OBJECTIVES The aims of this study were to develop and validate 2 simple scores for stratification of the risks of (1) any fragility (AF) and (2) major osteoporotic fracture (MOF) in type 2 diabetes (T2D) patients; we also compared the performance of these scores with that of the Fracture Risk Assessment Tool (FRAX) and its adjustments. DESIGN AND METHODS In this longitudinal cohort study, 1855 patients with T2D were enrolled from January 2015 to August 2019. Cox proportional hazard regression was used to model the 5-year risk of AF and MOF. These scores were internally validated using a bootstrap resampling method of 1000. RESULTS During a median follow-up of 5 years, 119 (6.42%) cases of AF and 92 (4.96%) cases of MOFs were identified. Both the concordance index (C-index) and calibration plots indicated improved identification performance using the newly established scores. Furthermore, these scores also showed improved outcomes regarding the decision curve analysis (DCA) and area under the curve (AUC) compared to the widely used FRAX and its derivatives. More importantly, these scores successfully separated T2D patients into risk groups according to significant differences in fracture incidence. CONCLUSIONS These novel scores enable simple and reliable fracture risk stratification in T2D patients. Future work is needed to validate these findings in external cohort(s).

中文翻译：

---

2 型糖尿病患者的简单脆性骨折风险评分：推导、验证、比较和风险分层研究。

目的 本研究的目的是开发和验证 2 种简单评分，用于对 2 型糖尿病 (T2D) 患者的 (1) 任何脆性 (AF) 和 (2) 严重骨质疏松性骨折 (MOF) 风险进行分层；我们还将这些分数的性能与骨折风险评估工具 (FRAX) 及其调整的性能进行了比较。设计和方法 在这项纵向队列研究中，2015 年 1 月至 2019 年 8 月期间入组了 1855 名 T2D 患者。使用 Cox 比例风险回归对 AF 和 MOF 的 5 年风险进行建模。这些分数使用 1000 的 bootstrap 重采样方法进行内部验证。 结果 在中位随访 5 年期间，发现了 119 例 (6.42%) AF 病例和 92 例 (4.96%) MOF 病例。一致性指数（C 指数）和校准图都表明使用新建立的分数提高了识别性能。此外，与广泛使用的 FRAX 及其衍生物相比，这些分数还显示决策曲线分析 (DCA) 和曲线下面积 (AUC) 的结果有所改善。更重要的是，这些评分根据骨折发生率的显着差异成功地将 T2D 患者分为危险组。结论 这些新颖的评分可以对 T2D 患者进行简单而可靠的骨折风险分层。未来的工作需要在外部队列中验证这些发现。