Predictive models for active suicidal ideation in cognitive decline: identifying risk factors

PURPOSE Suicide rates among older adults with cognitive decline represent a critical public health concern. Despite the association between cognitive decline and suicidality, predictive models for active suicidal ideation in this population remain underexplored. METHODS We conducted a retrospective study at the University Psychiatric Clinic Ljubljana, Slovenia, analyzing data from 1,889 patients (1,106 females and 783 males) with cognitive decline admitted between January 2019 and April 2024. Data were extracted from electronic health records and included sociodemographic, cognitive, clinical, psychiatric, and functional variables. Univariate logistic regression was initially used to identify correlates of active suicidal ideation. Subsequently, multivariate predictive modeling was performed using both Logistic Regression (LR) and Extreme Gradient Boosting (XGBoost). Recursive Feature Elimination (RFE) was applied to determine the key predictors, and SHAP values were utilized to enhance model interpretability. RESULTS Depressive symptoms, Mini-Mental State Examination score, duration of cognitive decline, past suicide attempts, antidementia medication use, and living arrangement emerged as key predictors. Both LR and XGBoost models demonstrated robust performance (ROC AUC: 0.81 and 0.85; PR AUC: 0.55 and 0.52, respectively). CONCLUSION Multivariate predictive models offer improved risk stratification for ASI, emphasizing the need for targeted interventions among individuals with cognitive decline.