Reference database for bone speed of sound measurement by a novel quantitative multi-site ultrasound device.

The nonuniform skeletal involvement in osteoporosis argues for multi-site evaluation. The Sunlight Omnisense (Sunlight Ultrasound Technologies, Israel) is a multi-site device that measures speed of sound (SOS) at the appendicular skeleton. We report the reference database for SOS at the radius (RAD), tibia (TIB), metatarsus (MTR) and phalanx (PLX). The database was obtained from 1521 healthy Israeli women (age 20-90 years) out of 2051 respondents. SOS was determined in 97.6% of the participants at the PLX, 96.4% at the TIB, 93.6% at the RAD and 85.1% at the MTR; it was not measurable in 0.5%. Short-term coefficient of variation was lowest at the RAD and always less than 1%. Maximal SOS was noted at 35-45 years of age in three of the sites (RAD 4169 m/s, MTR 3663 m/s, PLX 4047 m/s, respectively) but 10 years earlier at the TIB (3939 m/s). In the perimenopausal period (age 46-55 years), SOS was always lower in post- as compared with premenopausal women (p<0.05). Immediately following the menopause, SOS annually declined close to the short-term CV: 16, 34, 37 and 13 m/s at the RAD, PLX, MTR and TIB, respectively. The average age-stratified SOS values at various measurement sites were highly correlated at the population level (0.96-0.99), but less so at the individual level (0.40-0.57). Therefore, multi-site SOS measurements are better than single-site assessment. After 79 years of age, the average T-score at the RAD and PLX was <--2.5. This is similar to that of dual-energy X-ray absorptiometry (DXA)-determined spine bone mineral density (BMD) and somewhat lower than hip BMD. Equivalent T-score curves obtained by percentile adjustment of SOS at various sites to that of the RAD (at age group 60-69 years) reveal convergence and indicate that 52-68% of women older than 79 years are osteoporotic. In conclusion, multi-site peripheral SOS measurements reveal age-dependent bone changes with a high degree of measurement precision and indicate a prevalence of osteoporosis similar to that obtained by DXA.