Abstract:

Objective To investigate the factors associated with cognitive impairment (CI) in patients with stage 5 chronic kidney disease undergoing hemodialysis and to develop a risk prediction model. Methods A total of 184 patients with stage 5 chronic kidney disease undergoing hemodialysis, who were treated in the Department of Nephrology, Rheumatology, and Immunology of the First People's Hospital of Pingdingshan City from November 2022 to November 2025, were enrolled and randomly assigned in a 6∶2∶2 ratio to a modeling set (training set: 110 cases; test set: 37 cases) and an internal validation set (37 cases). Based on CI screening results, patients were divided into the CI group and the no-CI group. Univariate analysis, LASSO regression, and multivariate logistic regression were used to identify factors associated with CI, and a prediction model was developed. The predictive performance of the model was further validated using the test set. Results Among the 184 patients, 60 developed CI, with an overall incidence of 32.61%. After LASSO regression screening of 26 variables, multivariate logistic regression identified age (OR=1.074), years of education (OR=0.808), duration of dialysis (OR=1.074), serum albumin (OR=0.856), and use of beta-blockers (OR=4.359) as independent factors associated with CI (all P<0.05). The area under the curve (AUC) of the prediction model was 0.922 in the training set (95%CI: 0.866-0.978), 0.943 in the validation set (95%CI: 0.814-0.993), and 0.933 in the test set (95%CI: 0.800-0.989). The mean AUC of the model was 0.909 (95%CI: 0.814-0.993) on 5-fold cross-validation. The calibration curve showed high agreement between predicted and observed probabilities (P>0.05). According to the decision curve analysis, when the threshold probability ranged from 0.080 to 0.320, using this model for decision-making provided additional clinical net benefit. Conclusion The incidence of CI is relatively high in patients with stage 5 chronic kidney disease undergoing hemodialysis. Older age, longer duration of dialysis, and beta-blocker use are independent risk factors, whereas longer years of education and higher serum albumin levels are protective factors. The developed prediction model demonstrates good discrimination, calibration, and clinical utility, and may serve as a tool for early identification of patients at high risk of CI.

Key words: uremia, hemodialysis, cognitive impairment, influencing factors, nomogram