Insights from EEG Microstate Analysis on the Pathophysiology of Depression and Mechanisms of Seizure Therapy

Atluri, Sravya 1, 3 ;  Wong, Willy 1 ;  Blumberger, Daniel  2, 3 ;  Daskalakis, Zafiris J 2, 3 ;  Farzan, Faranak 2, 3

1. Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, ON, Canada; 2. Department of Psychiatry, University of Toronto, Toronto, ON, Canada; 3. Temerty Centre for Therapeutic Brain Intervention, Centre for Addiction and Mental Health, Toronto, ON, Canada

Background: Converging neuroimaging evidence suggests that specific patterns of brain network dysfunction at rest may contribute to the core symptoms associated with Major Depressive Disorder (MDD). The efficacy of seizure therapy to treat MDD is hypothesized to be associated with the resetting of abnormal neural connectivity in MDD. In this study, a distinct methodology known as electroencephalogram (EEG) microstate analysis, was used to evaluate functional brain network dynamics in patients with depression and after they received seizure therapy.

Methods: Resting-state 60-channel EEG data was obtained from 55 healthy subjects and 75 MDD patients. Of the 75 patients, 22 received Electroconvulsive Therapy (ECT) and 24 received Magnetic Seizure Therapy (MST). Resting-state EEG was recorded again at the completion of treatment. Three features were derived for each microstate class (A, B, C, D): frequency, duration and coverage time. The Hamilton Depression Scale and Montreal Cognitive Assessment test was used to clinically assess patients prior to and following treatment. 

Results: Seizure therapy was shown to have a significant impact on microstate characteristics. Results indicate the normalization of a hyperactive Map D (previously linked with the frontoparietal network) through a decrease in the coverage of Map D after treatment. This normalization was specific to responders of ECT. Furthermore, responders of seizure therapy (ECT and MST) also showed an increase in the duration of Map A. This increase in duration of Map A was shown to be correlated with improvement in depressive symptoms. Finally, a decrease in the frequency of Map B was shown to be specific to ECT responders only. This decrease in frequency of Map B was correlated with improvement in cognition.

Conclusions: Provided that microstates can be recorded with portable and inexpensive EEG technology, these findings have important practical implications in translational research for patients with neuropsychiatric disorders.