|
| 中老年骨质疏松性骨折危险因素及预测模型 |
| Risk factors and a predictive model for osteoporotic fracture in the middle age and elderly patients |
| |
| DOI：10.3969/j.issn.1006-7108.2026.02.003 |
| 中文关键词: 骨质疏松 骨折 I型胶原N端前肽 I型胶原C末端交联肽 抗骨质疏松治疗 预测模型 |
| 英文关键词:osteoporosis fracture type I collagen N-terminal propeptide type I collagen C-terminal cross-linked peptide anti-osteoporosis treatment predictive model |
| 基金项目: |
|
| 摘要点击次数: 9 |
| 全文下载次数: 0 |
| 中文摘要: |
| 目的 调查中老年骨质疏松性骨折的发生率和危险因素，并建立定量预测模型指导临床筛查。方法 将2020年6月-2024年3月由我院诊断的315例骨质疏松症中老年患者临床资料作为建模集，另选择2024年5月-2025年5月我院收治的102例骨质疏松症中老年患者临床资料作为验证集。受试者年龄50~80岁，通过双能X线骨密度仪检测T值＜– 2.5 SD。建模集根据是否发生脆性骨折分为骨折组（55例）和无骨折组（260例）。单因素比较两组一般资料（性别、年龄、基础疾病等）、血生化[包括骨转换标志物I型胶原N端前肽（PINP）和I型胶原C末端交联肽（CTX）等]以及临床治疗（补充维生素D、钙剂、抗骨质疏松治疗等），然后通过多因素Logistic回归筛选危险因素。结果 与无骨折组相比，骨折组年龄≥70岁，糖尿病例数增多，补充维生素D、钙剂和抗骨质疏松治疗减少，血肌酐和CTX水平升高，PINP、血钙和血磷水平下降（P＜0.05）。回归分析显示，年龄≥70岁（OR=1.235, 95% CI：1.102~1.526，P＜0.001）、糖尿病（OR=1.326, 95% CI：1.127~1.625，P＜0.001）和CTX（OR=1.859,9 5% CI：1.425~2.235，P＜0.001）是危险因素，PINP（OR=0.526, 95% CI：0.302~0.758，P＜0.001）和抗骨质疏松治疗（OR=0.758, 95% CI：0.542~0.926，P＜0.001）是保护因素。建立预测模型Y=0.211×（年龄）+0.282×（糖尿病）+0.620×（CTX）– 0.642×（PINP）– 0.277×（抗骨质疏松治疗）。受试者工作特征（ROC）显示，预测模型评估建模集与验证集骨折发生的曲线下面积（AUC）分别为0.876（95% CI：0.822~0.953，P＜0.001）和0.855（95% CI：0.801~0.932，P＜0.001），提示模型的预测性能良好。结论 老年骨质疏松症有较高的骨折发生风险，年龄、糖尿病、CTX和PINP水平以及抗骨质疏松治疗与骨折发生密切相关，据此建立预测模型对评估骨折发生风险有较好的应用潜能。受限于单中心研究、样本量较少和缺乏外部数据验证，临床应用仍需谨慎。 |
| 英文摘要: |
| Objective To investigate the risk factors of osteoporotic fracture in the middle age and elderly and establish a quantitative predictive model to guide clinical screening. Methods 315 middle age andelderly patients diagnosed with osteoporosis in our hospital from June 2020 to March 2024 were included as model set, other 102 middle age andelderly patients with osteoporosis from May 2024 to May 2025 were included as validation set. They were aged 50-80 years, with dual energy X-ray bone density value < – 2.5 SD. They were divided into fracture group of 30 cases and non-fracture group of 30 cases in model set based on whether fragility fracture occurred. Univariate comparison of general information (gender, age, underlying diseases, etc.), blood biochemistry[including bone turnover markers-type I collagen N-terminal propeptide (PINP) and type I collagen C-terminal cross-linked peptide (CTX)], and clinical treatment (Vit D calcium supplementation, anti-osteoporosis treatment, etc.) between two groups, followed by multivariate logistic regression to screen risk factors. Results Compared with non-fracture group, the age≥70 years old of fracture group and diabetes increased, Vit D calcium supplementation and anti-osteoporosis treatment decreased, serum creatinine and CTX levels increased, serum PINP,calcium and phosphorus levels decreased (P<0.05). Regression analysis showed that age≥70 years old (OR=1.235, 95% CI:1.102-1.526, P<0.001), diabetes (OR=1.326, 95% CI:1.127-1.625, P<0.001) and CTX (OR=1.859, 95% CI=1.425-2.235, P＜0.001) were risk factors, PINP (OR=0.526, 95% CI: 0.302-0.758, P＜0.001) and anti-osteoporosis treatment (OR=0.758, 95% CI:0.542-0.926, P<0.001) was a protective factor. The predictive model Y=0.211×(age)+0.282×(diabetes) +0.620×(CTX)–0.642×(PINP)–0.277×(anti-osteoporosis treatment) was established. Receiver operating characteristic (ROC) analysis showed that area under curve (AUC) for predicting fracture occurrence in model set and validation set were 0.876 (95% CI:0.822-0.953, P<0.001) and 0.855 (95% CI: 0.801-0.932, P＜0.001), which suggested that predictive performance of the model was better. Conclusion Middle age andelderly osteoporosis has a higher risk of fracture. Age, diabetes, CTX, PINP and anti-osteoporosis treatment are closely related to fracture. Therefore, establishing a predictive model has a good potential for assessing fracture risk. Due to limitations in a single center, small sample size, and lack of external data validation, clinical application still require caution. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|