Enhanced prediction of fracture risk combining vertebral fracture status and BMD

Abstract

Prevalent vertebral fractures are associated with increased fracture risk, but the magnitude of this effect across a range of BMD T-scores has not been quantified. In this analysis, for any given BMD T-score, incident fracture risk varied up to twelve fold when information regarding prevalent radiographic vertebral fracture status was considered.

Background: Clinical fracture risk evaluation of older women usually includes assessment of bone mineral density (BMD) but often not vertebral fracture status. In this analysis, we quantified the impact of vertebral fracture burden on two year fracture risk across a range of BMD T-scores.

Methods: Data were from 2,651 postmenopausal women who were assigned to the placebo groups of the Fracture Prevention Trial (median observation 21 months) and the Multiple Outcomes of Raloxifene Evaluation Trial (MORE; observation 2 years). Using the Genant visual semiquantitative criteria, we defined prevalent vertebral fracture status as: a) presence or absence of fracture; b) fracture number; c) maximum semi-quantitative (SQ) score (normal=0, mild fracture=1, moderate fracture=2, severe fracture=3); and d) spinal deformity index (SDI) score (sum of SQ scores of T4 to L4 vertebrae). Incident fractures over two years were identified via lateral spine radiographs and outside the spine by questioning of patients and review of radiographs or radiographic reports.

Results: Femoral neck BMD T-score provided significant information regarding fracture risk. Across the range of T-scores, vertebral fracture status provided additional prognostic information. The risk increased with increasing number and severity of prevalent vertebral fractures and SDI, a summary measure of spine fracture burden. Across a range of BMD values, prevalent spine fracture burden as assessed by SDI increased the risk of incident vertebral fractures by up to 12-fold, nonvertebral fractures by about twofold, and any fractures by up to sevenfold.

Conclusions: These findings indicate that at any given BMD T-score, the risk of incident vertebral, non-vertebral, and any fracture depended heavily on prevalent radiographic vertebral fracture status. Assessment of vertebral fracture status, in addition to BMD, provides practical and relevant clinical information to aid in predicting fracture risk in postmenopausal women.