社区老年人肌少-骨质疏松症风险预测模型的构建  

Construction of a risk prediction model of osteosarcopenia in the elderly in communities

作　　者：刘柯汝 何耀宇 王玉环[1] 陶晶[2] 王若仙 魏杉杉 何斌[2] LIU Keru;HE Yaoyu;WANG Yuhuan;TAO Jing;WANG Ruoxian;WEI Shanshan;HE Bin(Shihezi University School of Medicine,Xinjiang 832002 China;The Third Affiliated Hospital of Shihezi University School of Medicine)

机构地区：[1]石河子大学医学院,新疆832002 [2]石河子大学医学院第三附属医院

出　　处：《护理研究》2024年第24期4333-4340,共8页Chinese Nursing Research

基　　金：新疆生产建设兵团科技创新人才计划项目,编号:2022CB010-04。

摘　　要：目的:探讨社区肌少-骨质疏松症高危人群的影响因素,并构建风险预测模型。方法:2023年3月—7月,采用分层随机抽样的方法,抽取石河子市20个社区1051名老年人为研究对象,将其分为肌少-骨质疏松症高危组和非肌少-骨质疏松症高危组。采用单因素分析和LASSO回归初步筛选变量,再经过Logistic回归分析确定预测变量,使用R4.3.1软件构建列线图预测模型并进行验证。结果:1051名老年人肌少-骨质疏松症检出率为21.9%。年龄、性别、体质指数、合并症、骨折史、医保类型、是否饮用浓茶/咖啡或碳酸饮料、每日久坐时长、老年人活动情况、营养状况、社会衰弱及抑郁是社区肌少-骨质疏松症高危人群的影响因素。构建的风险预测模型拟合度良好,受试者工作特征曲线下面积为0.956,区分度良好;校准曲线实际值与预测值的平均绝对误差为0.014,具有良好的准确度;决策曲线结果显示临床有效性良好。结论:社区老年人肌少-骨质疏松症风险预测模型具有科学性与实用性,可用于社区肌少-骨质疏松症高危人群的筛查。Objective:To investigate the influencing factors of high risk population of osteosarcopenia in communities,and to establish a risk prediction model.Methods:From March to July 2023,a total of 1051 elderly people from 20 communities in Shihezi city were selected by stratified random sampling method and divided into high risk groups of osteosarcopenia and high risk groups of non⁃osteosarcopenia.Univariate analysis and LASSO regression were used to preliminarily screen variables,and then the predictive variables were determined by Logistic regression analysis.The Nomogram prediction model was constructed and verified by R4.3.1 software.Results:The incidence of osteosarcopenia in 1051 elderly patients was 21.9%.Age,sex,body mass index,comorbidities,history of fracture,type of health insurance,consumption of strong tea,coffee or carbonated beverages,daily sedentary time,activity of the elderly,nutritional status,social frailty and depression were the influencing factors for community groups at high risk of osteosarcopenia.The risk prediction model had a good fit,and the area under the receiver operating characteristic curve was 0.956,indicating a good differentiation.The average absolute error between the actual value and the predicted value of the calibration curve was 0.014,which has good accuracy.The results of decision curve showed good clinical effectiveness.Conclusions:The risk prediction model of osteosarcopenia in elderly people in communities is scientific and practical,and it can be used for the screening of community osteosarcopenia high⁃risk population.

关 键 词：社区 老年人 肌少-骨质疏松症 列线图 预测模型 

分 类 号：R473.2[医药卫生—护理学]