vertebral fracture than standard PA BMD or BMC, or estimates of volumetric density that are solely based on PA DXA scans.