老年膝骨关节炎合并肌少症的影响因素分析及风险预测模型构建  

Influencing factors and a risk forecasting model for knee osteoarthritis complicated with sarcopenia in the aged

作　　者：陈敏[1] 余泽芸 宋丹 周礼 熊小琴 CHEN Min;YU Zeyun;SONG Dan;ZHOU Li;XIONG Xiaoqin(The First People’s Hospital of Yibin,Yibin 644000,Sichuan,China)

机构地区：[1]宜宾市第一人民医院,四川宜宾644000

出　　处：《中医正骨》2025年第3期23-28,38,共7页The Journal of Traditional Chinese Orthopedics and Traumatology

基　　金：四川省卫生和健康委员会科研课题(21PJ011)。

摘　　要：目的:探讨老年膝骨关节炎(knee osteoarthritis,KOA)合并肌少症的影响因素,并构建老年KOA合并肌少症的风险预测模型。方法:选取2020年6月至2024年12月在宜宾市第一人民医院住院治疗的KOA患者为研究对象。将2020年6月至2024年6月纳入的患者归入模型组(用于模型建立),将2024年7—12月纳入的患者归入为验证组(用于模型验证)。采用亚洲肌少症工作组制定的肌少症诊断方法诊断肌少症,收集患者性别、年龄、体质量指数、KOA病程、Kellgren-Lawrence分级、蛋白质摄入量、膝关节疼痛视觉模拟量表(visual analogue scale,VAS)评分、西安大略和麦克马斯特大学骨关节炎指数(Western Ontario and McMaster Universities osteoarthritis index,WOMAC)分级、吸烟、酗酒、规范治疗、规律运动、合并基础疾病、钙剂补充、维生素D补充等信息。将模型组患者根据是否合并肌少症,分为合并肌少症组和不合并肌少症组。先对合并肌少症组和不合并肌少症组患者的相关信息进行单因素对比分析,对其中组间差异有统计学意义的因素进行Lasso回归分析,将Lasso回归分析筛选出来的因素用于多因素Logistic回归分析。采用R语言和rms程序包构建老年KOA合并肌少症的列线图预测模型。分别基于模型组和验证组数据,采用受试者操作特征(receiver operating characteristic,ROC)曲线和Hosmer-Lemeshow拟合优度检验分别评价老年KOA合并肌少症列线图预测模型的区分度和校准度。结果:共纳入模型组患者675例,其中合并肌少症组196例,不合并肌少症组479例;纳入验证组患者77例。合并肌少症组和不合并肌少症组患者年龄、KOA病程、Kellgren-Lawrence分级、蛋白质摄入量、膝关节疼痛VAS评分、WOMAC分级、酗酒、规范治疗、规律运动、合并基础疾病、维生素D补充情况比较,组间差异均有统计学意义[(73.8±5.2)岁,(68.3±4.6)岁,t=12.921,P=0.000;(26.5±3.9)个月,(19.6�Objective:To explore the influencing factors of knee osteoarthritis complicated with sarcopenia in the aged,and to construct a risk prediction model for KOA complicated with sarcopenia in the aged.Methods:The KOA patients hospitalized at The First People’s Hospital of Yibin from June 2020 to December 2024 were selected as the subjects.The ones admitted from June 2020 to June 2024 were assigned into the model group(for model building),while those from July 2024 to December 2024 into the validation group(for model validation).The information of the patients,including gender,age,body mass index(BMI),KOA duration,Kellgren-Lawrence(K-L)grade,protein intake,knee pain visual analogue scale(VAS)score,Western Ontario and McMaster Universities osteoarthritis index(WOMAC)grade,smoking,alcohol abuse,standardized treatment,regular exercise,combined with underlying diseases,calcium supplementation,and vitamin D supplementation,was collected,and the sarcopenia was diagnosed among the included KOA patients using the diagnostic methods developed by the Asian Working Group for Sarcopenia(AWGS).According to the results,the KOA patients with and without sarcopenia in model group were subgrouped into a sarcopenia group and a non-sarcopenia group.After that,a single-factor analysis was conducted on the extracted information of patients in the 2 subgroups,followed by a Lasso regression analysis on the factors with statistically significant differences between the 2 subgroups,based on which a multi-factor logistic regression analysis on the factors screened by Lasso regression analysis was performed.According to the findings,a nomogram prediction model for KOA complicated with sarcopenia in the aged was constructed using the R language and rms package,and the discrimination and calibration performance of the nomogram prediction model were analyzed and evaluated by using the receiver operating characteristic(ROC)curve and Hosmer-Lemeshow goodness-of-fit(GOF)test based on the data of model group and validation group,respectively. Result

关 键 词：骨关节炎 膝 肌肉衰减征 老年人 LOGISTIC模型 因素分析 统计学 列线图表 风险 预测 

分 类 号：R684[医药卫生—骨科学]