[Prevalence of frailty and associated factors among kidney transplant recipients].

Objective: To investigate the prevalence of frailty among kidney transplant recipients and to analyze the influential factors of frailty after kidney transplantation. Methods: We retrospectively included 201 kidney transplant recipients who were followed up in the Department of Urology, Beijing Chao-yang Hospital, Capital Medical University from November 2020 to May 2022. We investigated the prevalence of frailty based on the Fried Frailty Scale (including unexpected shrinking, slow walking speed, poor grip strength, low physical activity, and exhaustion). Then the logistic regression model and CART decision tree model were established separately to explore the influential factors of frailty after kidney transplantation. Results: Frail kidney transplant recipients accounted for 25.9% (n=52) of all participants. The age [M (Q1, Q3)] of the frailty group was higher than that of the non-frailty group, and the median ages of the two groups were 57(49, 62) and 46(38, 56) (P<0.001); the males accounted for 51.9% (n=27) and 62.4% (n=93), respectively. There was no significant difference in gender composition (P=0.244). Among the five components of Fried Frailty Scale, the incidence of unexpected shrinking was the lowest (19.4%, 39/201). In the frailty group, the frailty combination with the highest incidence was slow walking speed+low physical activity+exhaustion, which was 19.2% (10/52). The logistic regression model showed that advanced age (OR=1.062, 95%CI: 1.005-1.123), history of acute rejection (OR=16.776, 95%CI: 2.288-123.028), increased neutrophil/lymphocyte ratio (NLR) (OR=2.096, 95%CI: 1.158-3.792), and comorbidity (OR=10.600, 95%CI: 1.828-61.482) were risk factors for frailty among kidney transplant recipients, and high serum albumin level (OR=0.623, 95%CI: 0.488-0.795) was a protective factor. The CART decision tree grew in three layers with four terminal nodes, and three explanatory variables were screened out: serum albumin, NLR, and age. The accuracy, sensitivity, and specificity of the logistic regression model were 87.1% (95%CI: 82.5%-91.7%), 69.2% (95%CI: 54.7%-80.9%), and 93.3% (95%CI: 87.7%-96.6%), respectively. The area under the ROC curve (AUC) of the logistic regression model was 0.951 (95%CI: 0.923-0.978). The accuracy, sensitivity, and specificity of the CART decision tree model were 91.0% (95%CI: 87.0%-95.0%), 82.7% (95%CI: 69.2%-91.3%), and 94.0% (95%CI: 88.5%-97.0%), respectively. The AUC of the CART decision tree model was 0.883 (95%CI: 0.819-0.948). Conclusions: The prevalence of frailty among kidney transplant recipients in this study is 25.9%. Advanced age, history of acute rejection, low serum albumin level, increased NLR, and comorbidity are likely to be associated with the long-term frailty among kidney transplant recipients.目的： 调查肾移植受者衰弱的患病率并分析肾移植受者术后衰弱的相关因素。 方法： 回顾性纳入2020年11月至2022年5月在首都医科大学附属北京朝阳医院泌尿外科门诊随访的201例肾移植受者资料。以Fried 衰弱表型（包括意外体重下降、步行速度慢、握力差、体力活动少、疲惫5个方面）为诊断标准，调查肾移植受者衰弱患病情况。分别建立logistic回归模型和CART决策树模型，分析肾移植术后衰弱的影响因素。 结果： 肾移植受者衰弱患病率为25.9%（52例）。衰弱组年龄［M（Q1，Q3）］高于非衰弱组，分别为57（49，62）和46（38，56）岁（P<0.001）；男性分别占51.9%（27例）、62.4%（93例），性别构成差异无统计学意义（P=0.244）。在Fried 衰弱表型的5个组分中，意外体重下降发生率最低，为19.4%（39/201）。衰弱组发生率最高的衰弱组合为步行速度慢+低体力活动+疲惫，为19.2%（10/52）。logistic回归模型显示，高龄（OR=1.062，95%CI：1.005~1.123）、急性排斥史（OR=16.776，95%CI：2.288~123.028）、高中性粒细胞/淋巴细胞比值（NLR）（OR=2.096，95%CI：1.158~3.792）和患共病（OR=10.600，95%CI：1.828~61.482）是肾移植受者衰弱的危险因素，高血清白蛋白水平（OR=0.623，95%CI：0.488~0.795）是保护因素。CART决策树生长3层，共有4个终末节点，筛选出3个解释变量：血清白蛋白、NLR和年龄。logistic 回归模型的准确度为87.1%（95%CI：82.5%~91.7%），灵敏度为69.2%（95%CI：54.7%~80.9%），特异度为93.3%（95%CI：87.7%~96.6%），受试者工作特征（ROC）曲线下面积（AUC）为0.951（95%CI：0.923~0.978）；决策树模型的准确度为91.0%（95%CI：87.0%~95.0%），灵敏度为82.7%（95%CI：69.2%~91.3%），特异度为94.0%（95%CI：88.5%~97.0%），AUC为0.883（95%CI：0.819~0.948）。 结论： 本研究中肾移植受者衰弱的患病率为25.9%；高龄、急性排斥史、低血清白蛋白水平、NLR升高和患共病可能与肾移植受者术后长期衰弱相关。.