Mapping Depression Correlated Amygdala Communication.
Authors
Affiliations (5)
Affiliations (5)
- Department of Psychology, Renmin University of China, Beijing, 100872, China.
- Research Center for Social Psychological Science and Engineering, Renmin University of China, Beijing, 100872, China.
- Social Psychology Section, Behavioural and Human Movement Sciences, Vrije Universiteit Amsterdam, Amsterdam, 1081BT, The Netherlands.
- Department of Psychology, Renmin University of China, Beijing, 100872, China. [email protected].
- Research Center for Social Psychological Science and Engineering, Renmin University of China, Beijing, 100872, China. [email protected].
Abstract
The amygdala is a central hub for emotional processing, whose interactions with large-scale brain networks have been implicated in the pathophysiology of depression. Investigating how amygdala-centered connectivity relates to depression in a community-based sample, rather than focusing exclusively on patients with major depressive disorder, may provide broader insights into depression vulnerability. To this end, functional connectivity (via cross-correlation) and effective connectivity (via Granger causality) were computed using resting-state functional magnetic resonance imaging (fMRI) data from 161 adults obtained from the Nathan Kline Institute (NKI)/Rockland community sample. Subsequently, a behavior-guided windowed association analysis identified connectivity indicators correlated with depression severity (Beck Depression Inventory scores), whose contributions to depression prediction were subsequently ranked using machine learning based on a support vector machine. Specifically, higher Beck Depression Inventory (BDI) scores were associated with increased amygdala functional connectivity with the salience/ventral attention system and the dmPFC component of the default mode network, together with decreased connectivity involving dorsal attention, sensorimotor, and visual regions. Granger causality analyses further revealed predominantly reduced directional influences both to and from the amygdala, particularly between the amygdala and distributed default mode, salience/ventral attention, visual, and subcortical regions, suggesting weakened large-scale information exchange. Machine learning analyses identified amygdala-centered connectivity features, especially those involving the default mode and salience-related systems, as among the most informative predictors of depression severity. Overall, these findings delineate an amygdala-centered multi-system network pathology underlying depressive vulnerability, providing quantifiable neural signatures for early detection and potential targets for intervention.