Abstract
Trabecular bone score (TBS) is used for fracture prediction in adults, but its utility in children is limited by absence of appropriate reference values. We aimed to develop reference ranges for TBS by age, sex, and population ancestry for youth ages 5 to 20 years. We also investigated the association between height, body mass index (BMI), and TBS, agreement between TBS and lumbar spine areal bone mineral density (aBMD) and bone mineral apparent density (BMAD) Z-scores, tracking of TBS Z-scores over time, and precision of TBS measurements. We performed secondary analysis of spine dual-energy X-ray absorptiometry (DXA) scans from the Bone Mineral Density in Childhood Study (BMDCS), a mixed longitudinal cohort of healthy children (n = 2014) evaluated at five US centers. TBS was derived using a dedicated TBS algorithm accounting for tissue thickness rather than BMI. TBS increased only during ages corresponding to pubertal development with an earlier increase in females than males. There were no differences in TBS between African Americans and non-African Americans. We provide sex-specific TBS reference ranges and LMS values for calculation of TBS Z-scores by age and means and SD for calculation of Z-scores by pubertal stage. TBS Z-scores were positively associated with height Z-scores at some ages. TBS Z-scores explained only 27% and 17% of the variance of spine aBMD and BMAD Z-scores. Tracking of TBS Z-scores over 6 years was lower (r = 0.47) than for aBMD or BMAD Z-scores (r = 0.74 to 0.79), and precision error of TBS (2.87%) was greater than for aBMD (0.85%) and BMAD (1.22%). In sum, TBS Z-scores provide information distinct from spine aBMD and BMAD Z-scores. Our robust reference ranges for TBS in a well-characterized pediatric cohort and precision error estimates provide essential tools for clinical assessment using TBS and determination of its value in predicting bone fragility in childhood and adolescence.