Abstract：
The SPAH study is a population-based prospective cohort of Brazilian community-dwelling elderlies with higher fracture risk than observed in the studies used to construct the Brazilian FRAX model. In this study, the FRAX tool was a good fracture predictor within this high-risk elderly cohort, especially when calculated without bone density.
OBJECTIVE: To determine the performances of FRAX and age-dependent intervention thresholds according to National Osteoporosis Guideline Group (NOGG) guidelines with and without bone mineral density (BMD) regarding fracture prediction in community-dwelling elderly Brazilians.
METHODS: Seven hundred and five older adults (447 women; 258 men) were followed for 4.3 ± 0.8 years. FRAX risk for hip and major osteoporotic fractures with and without BMD was calculated at baseline. The bivariate analysis investigated the associations between the absolute probability of fracture (FRAX), as well as the age-dependent intervention thresholds (NOGG), and the incidence of vertebral fracture (VF), non-vertebral fracture (NVF), and major osteoporotic fractures (MOF), segregated by sex. Age-adjusted Poisson\'s multiple regression and ROC curves were constructed to determine FRAX and NOGG\'s accuracies as fracture predictors.
RESULTS: Fractures occurred in 22% of women and 15% of men. FRAX with and without BMD was higher in women with all types of fractures (p < 0.001). Only NOGG risk classification without BMD was associated with NVF (p = 0.047) and MOF (p = 0.024). FRAX was associated with NVF in the multiple regression, regardless of BMD. ROC curves of FRAX with and without BMD had AUCs of 0.74, 0.64, and 0.61 for NVF, VF, and MOF, respectively. The most accurate risk cutoffs for FRAX were 8% for MOF and 3% for hip fractures. No statistically significant associations were found in men.
CONCLUSIONS: FRAX predicted NVF more accurately than VF or MOF in elderlies, regardless of BMD. These results reiterate that FRAX may be used without BMD, even considering that Brazilian elderlies have known higher fracture risk.
OBJECTIVE: To determine the performances of FRAX and age-dependent intervention thresholds according to National Osteoporosis Guideline Group (NOGG) guidelines with and without bone mineral density (BMD) regarding fracture prediction in community-dwelling elderly Brazilians.
METHODS: Seven hundred and five older adults (447 women; 258 men) were followed for 4.3 ± 0.8 years. FRAX risk for hip and major osteoporotic fractures with and without BMD was calculated at baseline. The bivariate analysis investigated the associations between the absolute probability of fracture (FRAX), as well as the age-dependent intervention thresholds (NOGG), and the incidence of vertebral fracture (VF), non-vertebral fracture (NVF), and major osteoporotic fractures (MOF), segregated by sex. Age-adjusted Poisson\'s multiple regression and ROC curves were constructed to determine FRAX and NOGG\'s accuracies as fracture predictors.
RESULTS: Fractures occurred in 22% of women and 15% of men. FRAX with and without BMD was higher in women with all types of fractures (p < 0.001). Only NOGG risk classification without BMD was associated with NVF (p = 0.047) and MOF (p = 0.024). FRAX was associated with NVF in the multiple regression, regardless of BMD. ROC curves of FRAX with and without BMD had AUCs of 0.74, 0.64, and 0.61 for NVF, VF, and MOF, respectively. The most accurate risk cutoffs for FRAX were 8% for MOF and 3% for hip fractures. No statistically significant associations were found in men.
CONCLUSIONS: FRAX predicted NVF more accurately than VF or MOF in elderlies, regardless of BMD. These results reiterate that FRAX may be used without BMD, even considering that Brazilian elderlies have known higher fracture risk.
摘要：
SPAH研究是巴西社区居住的老年人的基于人群的前瞻性队列，其骨折风险高于用于构建巴西FRAX模型的研究。在这项研究中,FRAX工具在这个高风险的老年人队列中是一个很好的骨折预测指标，特别是在没有骨密度的情况下计算。
目的：根据国家骨质疏松指南组（NOGG）指南，确定FRAX的性能和年龄依赖性干预阈值，该指南对社区居住的老年巴西人的骨折预测有或没有骨密度（BMD）。
方法：对75名老年人（447名女性；258名男性）进行了4.3±0.8年的随访。在基线计算有和没有BMD的髋部和严重骨质疏松性骨折的FRAX风险。双变量分析调查了骨折绝对概率(FRAX)之间的关联，以及年龄依赖性干预阈值(NOGG)，和椎骨骨折（VF）的发生率，非椎体骨折(NVF),和严重的骨质疏松性骨折（MOF），按性别隔离。构建年龄调整的泊松多元回归和ROC曲线，以确定FRAX和NOGG作为骨折预测因子的准确性。
结果：22％的女性和15％的男性发生骨折。在所有类型骨折的女性中，有和没有BMD的FRAX均较高（p<0.001）。仅无BMD的NOGG风险分类与NVF（p=0.047）和MOF（p=0.024）相关。在多元回归中，FRAX与NVF相关，不管BMD。具有和不具有BMD的FRAX的ROC曲线对于NVF具有0.74、0.64和0.61的AUC,VF,MOF,分别。FRAX最准确的风险临界值为MOF的8％和髋部骨折的3％。在男性中没有发现统计学上显著的关联。
结论：FRAX比VF或MOF更准确地预测老年人的NVF，不管BMD。这些结果重申，FRAX可以在没有BMD的情况下使用，即使考虑到巴西老年人已知骨折风险较高。
目的：根据国家骨质疏松指南组（NOGG）指南，确定FRAX的性能和年龄依赖性干预阈值，该指南对社区居住的老年巴西人的骨折预测有或没有骨密度（BMD）。
方法：对75名老年人（447名女性；258名男性）进行了4.3±0.8年的随访。在基线计算有和没有BMD的髋部和严重骨质疏松性骨折的FRAX风险。双变量分析调查了骨折绝对概率(FRAX)之间的关联，以及年龄依赖性干预阈值(NOGG)，和椎骨骨折（VF）的发生率，非椎体骨折(NVF),和严重的骨质疏松性骨折（MOF），按性别隔离。构建年龄调整的泊松多元回归和ROC曲线，以确定FRAX和NOGG作为骨折预测因子的准确性。
结果：22％的女性和15％的男性发生骨折。在所有类型骨折的女性中，有和没有BMD的FRAX均较高（p<0.001）。仅无BMD的NOGG风险分类与NVF（p=0.047）和MOF（p=0.024）相关。在多元回归中，FRAX与NVF相关，不管BMD。具有和不具有BMD的FRAX的ROC曲线对于NVF具有0.74、0.64和0.61的AUC,VF,MOF,分别。FRAX最准确的风险临界值为MOF的8％和髋部骨折的3％。在男性中没有发现统计学上显著的关联。
结论：FRAX比VF或MOF更准确地预测老年人的NVF，不管BMD。这些结果重申，FRAX可以在没有BMD的情况下使用，即使考虑到巴西老年人已知骨折风险较高。
