Risk factors and bone mineral density in predicting the risk of fracture in postmenopausal women

Abstract

The Russian model FRAX®, an algorithm for estimating the 10-year absolute risk of fractures, which is based on the identification of risk factors that increase fracture risk, was proposed in 2012 to detect people at high risk for fracture.Objective: to estimate the sensitivity and specificity of the Russian model FRAX® versus dual energy X-ray absorptiometry (DEXA) for predicting high fracture risk.Patients and methods. In 2013–2014, the FRAX® questionnaire was retrospectively filled in provided that all information was available in the initial documents on 224 women aged 50 years and older (mean age, 62±7 years), examined at the V.A. Nasonova Research Institute of Rheumatology in 2003–2004. The risk of major osteoporotic fractures was assessed in accordance with the guidelines of the Russian Osteoporosis Association both with and without regard for bone mineral density (BMD+/BMD-) in the femoral neck. The diagnosis of osteoporosis (OP) was based on the WHO criteria using DEXA values.Results. At a primary examination, 96 (43%) patients had a history of minimal trauma fractures, 105 (47%) had OP in the vertebral column and/or femoral neck; the FRAX® (BMD-) values higher than the therapeutic intervention threshold were seen in 70 (31%) patients. 71 (31%) women had no risk factors included in the FRAX® questionnaire. In accordance with the current guidelines, therapy should be recommended for 146 (65%) patients. Over the 10-year period, minimal trauma fractures occurred in 106 (47%) women, including in 46 (40%) of the 128 patients who had no history of fractures. The sensitivity of the FRAX® algorithm with BMD- and BMD+ was 41% (31–51%) and 38% (29–48%) and its specificity was 77% (68–84%) and 82% (74–88%), respectively. The area un-der the ROC curve (AUC) was 0.66 for FRAX® BMD- and 0.69 for FRAX® BMD+. The sensitivity of BMD values in the spine for predicting OP fractures was greater than that of the FRAX® algorithm and was as high as 53% (43–63%) with a lower specificity of 61% (52–70%) (AUC, 0.61) and these values in the femoral neck were as follows: a sensitivity of 25% (18–35%) with a specificity of 89% (82–94%) (AUC 0.64). Conclusion. The Russian model FRAX® for major OP fractures, which is calculated both with and without regard for BMD in the femoral neck, and DEXA fail to identify in full measure patients who need antiosteoporotic therapy, which calls for further investigations providing a pharmacoeconomic rationale for the application of these methods.

Keywords

• frax
• диагностика остеопороза
• факторы риска
• прогнозирование переломов