Background Bone mineral density (BMD) may underestimate fracture risk since most patients with fracture had osteopenia or normal BMD. The WHO Fracture Risk Assessment Tool (FRAX) incorporated BMD and clinical risk factors (CRF) to estimate the 10-year probability of major osteoporotic fractures (MOF) and hip fracture. FRAX has been adopted by various guidelines to assess the need for therapeutic intervention. Whether FRAX is superior to BMD in identifying high-risk patients in need of anti-osteoporotic treatment is worth exploring.
Objectives To compare the ability of FRAX and BMD alone in discriminating patients with and without fracture, and their effectiveness in identifying anti-osteoporotic treatment needs
Methods 1300 rheumatic disease patients on long-term glucocorticoid were screened for vertebral fracture by radiograph from 7 rheumatology clinics. 220 ambulatory patients (110 with vertebral fracture and 110 without vertebral fracture) were recruited and assessed on CRF with FRAX questionnaire, BMD at the lumbar spine, total hip and femoral neck by dual-energy x-ray absorptiometry (DXA). The ten-year probability of MOF and hip fracture (FRAX score) with and without femoral neck BMD were calculated with the WHO FRAX tool.
Results Both groups of patients were matched in gender, disease type and cumulative glucocorticoid dose (p>0.1). Patients in the fracture group were older (63±14 vs 59±12 years, p<0.01), of lower body mass index [22.7±3.5vs 24±3.7kg/m2, p<0.05], with a higher prevalence of previous fracture [38 (34.5%) vs 5 (4.5%), p<0.01], use of vitamin D [77 (70%)vs 54 (49.1%), p<0.01] and bisphosphonates [23 (20.9%) vs 7 (6.4%), p<0.01]. A receiver operating characteristic (ROC) curve analysis was performed to determine the ability of FRAX score with or without BMD and BMD alone to discriminate fracture status (Table 1). BMD at the femoral neck and FRAX score with BMD were able to provide adequate discrimination with their area under curve (AUC) >0.70. With regards to treatment indication, only 53/110 (48.2%) of the patients with vertebral fracture were identified as having osteoporosis (T-score ≤ -2.5 at hip/spine) by DXA. In contrast, FRAX with BMD were able to identify 71/110 (64.5%) of the fracture patients whom treatment criteria were met, with a power of 0.97 and effect size of 0.38 (a medium effect) at the 0.05 level.
Conclusions In rheumatic patients on long-term glucocorticoid, FRAX score was more effective in identifying patients with intervention needs than BMD alone. It is suggested that FRAX should be used as a routine assessment to identify patients, even asymptomatic, indicated for anti-osteoporotic treatment.
Aspray, T. J. (2015). Fragility fracture: recent developments in risk assessment. Therapeutic Advances in Musculoskeletal Disease, 7(1), 17–25. http://doi.org/10.1177/1759720X14564562.
Rentero, M. L., Amigo, E., Chozas, N., Fernández Prada, M., Silva-Fernández, L., Abad Hernandez, M. A., ... del Pino-Montes, J. (2015). Prevalence of fractures in women with rheumatoid arthritis and/or systemic lupus erythematosus on chronic glucocorticoid therapy. BMC Musculoskeletal Disorders, 16(1), 300. article. https://doi.org/10.1186/s12891–015–0733–9.
Acknowledgements We would like to acknowledge the Research Grant Council of Hong Kong for funding support (RGC Ref No.14113714).
Disclosure of Interest None declared