Development and validation of a nomogram to predict suicidal behavior in female patients with mood disorder

INTRODUCTION: This study aims to explore the risk factors associated with suicidal behavior and establish predictive models in female patients with mood disorders, specifically using a nomogram of the least absolute shrinkage and selection operator (LASSO) regression. METHODS: A cross-sectional survey was conducted among 396 female individuals diagnosed with mood disorders (F30-F39) according to the International Classification of Diseases and Related Health Problems 10th Revision (ICD-10). The study utilized the Chi-Squared Test, t-test, and the Wilcoxon Rank-Sum Test to assess differences in demographic information and clinical characteristics between the two groups. Logistic LASSO Regression Analyses were utilized to identify the risk factors associated with suicidal behavior. A nomogram was constructed to develop a prediction model. The accuracy of the prediction model was evaluated using a Receiver Operating Characteristic (ROC) curve. RESULT: The LASSO regression analysis showed that psychotic symptoms at first-episode (β = 0.27), social dysfunction (β = 1.82), and somatic disease (β = 1.03) increased the risk of suicidal behavior. Conversely, BMI (β = −0.03), age of onset (β = −0.02), polarity at onset (β = −1.21), and number of hospitalizations (β = −0.18) decreased the risk of suicidal behavior. The area under ROC curve (AUC) of the nomogram predicting SB was 0.778 (95%CI: 0.730–0.827, p < 0.001). CONCLUSION: The nomogram based on demographic and clinical characteristics can predict suicidal behavior risk in Chinese female patients with mood disorders.