Network Structure and Central Symptoms of Depression in Older Korean Adults: A Community-Based Analysis

Abstract

Objective This study aimed to investigate the network structure of Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV)-defined depressive symptoms in a large, community-dwelling sample of older Korean adults, to better understand symptom interrelationships and their potential clinical implications in late-life depression.Methods Depressive symptoms were assessed using the Korean version of the Composite International Diagnostic Interview 2.1, which includes both DSM-IV and 10th revision of the International Classification of Diseases symptom assessments. A total of 16 DSM-IV-defined symptom variables were derived based on diagnostic criteria and included in the network analysis (n=2,007). Network analysis was conducted using phi coefficients, community detection, and centrality and predictability. Sex-specific network structures were also compared.Results Two distinct symptom clusters were identified. Suicide attempt emerged as one of the symptoms with relatively high betweenness centrality and predictability, suggesting a potentially distinctive position within the network, despite its relatively peripheral placement in the network visualization. Low energy, reduced self-confidence, and concentration difficulties also showed high degree and closeness centrality. The network among female participants was denser, particularly around suicide-related symptoms, and older female participants showed higher symptom prevalence and a slightly higher mean age.Conclusion These findings suggest that suicide-related symptoms may play a distinctive and potentially bridging role within the depressive symptom network of older Korean adults, rather than being uniformly central across all metrics. These results highlight the importance of considering symptom heterogeneity, sex differences, and cultural context when developing targeted prevention and intervention strategies in geriatric mental health.

The authors' abstract, as published at the source. Psychiatry Investigation, 2026 · DOI ↗