作者：邓银辉 ，李娜 ，王亚如 ，熊忱 ，邹小芳

单位：

1.510150广东省广州市，广州医科大学附属第三医院护理部　2.511436广东省广州市，广州医科大学护理学院

单位（英文）：

1.Nursing Department, the Third Affiliated Hospital of Guangzhou Medical University, Guangzhou 510150, China 2.School of Nursing, Guangzhou Medical University, Guangzhou 511436, China

关键词：

糖尿病，2型；高血压；老年人；住院病人；认知衰弱；列线图

关键词（英文）：

Diabetes mellitus, type 2; Hypertension; Elderly; Inpatients; Cognitive frailty; Nomograms

中图分类号：

R 587.1 R 544.1

DOI：

10.12114/j.issn.1008-5971.2023.00.311

基金项目：

2022年广东省卫生健康委卫生健康适宜技术推广项目（202206230855074874）

摘要：

目的　分析老年糖尿病合并高血压住院患者发生认知衰弱（CF）的影响因素，并构建其风险预测列 线图模型。方法　2022年9月至2023年3月，采用便利抽样法选取广州医科大学附属第三医院收治的老年糖尿病合并高 血压住院患者252例为研究对象。分别采用衰弱表型量表、蒙特利尔认知评估量表、一般资料调查表、微型营养评定 短表、简版老年抑郁量表、Barthel指数量表、Morse跌倒评估量表调查患者的衰弱情况、认知功能、一般资料、营养状 况、抑郁状态、生活自理能力及跌倒风险，统计患者CF发生情况。老年糖尿病合并高血压住院患者发生CF的影响因素 分析采用多因素Logistic回归分析；使用R语言中的“rms”程序包构建老年糖尿病合并高血压住院患者发生CF风险预 测列线图模型，采用ROC曲线评价该列线图的区分度，采用Hosmer-Lemeshow拟合优度检验评价该列线图模型的拟合 程度，采用Brier得分及校准曲线评价该列线图模型的校准度。结果　本研究共发放252份问卷，回收有效问卷251份， 有效回收率为99.6% 。 251例老年糖尿病合并高血压住院患者发生 CF 76例，CF发生率为30.3%。按照7∶3的比例将患 者分为建模集（ n=175 ）和验证集（ n=76）。多因素Logistic回归分析结果显示，年龄、脑力活动、抑郁、生活自理能 力、跌倒风险是老年糖尿病合并高血压住院患者发生CF的独立影响因素（P＜0.05）。基于上述影响因素构建老年糖 尿病合并高血压住院患者发生CF风险预测列线图模型。ROC曲线分析结果显示，该列线图模型预测建模集老年糖尿病 合并高血压住院患者发生CF的AUC为0.893〔95%CI（0.842，0.945）〕，Brier得分为0.116分；该列线图模型预测验证 集老年糖尿病合并高血压住院患者发生CF的AUC为0.836〔95%CI （0.741 ，0.930 ）〕，Brier得分为0.146分。Hosmer Lemeshow拟合优度检验结果显示，该列线图模型拟合较好（χ 2 =5.97 ，P=0.65 ）。校准曲线分析结果显示，该列线图 模型预测建模集老年糖尿病合并高血压住院患者CF的发生概率与实际发生率相近。结论　年龄、脑力活动、抑郁、生 活自理能力、跌倒风险是老年糖尿病合并高血压住院患者发生CF的独立影响因素，而基于上述影响因素构建的列线图 模型对老年糖尿病合并高血压住院患者发生CF具有良好的区分度、校准度。

英文摘要：

Objective To analyze the influencing factors of cognitive frailty (CF) in elderly inpatients with diabetes mellitus and hypertension, and construct its risk prediction nomogram model. Methods From September 2022 to March 2023, a total of 252 elderly inpatients with diabetes mellitus and hypertension admitted to the Third Affiliated Hospital of Guangzhou Medical University were selected as subjects by convenience sampling method. The Fried's Frailty Phenotype, Montreal Cognitive Assessment, General Information Questionnaire, Mini Nutritional Assessment-Short Form, Geriatric Depression Scale, Barthel Index Scale and Morse Fall Scale were used to investigate the frailty, cognitive function, general information, nutritional status, depression, self-care ability and fall risk of patients, and the occurrence of CF in patients was recorded. Multivariate Logistic regression analysis was used to analyze the influencing factors of CF in elderly inpatients with diabetes mellitus and hypertension; the "rms" package in R language was used to construct a nomogram model for predicting CF risk in elderly inpatients with diabetes mellitus and hypertension. The ROC curve was used to evaluate the discrimination of the nomogram model, the Hosmer-Lemeshow goodness of fit test was used to evaluate the fitting degree of the nomogram model, and the Brier score and calibration curve were used to evaluate the calibration degree of the nomogram model. Results A total of 252 questionnaires were distributed in this study, and 251 valid questionnaires were recovered, with an effective recovery rate of 99.6%. There were 76 cases of CF in 251 elderly inpatients with diabetes mellitus and hypertension, and the incidence of CF was 30.3 %. According to the ratio of 7∶3, the patients were divided into the modeling set ( n=175) and the validation set (n=76) . Multivariate Logistic regression analysis results showed that age, mental activity, depression, self-care ability and fall risk were independent influencing factors of CF in elderly inpatients with diabetes mellitus and hypertension (P < 0.05) . Based on the above influencing factors, the nomogram model for predicting CF in elderly inpatients with diabetes mellitus and hypertension was constructed. ROC curve analysis showed that the AUC of the nomogram model for predicting CF in elderly inpatients with diabetes mellitus and hypertension in the modeling set was 0.893 [95%CI (0.842, 0.945) ] , and the Brier score was 0.116; and the AUC of the nomogram model for predicting CF in elderly inpatients with diabetes mellitus and hypertension in the validation set was 0.836 [95%CI (0.741, 0.930) ] , and the Brier score was 0.146. The results of Hosmer-Lemeshow goodness-of-fit test showed that the nomogram model fit well ( χ 2 =5.97, P=0.65) . The results of calibration curve analysis showed that the incidence of CF in elderly patients with diabetes mellitus and hypertension in the modeling set predicted by the nomogram model was consistent with the actual incidence. Conclusion Age, mental activity, depression, self-care ability and fall risk are independent influencing factors of CF in elderly inpatients with diabetes mellitus and hypertension. The nomogram model constructed based on the above influencing factors has a good discrimination and calibration for predicting CF in elderly inpatients with diabetes mellitus and hypertension.

参考文献：