Abstract: Objective To construct and validate of a 3-year recurrent stroke prediction model for elderly patients with acute ischemic stroke (AIS) involving the anterior circulation. Methods A total of 281 patients with anterior circulation AIS admitted to the Second Affiliated Hospital of Guangdong Medical University from January 2021 to June 2022 were retrospectively included as the research subjects. Based on the 36-month follow-up data, the patients were divided into the recurrence group and the non-recurrence group. Univariate and multivariate logistic regression analysis were used to investigate the independent risk factors for recurrent stroke in elderly patients within 3 years, and a nomogram prediction model was constructed. Subsequently, internal validation was conducted through calibration curves, area under the curve (AUC), and decision curve analysis (DCA). Results Among 281 patients, 93 (33.09%) elderly patients with anterior circulation AIS experienced recurrent stroke within 3 years. The results of multivariate logistic regression analysis showed that the time from onset to admission, modified Barthel index(MBI), high-sensitivity C-reactive protein and malondialdehyde (MDA) were independent risk factors for the risk of recurrent stroke within 3 years in patients with anterior circulation AIS (P<0.05), while end-distolic velocity (Vd) and superoxide dismutase (SOD) were independent protective factors (P<0.05). The internal performance evaluation results of the model showed that the Brier score of this model was 0.0146, the calibration slope was 1.0073, and the calibration intercept was -0.0024. The receiver operating characteristic curve indicated that the AUC value of this model for the risk of recurrent stroke within 3 years in elderly patients with anterior circulation AIS was 0.978 (95%CI:0.956-0.983). After 1000 internal validations by Bootstrap, the corrected AUC of this nomogram model was 0.942 (95%CI:0.921-0.965). DCA further confirmed that the model demonstrated good clinical net benefits over a wide range of threshold probabilities. Conclusions The nomogram model based on the time from onset to admission, MBI scores, MDA, hs-CRP, Vd, and SOD accurately assesses the risk of recurrent stroke within 3 years in elderly patients with anterior ciculation AIS, demonstrating good predictive efficacy and clinical utility.

Key words: acute ischemic stroke, recurrent stroke, clinical parameter, hemodynamic parameter, nomogram