社区老年人高血压合并糖尿病临床预测模型的建立与验证  

Establishment and verification of clinical prediction model of hypertensioncomplicated with diabetes in elderly people in community

作　　者：朱先尚 毛华博 程虎 王建成[2,4] 徐银娟 ZHU Xianshang;MAO Huabo;CHENG Hu;WANG Jiancheng;XU Yin juan(First Clinical Medical College,Gansu University of Traditional Chinese Medicine,Lanzhou,Gansu 730000,China;不详)

机构地区：[1]甘肃中医药大学第一临床医学院,甘肃兰州730000 [2]甘肃省人民医院全科医学科,甘肃兰州730000 [3]兰州大学公共卫生学院,甘肃兰州730000 [4]甘肃卫生职业学院临床医学院,甘肃兰州730000

出　　处：《中华全科医学》2025年第5期737-741,793,共6页Chinese Journal of General Practice

基　　金：国家自然科学基金项目(72264002);中国医药教育协会重大项目(2022KTZ010);甘肃省人民医院院内科研基金项目(22GSSYD-54)。

摘　　要：目的 分析社区老年高血压患者发生糖尿病的危险因素,并构建预测模型,以期为社区老年高血压患者合并糖尿病的防控提供参考依据。方法 收集2023年4月—2024年3月在团结新村社区卫生服务中心参与慢性病管理的1 859例≥65岁老年高血压患者的临床资料。将数据采用简单随机抽样的方法按7∶3的比例拆分为训练集(1 301例)和验证集(558例),通过LASSO回归对变量降维,进行多因素logistic回归分析研究高血压合并糖尿病的危险因素,构建Nomogram预测模型,并使用验证集进行内部验证。使用受试者工作特征曲线评估模型区分度,绘制校正曲线并进行H-L检验评估模型一致性,通过决策分析评估模型有效性。结果 在1 859例社区老年高血压患者中,有546人(29.37%)患有糖尿病。多因素logistic回归显示,慢性病家族史、乏力、胸闷和嗜糖均为高血压合并糖尿病的独立危险因素;而认知功能较好和日常锻炼则为保护性因素。训练集构建预测模型的AUC为0.873(95%CI:0.850~0.873),校正曲线斜率接近1,且H-L拟合优度检验χ^(2)=6.511,P=0.260,一致性较好。决策曲线提示模型能够产生净获益。结论 该社区老年人高血压合并糖尿病的患病率较高,基于社区可测量的6个临床特征构建的预测模型在预测高血压合并糖尿病的发生时表现出良好的预测能力和临床应用价值。Objective The objective of this study is two-fold:firstly,to analyze the risk factors associated with diabetes in elderly patients suffering from hypertension in the community,and secondly,to construct a prediction model.The ultimate aim of this study is to provide a reference for the prevention and management of diabetes in elderly patients suffering from hypertension in the community.Methods From April 2023 to March 2024,the clinical data of 1859 elderly patients with hypertension aged 65 years and over who participated in chronic disease management in the Tuanjie Community Health Service Centre were collected for the purpose of the study.The data were randomly partitioned into a training set comprising 1301 cases and a validation set consisting of 558 cases,at a ratio of 7∶3 by means of a simple random sampling method.The LASSO regression method was employed to reduce the dimensionality of the variables,while multivariate logistic regression analysis was conducted to investigate the risk factors associated with hypertension in conjunction with diabetes.Subsequently,a Nomogram prediction model was developed.The validation set was utilized for the purpose of internal validation.The receiver operating characteristic curve(AUC)was employed to evaluate the discrimination of the model,the calibration curve was delineated and the Hosmer-Lemeshow test was used to evaluate the consistency of the model,and the decision analysis was used to evaluate the validity of the model.Results In a study of 1859 elderly hypertensive patients,546(29.37%)were found to have diabetes mellitus.Multivariate logistic regression analysis demonstrated that a family history of chronic diseases,fatigue,chest tightness and sugar addiction were independent risk factors for hypertension combined with diabetes.Nevertheless,superior cognitive function and the undertaking of daily exercise were found to be protective factors.The area under the curve(AUC)of the prediction model constructed from the training set was 0.873(95%CI:0.850-0.873).The cal

关 键 词：高血压 糖尿病 社区 老年人 预测模型 

分 类 号：R544.1[医药卫生—心血管疾病] R587.1[医药卫生—内科学]