Objective:To investigate the risk factors of post-stroke cognitive impairment (PSCI) in patients with acute ischemic stroke, to establish a nomogram predictive model to help clinicians predict and intervene in the people who are prone to PSCI in advance, and to improve the recognition, intervention and prevention of the disease at an early stage, so as to provide a new way of thinking for the diagnosis and treatment of PSCI.Methods:Totally 330 patients with acute ischemic stroke hospitalized in the Department of Neurology, the First Bethune Hospital of Jilin University from January 2021 to June 2023 were collected. Their general clinical data, laboratory examination, imaging examination, and neuropsychological assessment data were collected. Neuropsychological scales assessment was completed within 7 days of the onset of acute ischemic stroke as a baseline value. The patients were followed up with neuropsychological scales assessment 6 months after the onset of stroke, and according to the results of the Montreal Cognitive Assessment (MoCA) scale assessment 6 months later, the patients were divided into PSCI group (143 patients) and post-stroke non-cognitive impairment (PSNCI) group (147 patients) (40 patients were removed from the study after 6 months, and a total of 290 patients were finally included in the study). Comparisons of general clinical information between the PSCI and PSNCI groups were first performed using statistical methods; then more influential predictors were selected using least absolute shrinkage and selection operator (LASSO) regression method and included in multifactor Logistic regression analyses to create a nomogram predictive model. Internal validation was performed by repeating the sampling 1 000 times using the bootstrap method; receiver operating characteristic (ROC) curve and area under the curve (AUC) were plotted to analyze the discrimination of the predictive model; the accuracy of the model was assessed using calibration curves; and a decision curve analysis (DCA) diagram was plotted to assess the clinical utility of the model.Results:Age, education level, critical area cerebral infarction, low-density lipoprotein-cholesterol (LDL-C), cerebral white matter hyperintensity (WMH), and cerebral atrophy were selected as the predictors of the nomogram predictive model by LASSO regression, and the results of multifactor Logistic regression analysis showed that these predictors were independent risk factors for PSCI in patients with acute ischemic stroke; the risk predictive model established was validated, and the results showed that the AUC of the present predictive model was 0.890, and the AUC of the internally validated predictive model was 0.940, suggesting that the model had a good degree of differentiation; the good fit between the calibration curve and the actual prediction results indicated that the model had good accuracy; the DCA results showed that the model can be well applied in clinical practice.Conclusion:The nomogram predictive model consisting of age, education level, critical area cerebral infarction, LDL-C, WMH, and cerebral atrophy has good differentiation, accuracy, and clinical utility, and can be used in practical clinical practice, which can help clinicians screen patients who are prone to PSCI, and intervene in a timely manner to achieve better clinical outcomes.

Objective:To investigate the risk factors of post-stroke cognitive impairment (PSCI) in patients with acute ischemic stroke, to establish a nomogram predictive model to help clinicians predict and intervene in the people who are prone to PSCI in advance, and to improve the recognition, intervention and prevention of the disease at an early stage, so as to provide a new way of thinking for the diagnosis and treatment of PSCI.Methods:Totally 330 patients with acute ischemic stroke hospitalized in the Department of Neurology, the First Bethune Hospital of Jilin University from January 2021 to June 2023 were collected. Their general clinical data, laboratory examination, imaging examination, and neuropsychological assessment data were collected. Neuropsychological scales assessment was completed within 7 days of the onset of acute ischemic stroke as a baseline value. The patients were followed up with neuropsychological scales assessment 6 months after the onset of stroke, and according to the results of the Montreal Cognitive Assessment (MoCA) scale assessment 6 months later, the patients were divided into PSCI group (143 patients) and post-stroke non-cognitive impairment (PSNCI) group (147 patients) (40 patients were removed from the study after 6 months, and a total of 290 patients were finally included in the study). Comparisons of general clinical information between the PSCI and PSNCI groups were first performed using statistical methods; then more influential predictors were selected using least absolute shrinkage and selection operator (LASSO) regression method and included in multifactor Logistic regression analyses to create a nomogram predictive model. Internal validation was performed by repeating the sampling 1 000 times using the bootstrap method; receiver operating characteristic (ROC) curve and area under the curve (AUC) were plotted to analyze the discrimination of the predictive model; the accuracy of the model was assessed using calibration curves; and a decision curve analysis (DCA) diagram was plotted to assess the clinical utility of the model.Results:Age, education level, critical area cerebral infarction, low-density lipoprotein-cholesterol (LDL-C), cerebral white matter hyperintensity (WMH), and cerebral atrophy were selected as the predictors of the nomogram predictive model by LASSO regression, and the results of multifactor Logistic regression analysis showed that these predictors were independent risk factors for PSCI in patients with acute ischemic stroke; the risk predictive model established was validated, and the results showed that the AUC of the present predictive model was 0.890, and the AUC of the internally validated predictive model was 0.940, suggesting that the model had a good degree of differentiation; the good fit between the calibration curve and the actual prediction results indicated that the model had good accuracy; the DCA results showed that the model can be well applied in clinical practice.Conclusion:The nomogram predictive model consisting of age, education level, critical area cerebral infarction, LDL-C, WMH, and cerebral atrophy has good differentiation, accuracy, and clinical utility, and can be used in practical clinical practice, which can help clinicians screen patients who are prone to PSCI, and intervene in a timely manner to achieve better clinical outcomes.