基于列线图构建老年骨质疏松性腰椎骨折影响因素预测模型及应用价值评估  

Development and evaluation of a nomogram-based prediction model for influencing factors of osteoporotic lumbar vertebral fractures in the elderly and its application value

作　　者：杜万里 杜洋 李克乾[3] 刘巍巍 DU Wanli;DU Yang;LI Keqian;LIU Weiwei(School of Basic Medicine,School of Medicine,Hubei University of Science and Technology,Hubei Province,Xianning437100,China;Department of Traumatology,Suizhou Central Hospital,Hubei University of Medicine,Hubei Province,Suizhou441300,China;Department of Spine Surgery,Suizhou Central Hospital,Hubei University of Medicine,Hubei Province,Suizhou441300,China)

机构地区：[1]湖北科技学院医学部基础医学院,湖北咸宁437100 [2]湖北医药学院附属随州市中心医院创伤骨科,湖北随州441300 [3]湖北医药学院附属随州市中心医院脊柱外科,湖北随州441300

出　　处：《中国医药导报》2024年第15期9-12,共4页China Medical Herald

基　　金：湖北省卫生健康委员会科研项目(WJ2019F140)。

摘　　要：目的 通过列线图构建老年骨质疏松性腰椎骨折的影响因素预测模型,并评估其准确性。方法 纳入2020年1月至2023年1月湖北医药学院附属随州市中心医院收治的骨质疏松性腰椎骨折老年患者82例为观察组,另选取同期在湖北医药学院附属随州市中心医院体检的健康老年人82名设为对照组。收集其一般资料及相关临床资料,分析老年骨质疏松性腰椎骨折的影响因素,构建列线图预测模型;列线图预测模型评价采用校正曲线和决策曲线分析。结果 两组年龄、吸烟史、骨折史、骨质疏松分级、血清脂联素(APN)、Ⅰ型胶原β降解产物(β-CTX)、Ⅰ型前胶原N-端前肽(PⅠNP)、骨钙素(OC)、骨碱性磷酸酶(BAP)、25羟基维生素D_(3)[25(OH)D_(3)]及腰椎骨密度(BMD)比较,差异有统计学意义(P<0.05)。年龄、吸烟史、骨折史、骨质疏松分级、APN、β-CTX、PⅠNP、OC、BAP、25(OH)D_(3)、BMD均是老年骨质疏松性腰椎骨折的影响因素(P<0.05)。列线图预测模型拟合度良好;风险阈值在0.2~0.8时,列线图预测模型具有最大净效益。列线图预测模型预测老年骨质疏松性腰椎骨折的曲线下面积为0.980。结论 老年骨质疏松性腰椎骨折是一个多因素导致的复杂问题。通过对相关风险因素(年龄、吸烟史、骨折史、骨质疏松分级、APN、β-CTX、PⅠNP、OC、BAP、25(OH)D_(3)、BMD)的全面认知有望减少老年人骨质疏松性腰椎骨折的风险。Objective To create a risk prediction model for osteoporotic lumbar fractures in the elderly through a nomogram,and to evaluat its precision.Methods A total of 82 elderly patients with osteoporotic lumbar fractures admitted to Suizhou Central Hospital,Hubei University of Medicine from January 2020 to January 2023 were included as the observation group,In addition,82 healthy elderly people who underwent physical examination in Suizhou Central Hospital,Hubei University of Medicine were selected as the control group.The general data and related clinical data were collected,and the influencing factors of osteoporotic lumbar fractures in the elderly were analyzed.Correction curve and decision curve analysis are used to evaluate the nomogram prediction model.Results There were statistically significant differences in age,smoking history,fracture history,osteoporosis grade,serum adiponectin(APN),β-C-terminal telopeptides of type Ⅰ collagen(β-CTX),procollagen type Ⅰ N-terminal propeptide(PⅠNP),osteocalcin(OC),bone alkaline phosphatase(BAP),25 hydroxyvitamin D_3(25 [OH] D_3),and lumbar bone mineral density(BMD) between the two groups(P<0.05).Age,smoking history,fractures history,osteoporosis grade,APN,β-CTX,PⅠNP,OC,BAP,25(OH)D_3,and BMD were all independent risk factors for osteoporotic lumbar vertebral fractures in the elderly(P<0.05).The nomogram prediction model has a good fit.When the risk threshold was 0.2-0.8,the nomogram prediction model has the greatest net benefit.The nomogram prediction model predicted that area under the curve of osteoporotic lumbar fractures in the elderly was 0.980.Conclusion Osteoporotic lumbar fracture in the elderly is a complex problem caused by multiple factors.Comprehensive knowledge of relevant risk factors(age,smoking history,fracture history,osteoporosis grade,APN,β-CTX,PⅠNP,OC,BAP,25(OH) D_3,BMD) is expected to reduce the risk of osteopolytic lumbar fracture in the elderly.

关 键 词：骨质疏松性腰椎骨折 骨转换生化指标 危险因素 LOGISTIC回归分析 

分 类 号：R687.3[医药卫生—骨科学]