Abstract
Alexithymia, characterized by difficulty in expressing and recognizing emotions, is prevalent among young and middle-aged stroke survivors and can significantly impact rehabilitation outcomes. This study aims to develop and validate a dynamic nomogram to predict the risk of alexithymia in this population. This cross-sectional study was conducted from November 2022 to August 2023 at two tertiary hospitals in Jinzhou City and Cangzhou City, enrolling 319 patients. Predictive factors for alexithymia, such as Activities of Daily Living (ADL) scores, social support levels, lesion location, educational background, and National Institutes of Health Stroke Scale (NIHSS) scores, were identified through univariate and multivariate analyses. These factors were integrated into a web-based dynamic nomogram. The model's accuracy was evaluated using Receiver Operating Characteristic (ROC) curves and 1000 bootstrap resamples. In the training cohort, 47.8% of patients were diagnosed with alexithymia. The nomogram demonstrated excellent fit and reliability, with an Area Under the Curve (AUC) of 0.837 (95% CI: 0.787-0.889) in the training cohort and 0.847 (95% CI: 0.767-0.928) in the validation cohort, enabling reliable early detection of alexithymia. The dynamic nomogram provides healthcare professionals with an important tool for early detection and management of alexithymia in young and middle-aged stroke survivors. While the model shows high predictive accuracy, its applicability may be limited to similar clinical settings. Future studies should evaluate its utility across diverse healthcare systems. This tool has the potential to significantly improve rehabilitation outcomes by supporting personalized therapeutic strategies and interventions.