Abstract:

Objective A Nomogram model of cognitive frailty was constructed and validated in hospitalized older adults，providing a reference for early screening，intervention and personalized management of cognitive frailty. Methods A convenience sampling approach was employed to recruit 322 elderly inpatients from a tertiary hospital in Beijing between October 2024 and February 2025 as study participants，and data were collected using the General Information Questionnaire，the Short Form-Mini-Nutritional Assessment，the Asens Insomnia Scale，the Activity of Daily Living Rating，the Self-Rating Anxiety Scale，the Geriatric Depression Scale-15，the Social Support Rating Scale，the Frailty Phenotype scale，the Subjective Cognitive Decline Questionnaire-9，the Mini-Mental State Examination，and the Clinical Dementia Rating. Lasso-Logistic regression was used to screen the variables，R software was used to draw the nomogram model；Bootstrap method was used for internal validation. Results Lasso-Logistic regression screened 8 predictors of age，depression，anxiety，support utilization，nutritional status，literacy，physical activity，and chronic pain，with an area under the subject operating characteristic curve of 0.830（95%CI：0.787-0.873），a sensitivity of 0.764，a specificity of 0.730，an accuracy of 0.748，and a calibrated curve，Brier score，and Hosmer-Lemeshow test（P=0.774） all showed that the model fit was good. Conclusion The Lasso-Logistic regression-based nomogram model of cognitive frailty in hospitalized older adults has good predictive performance and clinical utility，and can be used as a reference for early identification and intervention of cognitive decline in hospitalized older adults.

Key words: Aged, Cognitive Frailty, Nomogram, Geriatric Nursing

摘要：

目的 构建并验证住院老年患者认知衰弱列线图模型，为认知衰弱的早期筛查、干预和个性化管理提供参考。方法 采用便利抽样法，于2024年10月—2025年2月，选取北京市某三级甲等医院的322例住院老年患者为调查对象，使用一般资料调查表、微型营养评定简表、阿森斯失眠量表、日常生活能力量表、焦虑自评量表、简版老年抑郁量表、社会支持评定量表、Fried衰弱表型量表、中文版主观认知下降问卷、简易精神状态检查量表、临床痴呆评估量表收集资料。使用Lasso-Logistic回归筛选变量，采用R软件绘制列线图模型，采用Bootstrap法进行内部验证。结果 Lasso-Logistic回归筛选出年龄、抑郁、焦虑、支持利用度、营养状况、文化程度、体育锻炼及慢性疼痛8个预测因素，预测模型的受试者操作特征曲线下面积为0.830（95%CI：0.787~0.873），灵敏度为0.764，特异度为0.730，准确率为0.748，校准曲线、Brier得分以及Hosmer-Lemeshow检验结果（P=0.774）均显示模型拟合优度良好。结论 基于Lasso-Logistic回归构建的住院老年患者认知衰弱列线图模型预测性能和临床实用性良好，可为住院老年患者认知衰弱的早期识别和干预提供参考。

关键词: 老年人, 认知衰弱, 列线图, 老年护理