Ultrasound transmission speed and ultrasound bone profile score (UBPS) of the phalanges in normal women and women with osteoporosis

The distal metaphysis of the first phalanx of the fingers II-V is, like the vertebral body, a useful site for the measurement of mineralisation and structure of the bone because of the simultaneous presence of compact and trabecular bone. With an ultrasound device (DBM sonic 1200, IGEA, Italy), we measured the adSOS (the amplitude dependent speed of sound) and the UBPS (ultrasound bone profile score), a score which is calculated from the graphic traces of the receiving probe with an expert system which uses fuzzy-logic at phalanges II-IV, as well as bone mineral density (BMD) at lumbar spine using dual X-ray absorptiometry (DXA). Precision of the measurements was as follows: adSOS: short-time-CV% = 0.576, long-time-CV% = 1.1, SCV% = 5.9, RMSSD% = 1.825. UBPS: short-time CV% = 5.95. There was no correlation between adSOS or UBPS and lumbar BMD (DXA). There was a significant positive correlation between adSOS and UBPS, r = 0.804 (p < 0.00001). The validity of adSOS and UBPS was examined in 25 young and healthy women (mean age: 33.4 year), 15 postmenopausal healthy women (mean age: 58.5 years), 17 women with osteopenia, (mean age: 52.4 years), as defined by a t-score between -1 to -2.5 SD as lumbar BMD (DXA), and 20 women with osteoporosis and vertebral fractures (mean age: 61.4 years). We compared the healthy postmenopausal women and the women with osteoporotic vertebral fractures, the z-score of the adSOS was below minus 1.5 SD and UBPS was below 40, sensitivity was 0.7 for adSOS, and 0.85 for UBPS, with a specificity 0.97 for adSOS, and of 0.93 for UBPS; positive predictive value: adSOS: 0.93, UBPS: 0.85. AdSOS declined with age (r = 0.694, p = 0.021); the UBPS was not age dependent (r = -0.15, p = n.s.). The ROC-curve shows a value of 0.96 for adSOS and 0.94 for UBPS. AdSOS and UBPS could discriminate well between the healthy controls and the women with osteopenia or vertebral fractures (p < 0.00001). These results show that adSOS and UBPS are precise parameters to be measured at the phalanges. The detection level of pathological changes in osteoporosis are similar between adSOS and lumbar BMD (DXA) and improved by using the UBPS. This might be explained by the influence of structural changes in bone on UBPS, rather than change in bone mineral alone. Prospective studies have to clarify the role of adSOS and UBPS in fracture prediction.