Abstract
Introduction
DXA areal-bone-mineral-density (aBMD) is used clinically as a surrogate for true volumetric-BMD to assess bone fragility. Trabecular-Bone-Score (TBS) provides an assessment of bone quality based on the DXA-derived two-dimensional images. Calculated from bone area (BA), aBMD may under- or overestimate true BMD in individuals with relatively low and high BA respectively. This study investigated relationships between BA at the lumbar-spine (L1–L4) and measurements of BMD and TBS.
Method
Lumbar spine scans were performed (GE Lunar Prodigy) on 114 women (mean 53 yrs). The study population was divided by L1–L4 BA using the 20th and 80th centiles, and BMD v TBS correlations calculated for the subgroups. BMD and TBS, converted to Z-scores, were correlated with BA.
Results
For the whole group, r = 0.23 for BMD v BA, whereas r = −0.12 for TBS v. BA: the correlation between BMD and TBS was r=0.38. Correlations between BMD and TBS for BA subgroups were r = 0.23, r=0.44 and r=0.59 for the >20th, 20th −80th and >80th centile groups respectively.
Conclusion
Low correlation between BMD and TBS for the entire group indicates that TBS reflects some bone quality properties unrelated to BMD. The correlation of BMD with BA suggests that it incorrectly estimates true BMD at BA extremes. Correlation for BMD v TBS in the lowest area-subgroup was substantially lower than for the middle and highest area subgroups suggesting that TBS may afford valuable complementary information on bone fragility at the lowest extreme of BA where distortions in BMD measurement potentially occur.
