An Electrophysiological Biomarker That May Predict Treatment Response to ECT.


Journal Article

OBJECTIVE: Electroconvulsive therapy (ECT) is the most effective treatment for major depression but also carries risk of cognitive side effects. The ability to predict whether treatment will be effective before initiation of treatment could significantly improve quality of care, reduce suffering, and diminish costs. We sought to carry out a comprehensive and definitive study of the relationship between the background electroencephalography (EEG) and therapeutic response to ECT. METHODS: Twenty-one channel resting EEG was collected pre-ECT and 2 to 3 days after ECT course from 2 separate data sets, one to develop an EEG model of therapeutic response (n = 30) and a second to test this model (n = 40). A 3-way principal components analysis was applied and coherence and spectral amplitude across 6 frequency bands were examined. The primary outcome measure was the Montgomery-Asberg Rating Scale (MADRS). RESULTS: Four patterns of amplitude and coherence along with baseline MADRS score accounted for 85% of the variance in posttreatment course MADRS score in study 1 (R = 0.85, F = 11.7, P < 0.0002) and 53% of the variance in MADRS score in study 2 (R = 0.53, F = 5.5, P < 0.003). Greater pre-ECT course anterior delta coherence accounted for the majority of variance in therapeutic response (study 1: R = 0.44, P = 0.01; study 2: R = 0.16, P = 0.008). CONCLUSIONS: These results suggest a putative electrophysiological biomarker that can predict therapeutic response before a course of ECT. Greater baseline anterior delta coherence is significantly associated with a better subsequent therapeutic response and could be indicative of intact circuitry allowing for improved seizure propagation.

Full Text

Duke Authors

Cited Authors

  • Scangos, KW; Weiner, RD; Coffey, EC; Krystal, AD

Published Date

  • June 2019

Published In

Volume / Issue

  • 35 / 2

Start / End Page

  • 95 - 102

PubMed ID

  • 30531398

Pubmed Central ID

  • 30531398

Electronic International Standard Serial Number (EISSN)

  • 1533-4112

Digital Object Identifier (DOI)

  • 10.1097/YCT.0000000000000557


  • eng

Conference Location

  • United States