The Incidence of Frailty and Construct a Risk Prediction Nomogram Model of Frailty in Hemodialysis Patients
نویسندگان
1 Department of Nephrology and Rheumatology，Xinan International Hospital Affiliated to Jiaxing University, Jiaxing 314000, Zhejiang Province, China
2 Jiaxing Zhejiang Health Examination Outpatient Department, Jiaxing 314000, Zhejiang Province, China
3 Department of Nephrology and Rheumatology，Xinan International Hospital Affiliated to Jiaxing University, Jiaxing 314000, Zhejiang Province, China
4 Department of Nephrology and Rheumatology，Xinan International Hospital Affiliated to Jiaxing University,Zhejiang, China
5 Department of Nephrology and Rheumatology，Xinan International Hospital Affiliated to Jiaxing University, Jiaxing 314000, Zhejiang Province, China
doi
10.22034/ircmj.2025.501188.1813چکیده
Background and Objectives: Frailty is common in hemodialysis patients, affecting quality of life and clinical outcomes. Early identification can guide interventions, yet reliable predictive models for frailty risk remain limited. To determine the incidence of frailty in hemodialysis patients and develop a nomogram model for predicting frailty risk, aiding early identification and intervention. Methods: A cross-sectional study with 400 hemodialysis patients collected clinical data, laboratory indicators, social support, depression, anxiety, and sleep quality. Frailty was assessed using the Fried Frailty Phenotype scale. Univariate and logistic regression analyses identified risk factors for frailty, which informed a predictive nomogram model. Model performance was evaluated via ROC curve, calibration curve, and decision curve analysis (DCA). Results: Frailty incidence was 44.00%. Significant differences were observed between frailty and non-frailty groups in age, sex, education, diabetes, cardiovascular disease, serum albumin, social support, depression, anxiety, and sleep quality (P < 0.05). Logistic regression revealed that older age (OR=1.086, P=0.002), cardiovascular disease (OR=2.741, P=0.030), low serum albumin (OR=5.569, P<0.001), depression (OR=26.578, P<0.001), anxiety (OR=13.471, P<0.001), and poor sleep quality (OR=2.992, P=0.014) were significant risk factors. The nomogram demonstrated high predictive accuracy (AUC: 0.974 training, 0.980 validation). Calibration and DCA curves confirmed its reliability and clinical utility. Conclusion: Age, cardiovascular disease, serum albumin, depression, anxiety, and sleep quality were key risk factors. The nomogram offers a practical and accurate tool for early frailty identification and intervention.
