Effect of excluding fractured or abnormal vertebrae on the trabecular bone score measurement

Brief rationale: The use of L1-L4 vertebrae, without exclusions, has been recommended for trabecular bone score (TBS) measurements.

Main result: Excluding abnormal and fractured vertebrae affected the TBS. Significance of the paper: Fracture or degenerative abnormality may not affect TBS. The preferred action may involve including all levels without exclusions.

Purpose: The use of L1-L4 vertebrae, without exclusions, has been recommended for trabecular bone score (TBS) measurements. We aimed to investigate the effect of excluding fractured or abnormal vertebrae from TBS.

Methods: Dual-energy X-ray absorptiometry images of 2767 participants, including 1080 without excluded vertebrae, 556 with fractured vertebrae, and 1131 with abnormal vertebrae showing a 1.0 T-score difference compared to the adjacent vertebrae, were retrospectively reviewed. Differences between TBS measurements with and without fractured or abnormal vertebrae were evaluated.

Results: Among 1080 participants without excluded vertebrae, TBS was 1.234 at L1, 1.296 at L2, 1.308 at L3, and 1.301 at L4. A significantly higher mean TBS was seen after excluding L1, whereas a significantly lower mean TBS was seen after excluding L2-L4. In the 556 participants with fractured vertebrae, excluding the involved level from the TBS calculation led to a significant difference in the total sample, women, ≥ 70 years old, and overweight subgroups. A significantly higher mean TBS was seen after excluding the fractured L1, whereas a significantly lower mean TBS was seen after excluding fractures at L2-L4. Among the 1131 participants with abnormal vertebrae, excluding the involved level from the TBS led to a significant difference in age, sex, and body mass index subgroups. Excluding abnormal L1 and L4 vertebrae led to a significantly higher and lower mean TBS, respectively.

Conclusion: Excluding fractured or abnormal vertebrae led to differences in TBS across various subgroups. Regarding the effect of vertebral level exclusion, the observed differences may be attributed to the systematic intervertebral variation, which is unrelated to any effect from fractures or degenerative abnormalities.