A Clinical-Metabolic Prediction Model for Suicidal Behaviors Risk Stratification in First-Admission Major Depressive Disorder: A Cross-Sectional Analysis

Neuropsychiatr Dis Treat. 2025 Jul 5;21:1339-1348. doi: 10.2147/NDT.S519787. eCollection 2025.

ABSTRACT

BACKGROUND: The clinical characteristics and biomarkers of suicidal behaviors (SB) in first-hospitalized patients with Major Depressive Disorder (MDD) remain poorly understood. This study aimed to investigate the prevalence, clinical correlates, and metabolic disturbances of SB in first-admission MDD patients in China, integrating psychosocial and biological markers to establish a predictive model.

METHODS: A cross-sectional analysis was conducted on 981 first-admission MDD inpatients. Sociodemographic data, clinical symptom severity (17-item Hamilton Depression Rating Scale [HAMD-17], 14-item Hamilton Anxiety Rating Scale [HAMA-14], PANSS positive subscale [PSS], Clinical Global Impression-Severity Index [CGI-SI]), and metabolic parameters (lipid profile, fasting glucose, thyroid function) were collected. SB was assessed using the Columbia-Suicide Severity Rating Scale (C-SSRS). Binary logistic regression and ROC analysis identified correlates and model performance.

RESULTS: The prevalence of SB was 13.46% (132/981). SB patients exhibited significantly higher psychotic symptoms, anxiety severity, and illness severity, along with elevated waist circumference (WC), diastolic blood pressure (DBP), total cholesterol (TC), and thyroid-stimulating hormone (TSH). Logistic regression identified HAMA (OR=1.72, 95% CI=1.25-2.37), PSS (OR=1.58, 95% CI=1.13-2.21), CGI-SI (OR=1.45, 95% CI=1.08-1.95), and TC (OR=1.32, 95% CI=1.04-1.68) as factors independently associated with SB (all p<0.05). The combined model of PSS, HAMA, and CGI-SI demonstrated strong discriminative accuracy (AUC=0.87, 95% CI: 0.83-0.91). Linear regression further linked HAMA scores to SB severity (β=0.21, p=0.029).

CONCLUSION: SB in first-hospitalized MDD patients correlates with anxiety symptoms, psychotic features, and metabolic dysregulation. A multidimensional model integrating clinical and metabolic indicators is associated with high-risk individuals, supporting targeted prevention strategies.

PMID:40636800 | PMC:PMC12240159 | DOI:10.2147/NDT.S519787