Bone Abstracts

Glucocorticoids inhibit osteoblast function and cause an increase in osteoblast and osteocyte apoptosis. Bone remodelling defects occur resulting in an increase in fracture risk that cannot be fully explained by decreases in bone mineral density (BMD). We propose that this may be due to alterations in bone quality. Trabecular bone score (TBS) correlates with 3D bone micro-architectural parameters and can be derived directly from grey-level variations within 2D DXA images.

We assessed the ability of BMD, TBS, and BMD with TBS (BMD+TBS) to discriminate between healthy women and i) glucocorticoid-treated women and ii) women with recent fractures.

Locally recruited older women (n=484, ages 55–79 years) had either i) taken prednisolone ≥5 mg/day (or equivalent) for >3 months (n=64, average dose range=5.0–20.0 mg/day) or ii) sustained a recent fracture of the distal forearm (n=46), proximal humerus (n=37), vertebra (n=30), or proximal femur (n=28). They were compared to healthy population-based women without prevalent fractures (n=279). Lumbar spine BMD was measured by DXA (Hologic QDR 4500A) and TBS values were derived following scan image reanalysis using TBS – Clinical Data Analysis software v1.6 (Med-Imaps).

BMD+TBS values were calculated using logistic regression analysis. The discriminatory ability; area under the curve (AUC); of BMD, TBS and BMD+TBS for prevalent fracture or glucocorticoid use was determined using receiver operator characteristic (ROC) analysis. The AUCs for i) BMD and TBS; ii) BMD and BMD+TBS; and iii) TBS and BMD+TBS were compared using pairwise comparisons of ROC curves (P<0.05).