Bone ultrasound

  • Cost much less than DEXA
  • Reproducibility not quite as good as DEXA
  • Relative risk for hip fracture predicted as well by ultrasound of the calcaneus as by DEXA of the hip
  • Fracture prediction is independent of DEXA, but adding ultrasound to DEXA does not enhance prediction
  • in vitro studies show that ultrasound is mainly measuring the bone mass; DEXA better able to predict bone strength
  • debate about whether treatment can be monitored by ultrasound

    New! Recently Bauer and colleagues presented an abstract at the 2004 ASBMR comparing ultrasound of the heel and DEXA of the hip in 5995 men in the MrOs study. The tecniques showed relative risk of 1.7 for ultrasound for clinical fractures, compared to 1.9 for DEXA. For hip fractures the risks were 2.1 vs 3.3, not significantly different. Areas under the ROC curves were both .84. Combining the tests did not improve fracture prediction.

    There have been many articles since the tables below, but most of them have similar results. Updating is planned.

    Precision of calcaneal measurements
    Model Instrument Interobserver
    Achilles 2.74 %2.48 %
    UBA575 2.01 %4.97 %
    DEXA 1.23 %1.91 %
    Greenspan, S. L.(1997). Precision and discriminatory ability of calcaneal bone assessment technologies. J Bone Miner Res 12(8): 1303-13.

    Relative risk of fracture as predicted by ultrasound
    Author yr Design N Relative risk Independent
    Bauer, D. C. 97 pros x 2yrs 6189 2.0 yes
    Hans, D. 96 pros x 2yrs 5662 2.0 yes
    Pluijm 97 pros x 2yrs 710 ** .
    Heaney, R. P. 95 pros x 2yrs 130 2.1 .
    Thompson, P. 98 xs 3180 1.4 .
    Stegman, M. R. 95 xs 1428 1.5 .
    Bauer, D. C. 95 xs 442 1.8 yes
    Cepollaro, C. 97 xs 397 3.5* yes
    Ross, P. 95 retro 544 1.8 yes
    ** Risk 4.3 lowest to highest tercile. * risk for DEXA was 7.5

    in vitro studies of bone mechanics and ultrasound
    AuthorYrNFactorvs USvs BMDUS vs BMD Additive?
    Bouxsein, M. L.9731Compression strength.48-.63.66-.88..
    Langton, C. M.9620Compressive strength.74-.76.880.88no
    Lochmuller, E. M.9849Failure load.23.38.some
    Lochmuller, E. M.9858Failure load.67.65..
    Nicholson, P. H.9764Femoral strength.40-.47.71-.880.37no
    Cheng, X. G.9762Ultimate load.10.640.76no
    Hodgskinson, R.9718Young's modulus.95.87.some
    Nicholson, P. H.9923Young's modulus.26-.83..yes
    This table shows correlations (r-squared) between mechanical factors and ultrasound and bone density, as well as the r-squared between ultrasound and bone density. In most studies the mechanical factors are more closely related to the density than to ultrasound measurements.

    Structural measurements
    Hans, D.9517 Related to trabecular thickness but not other parameters
    Hausler, K. D.9915 Not correlated to histomorphometry
    Nicholson, P. H.9870 Depends on direction, mostly measures density
    Trebacz, H.9941 Strongly correlated with BV/TV; not other parameters

    Updated 10/15/04