Correlation between osteoporotic fracture risk in Brazilian postmenopausal women calculated using the FRAX with and without the inclusion of bone densitometry data

• Published in: Archives of osteoporosis 2016-12, Vol.11 (1), p.16-16, Article 16
• Main Authors: Bastos-Silva, Yasmin, Aguiar, Luiza Borges, Pinto-Neto, Aarão M., Baccaro, Luiz Francisco, Costa-Paiva, Lúcia
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Algorithms; Bone Density; Bone Diseases, Metabolic - complications; Brazil; Cross-Sectional Studies; Densitometry - methods; Endocrinology; Female; Hip Fractures - etiology; Humans; Logistic Models; Medicine; Medicine & Public Health; Middle Aged; Original Article; Orthopedics; Osteoporosis - complications; Osteoporotic Fractures - etiology; Postmenopause; Reproducibility of Results; Risk Assessment - methods; Risk Factors; Statistics, Nonparametric
• ISSN: 1862-3522; 1862-3514
• DOI: 10.1007/s11657-015-0255-y

Summary The risks of fracture were calculated in 402 postmenopausal Brazilian women by FRAX with or without the inclusion of bone densitometry values. The correlation between the risk of major osteoporosis fracture or hip fracture calculated by FRAX with or without BMD was similar in this population, 0.76 and 0.64, respectively. Objective The objective of this study is to evaluate the degree of agreement between the 10-year fracture risk in postmenopausal Brazilian women, calculated using the WHO Fracture Risk Assessment Tool (FRAX)-Brazil, with and without the inclusion of bone mineral density (BMD) values. Methods A cross-sectional study was conducted with 402 postmenopausal women (≥40 years) who had undergone bone densitometry prior to initiating any pharmacological treatment for osteopenia or osteoporosis. The risks of a major osteoporosis fracture or hip fracture were calculated according to FRAX-Brazil either using clinical risk factors alone or with the inclusion of BMD value. The tests used were intraclass correlation coefficient, Mann-Whitney test, and univariate linear regression analysis. Results When the patients were classified according to the cutoff point defined as determining a high risk of a major osteoporosis fracture (≥20 %), 0.75 % were defined as high risk when BMD values were included and 1 % when they were not. With respect to the hip, 5.22 % were defined as having a high risk of fracture (≥3 %) when BMD values were included and 11.44 % when they were not. Intraclass correlation coefficient between the FRAX-Brazil assessed risk with and without the inclusion of BMD values was 0.76 (95 % CI 0.716–0.799) for a major osteoporosis fracture and 0.64 (95 %CI 0.583–0.698) for a hip fracture. Conclusion The correlation found for the FRAX-Brazil score obtained with and without the inclusion of BMD values was good for the risk of a major osteoporosis fracture and moderate for the risk of a hip fracture. The fracture risk calculated was similar in this population irrespective of whether or not BMD values were included in the algorithm.