Publication Details and Research Team
A new study published in the Journal of Affective Disorders examines resting-state functional connectivity alterations associated with suicide attempt history and childhood trauma among individuals with major depressive disorder. The research, appearing in Volume 412 on November 1, 2026, as article 122111, was led by Minjee Jung alongside Jihoon Park, Youbin Kang, Daun Shin, June Kang, JeYoung Jung, Marcus Kaiser, Dorothee P. Auer, Byung-Joo Ham, and Kyu-Man Han. Readers can access the full publication at https://www.sciencedirect.com/science/article/abs/pii/S0165032726009638.
Context of Major Depressive Disorder and Suicidality
Major depressive disorder affects millions worldwide and carries elevated risks for suicidal behavior. A prior suicide attempt stands as the strongest predictor of future suicide mortality in this population. Researchers have long sought biological markers that distinguish those with attempt histories from others diagnosed with the disorder. This latest work contributes to that effort by mapping large-scale brain network differences using resting-state functional magnetic resonance imaging.
The study recruited 204 adults aged 19 to 59 years, including 123 individuals with major depressive disorder and 81 healthy controls. Within the patient group, 61 had a lifetime history of at least one suicide attempt, designated the suicidal depression group, while 62 had no such history. All participants underwent scanning at Korea University Anam Hospital in Seoul between 2015 and 2021.
Neuroimaging Approach and Analytical Methods
Investigators constructed functional connectivity matrices from resting-state scans using the Schaefer 400-node atlas. They applied network-based statistics to detect distributed subnetworks showing group differences without limiting analysis to predefined regions. Complementary graph-theoretical measures assessed network strength, efficiency, and centrality within identified subnetworks.
Functional connectivity refers to statistical associations between blood-oxygen-level-dependent signals across brain regions during rest. Network-based statistics identifies clusters of connections that differ significantly between groups while controlling for multiple comparisons. Graph theory then quantifies how information flows through those connections, revealing whether networks appear more or less integrated or centralized.
Primary Connectivity Findings
Network-based statistics revealed a significant subnetwork with lower functional connectivity in the suicidal depression group compared with the non-suicidal depression group. This subnetwork spanned the visual, somatomotor, dorsal attention, frontoparietal, and default mode networks. The pattern reached family-wise error corrected significance at p = 0.038.
Relative to healthy controls, the suicidal depression group displayed both hypo- and hyperconnected subnetworks, whereas the non-suicidal depression group showed no statistically significant connectivity differences. Within the suicidal depression versus non-suicidal depression subnetwork, graph-theoretical analysis indicated reduced strength, global efficiency, and betweenness centrality.
Associations with Clinical Measures
Across the major depressive disorder sample, mean subnetwork connectivity, somatomotor-visual connectivity, and default mode-default mode connectivity correlated negatively with suicidal ideation severity. In the suicidal depression group specifically, somatomotor-somatomotor connectivity showed an inverse relationship with scores on the childhood trauma questionnaire subscale for sexual abuse.
These correlations suggest that the observed hypoconnectivity may track with both current ideation intensity and particular forms of early adversity. The somatomotor network involvement points to potential disruptions in sensory-motor integration that could influence emotional regulation or impulse control pathways.
Photo by Bhautik Patel on Unsplash
Interpretation of Distributed Network Disruption
The findings point to hypoconnectivity across sensory, attentional, and self-referential systems as a potential connectome-level signature of suicide vulnerability within major depressive disorder. Rather than isolated focal abnormalities, the pattern involves coordinated failures of integration among multiple canonical networks. This distributed architecture aligns with emerging views that suicidality reflects broad alterations in how the brain processes sensory input, directs attention, and maintains self-referential thought.
Childhood sexual abuse emerged as a specific correlate within the suicidal depression subgroup, consistent with evidence that distinct trauma subtypes can imprint differently on sensorimotor and cognitive networks. The study thereby links early experience to later network organization in a clinically relevant population.
Implications for Mental Health Research and Practice
These results advance understanding of biological heterogeneity within major depressive disorder. Clinicians and researchers may eventually use such connectivity profiles to stratify patients by risk or to monitor treatment response. The emphasis on connectome-wide approaches complements narrower seed-based or region-of-interest methods that have dominated earlier work.
Future studies could test whether the identified subnetwork predicts prospective suicide attempts or responds to targeted interventions such as neuromodulation or trauma-focused therapies. Longitudinal designs would clarify whether the observed patterns represent state or trait features.
Broader Neuroimaging Landscape
This publication builds on prior connectome studies in mood disorders and suicidality. It extends earlier multivariate pattern analyses of the same cohort by shifting focus to subnetwork topology. Complementary work in transdiagnostic samples has similarly highlighted hypoconnectivity involving visual, somatomotor, and salience systems, suggesting partial convergence across diagnostic boundaries.
Funding support came from the National Research Foundation of Korea and Korea University, underscoring institutional investment in psychiatric neuroimaging within the region. The absence of reported conflicts of interest strengthens confidence in the reported outcomes.
Considerations for Academic and Research Communities
Neuroimaging research of this scope requires interdisciplinary teams spanning psychiatry, neuroscience, engineering, and statistics. Early-career researchers interested in functional connectivity methods or suicide risk biomarkers may find expanding opportunities in collaborative projects that integrate large-scale datasets with graph-analytic techniques.
Institutions seeking to strengthen mental health research portfolios could prioritize training in network-based statistics and open-source connectomics pipelines. Such capacity building supports both basic discovery and translational applications aimed at reducing suicide mortality.
Future Directions and Open Questions
Key questions remain about generalizability beyond the Korean clinical sample and the influence of medication status or illness duration on the observed patterns. Replication in independent cohorts, including those with diverse cultural and demographic backgrounds, will be essential.
Integration with other modalities such as structural imaging, electroencephalography, or peripheral biomarkers could yield multimodal signatures with improved predictive utility. Computational modeling of the identified subnetworks may further illuminate mechanisms linking connectivity changes to behavior.
Photo by Vitaly Gariev on Unsplash
Conclusion
The study by Jung, Park, Kang, and colleagues provides a detailed map of resting-state functional connectivity differences that distinguish major depressive disorder patients with versus without suicide attempt histories. By linking these network alterations to suicidal ideation and childhood sexual abuse, the work offers a foundation for refined risk models and targeted research agendas. The full article is available at the provided ScienceDirect link for interested investigators and clinicians.
