FUNCTIONAL NETWORKS THAT UNDERLIE EMOTION PROCESSING, DEPRESSION, AND ANXIETY: AN EEG APPLICATION OF CONNECTOME PREDICTIVE MODELING

Ashley Murray

Large scale distributed networks (i.e., functional connectivity) facilitate complex behaviors unique to humans. Dissonance within networks is present in mental health disorders, such as depression and anxiety, which also exhibit impairment in processing facial emotions. Brain activity while not overtly performing a task (i.e., at rest) has also provided evidence that there are stable, recurring patterns of brain activity that serve as the basis for functional activity, which allows for faster and more efficient responses to environmental changes. Connectome predictive modeling (CPM) is a useful methodology for identifying networks associated with individual behavioral outcomes. This study used electroencephalography (EEG) to capture brain activity in an eyes-closed resting state and while discriminating between four negative emotions, one of which served as the target per block. Anxiety, depression, and social anxiety symptom severity were captured. The aims of this study were to identify whether an emotion network could be found using CPM, identify whether that network was active at rest and could predict performance on the emotion task, and whether the task-active or resting state networks related to levels of anxiety, depression, and social anxiety. CPM was carried out on the EEG data, using behavioral accuracy measures of d’, target, and nontarget. Four positive networks were found to predict performance: delta d’, delta nontarget, beta d’, and beta target, indicating that greater connectivity in the delta and beta bands meant better performance on nontarget and target emotions, respectively. Using these networks, a common network was identified, applied to resting-state brain activity, and used to predict performance. The delta d’ network was the sole network able to predict task-based performance (d’) from resting state brain activity. This meant that greater connectivity in the delta d’ network at rest was associated with better d’ performance on the emotion task. Finally, it was found that beta d’ and beta target networks were positively related to anxiety and depression levels. Delta networks may represent cognitive processes present during the task (e.g., decision making and concentration) but may also represent processing of different negative facial emotions. The beta networks were also found to relate to depression and anxiety; thus, they may represent increased sensitivity to highly arousing emotions, which is present across both depression and anxiety. Together, delta supports differentiation across emotions, and beta supports heightened sensitivity to valence.

FUNCTIONAL NETWORKS THAT UNDERLIE EMOTION PROCESSING, DEPRESSION, AND ANXIETY: AN EEG APPLICATION OF CONNECTOME PREDICTIVE MODELING

Files and links (1)

Abstract

Metrics

Details