作者：王彦，刘媛

单位：

225001江苏省扬州市，扬州大学附属医院老年全科医学科 通信作者：刘媛，E-mail：liuyyz1982@163.com

Units：

Geriatric Medicine, Affiliated Hospital of Yangzhou University, Yangzhou 225001, China Corresponding author: LIU Yuan, E-mail: liuyyz1982@163.com

关键词：

高血压; 老年人; 住院; 认知衰弱; 影响因素分析; 列线图模型;

Keywords：

Hypertension; Aged; Hospitalization; Cognitive frailty; Root cause analysis; Nomogram model

CLC：

DOI：

10.12114/j.issn.1008-5971.2022.00.167

Funds：

摘要：

目的 分析住院老年高血压患者认知衰弱的影响因素,并构建其列线图模型。方法 选取2020年1月至2022年1月扬州大学附属医院收治住院的老年高血压患者221例,按照7∶3比例将其随机分为建模组(n=155)和验证组(n=66)。建模组根据患者是否合并认知衰弱分为衰弱亚组(n=41)及非衰弱亚组(n=114)。收集患者的一般资料,使用Fried衰弱表型量表评估患者认识衰弱情况,使用简易智力状态检查量表(MMSE)评估患者认知功能,使用抑郁自评量表(SDS)及焦虑自评量表(SAS)评估患者抑郁、焦虑情况,使用微型营养评估(MNA)量表评估患者营养状态,使用阿森斯失眠评估量表(AIS)评估患者睡眠情况。采用多因素Logistic回归分析探讨住院老年高血压患者发生认知衰弱影响因素并构建其列线图模型,采用H-L拟合优度检验评估列线图模型在模型组及验证组的预测有效性;采用Bootstrap法重复抽样1 000次进行验证,计算一致性指数(CI);绘制校正曲线和ROC曲线以分析列线图模型在模型组及验证组的预测概率与实际概率的一致性和区分度。结果 221例患者中,发生认知衰弱58例,发生率为26.2%。衰弱亚组与无衰弱亚组年龄、合并基础疾病者所占比例、高血压分级、合并营养不良者所占比例、合并失眠者所占比例比较,差异有统计学意义(P<0.05);多因素Logistic回归分析结果显示,年龄[OR=8.283,95%CI(2.809,24.425)]、高血压分级[OR=5.017,95%CI(1.448,17.385)]、合并营养不良[OR=7.035,95%CI(2.451,20.193)]、合并失眠[OR=5.151,95%CI(1.830,14.499)]是住院老年高血压患者发生认知衰弱的影响因素(P<0.05)。构建列线图模型,年龄≥70岁为100分,高血压分级为3级为74分,合并营养不良为92分,合并失眠为78分。H-L拟合优度检验结果显示,列线图模型的拟合效果良好,建模组χ2=6.423,P=0.431;验证组χ2=6.174,P=0.352。建模组CI为0.886[95%CI(0.812,0.968)],验证组CI为0.781[95%CI(0.742,0.934)]。校正曲线分析结果显示,列线图模型预测建模组、验证组住院老年高血压患者发生认知衰弱风险的预测概率与实际概率基本吻合。ROC曲线分析结果显示,列线图模型预测住院老年高血压患者发生认知衰弱建模组的AUC为0.883[95%CI(0.822,0.929)],灵敏度、特异度分别为95.12%、72.81%;验证组的AUC为0.770[95%CI(0.696,0.834)],灵敏度、特异度分别为82.93%、68.42%。结论 年龄≥70岁、高血压分级为3级、合并营养不良、合并失眠是住院老年高血压患者发生认知衰弱的危险因素,基于上述因素构建的列线图模型具有较高的有效性、一致性和区分度,可用于临床预测住院老年高血压患者认知衰弱的发生。

Abstract：

【Abstract】　Objective To analyze the influencing factors of cognitive frailty in hospitalized elderly hypertensivepatients, and to construct its nomogram model. Methods A total of 221 hospitalized elderly hypertensive patients admitted toAffiliated Hospital of Yangzhou University from January 2020 to January 2022 were selected and randomly divided into modelinggroup (n=155) and validation group (n=66) according to the ration of 7∶3, and the patients in modeling group were dividedinto frail subgroup (n=41) and non-frail subgroup (n=114) according to whether combined with cognitive frailty. The generalinformation of the patients were collected, Fried Frailty Phenotype Scale were used to assess cognitive frailty, Mini-mental StateExamination (MMSE) was used to assess cognitive function, Self-rating Depression Scale (SDS) and Self-rating Anxiety Scale(SAS) were used to assess depression and anxiety states, Mini Nutritional Assessment (MNA) Scale was used to assess nutritionalstatus, and Athens Insomnia Scale (AIS) was used to assess sleep. The influencing factors of cognitive frailty in hospitalized elderlyhypertensive patients were analyzed by multivariate Logistic regression analysis, and nomogram model was constructed. The H-Lgoodness-of-fit test was used to evaluate the validity of the nomogram model in modeling group and validation group, the C-index(CI) was calculated by repeated sampling 1 000 times by Bootstrap method, and the calibration curve and ROC curve were drawnto analyze the consistency and discrimination between predicted probability and actual probability of the nomogram model inmodeling group and validation group. Results Among the 221 patients, 58 patients had cognitive frailty, with an incidence of26.2%. There were significant differences in age, proportion of patients with basic diseases, hypertension grade, proportion ofpatients with malnutrition, and proportion of patients with insomnia between frail subgroup and non-frail subgroup (P <0.05) .Multivariate Logistic regression analysis showed that age [OR=8.283, 95%CI (2.809, 24.425) ] , hypertension grade [OR=5.017,95%CI (1.448, 17.385) ] , combined with malnutrition [OR=7.035, 95%CI (2.451, 20.193) ] and combined with insomnia[OR=5.151, 95%CI (1.830, 14.499) ] were the influencing factors of cognitive frailty in hospitalized elderly hypertensive patients(P<0.05) . The nomogram model was constructed, and the score of age ≥ 70 years was 100 points, hypertension of grade 3 was 74points, combined with malnutrition was 92 points, and combined with insomnia was 78 points. The H-L goodness of fit test showedthat the fitting effect of the nomogram model was good, modeling group χ2 =6.423, P=0.431; validation group χ2 =6.174, P=0.352.CI in the modeling group was 0.886 [95%CI (0.812, 0.968) ] , CI in the validation group was 0.781 [95%CI (0.742, 0.934) ] . Thecalibration curve analysis showed that predicted probability of the nomogram model for predicting cognitive frailty in hospitalizedelderly hypertensive patients in the modeling group and validation group were basically consistent with the actual probability. TheROC curve analysis showed that the AUC of the nomogram model for predicting the occurrence of cognitive frailty in hospitalizedelderly hypertensive patients in the modeling group was 0.883 [95%CI (0.822, 0.929) ] , and the sensitivity and specificity were95.12% and 72.81%, respectively; the AUC in the validation group was 0.770 [95%CI (0.696, 0.834) ] , and the sensitivityand specificity were 82.93% and 68.42%, respectively.Conclusion Age ≥ 70 years, hypertension of grade 3, combined withmalnutrition, and combined with insomnia are the risk factors of cognitive frailty in hospitalized elderly hypertensive patients,the constructed nomogram model based on the above factors has high validity, consistency and discrimination, and can be used toclinically predict the occurrence of cognitive frailty in hospitalized elderly hypertensive patients.

ReferenceList：