J. Damilakis et al., Imaging ultrasonometry of the calcaneus: Optimum T-score thresholds for the identification of osteoporotic subjects, CALCIF TIS, 68(4), 2001, pp. 219-224
The purpose of the present study was to (1) examine the age dependence of T
-score results for calcaneal imaging ultrasonometry and dual X-ray absorpti
ometry of the axial skeleton and (2) determine the optimum T-score threshol
ds appropriate for broadband ultrasound attenuation (BUA) and speed of soun
d (SOS) measurements. A total of 453 healthy women aged 20-79 years were in
cluded in the study. All study participants underwent bone mineral density
(BMD) measurements of the lumbar spine, femoral neck, total hip and calcane
al measurements of the BUA and SOS. An imaging ultrasound device (UBIS, DMS
, France) was used for the ultrasound measurements. T-scores were calculate
d using a subgroup of 71 healthy women aged 20-35 years to estimate the mea
n value of young normals and SD for BUA, SOS, and BMD. The age-related decl
ine in both BUA and SOS T-scores was slower than that in the equivalent fig
ures obtained by BMD measurements. The optimum T-score thresholds estimated
by receiver operating characteristic (ROC) analysis were -1.3 for BUA and
-1.5 for SOS. Using the optimum threshold, the sensitivity and specificity
for BUA was 68% and 83%, respectively. Corresponding values for SOS were 63
% and 79%. Utilizing calculated optimum T score thresholds for BUA and SOS,
the agreement among BUA, SOS, and BMD at the femoral neck was improved com
pared with that found using the T-score of less than or equal to -2.5 crite
rion. In conclusion, the definition of osteoporosis by a T-score of less th
an or equal to -2.5 was not applicable to imaging ultrasonometry of the cal
caneus. Optimum T-score thresholds were determined for both BUA and SOS sui
table to Ubis QUS device.