Alpha Wavelet Power as a Biomarker of Antidepressant Treatment Response in Bipolar Depression

TL;DR

This study introduces a novel EEG-based biomarker using alpha wavelet power to predict antidepressant treatment response in bipolar depression with high accuracy from a single measurement.

Contribution

It develops a new EEG biomarker based on alpha wavelet power that can predict treatment response after one measurement, unlike previous methods requiring multiple assessments.

Findings

Alpha wavelet power is significantly higher in responders.

The classification algorithm achieved 90% sensitivity and 100% specificity.

The biomarker is effective with a single EEG measurement.

Abstract

There is mounting evidence of a link between the properties of electroencephalograms (EEGs) of depressive patients and the outcome of pharmacotherapy. The goal of this study was to develop an EEG biomarker of antidepressant treatment response which would require only a single EEG measurement. We recorded resting, 21-channel EEG in 17 inpatients suffering from bipolar depression in eyes closed and eyes open conditions. The EEG measurement was performed at the end of the short washout period which followed previously unsuccessful pharmacotherapy. We calculated the normalized wavelet power of alpha rhythm using two referential montages and an average reference montage. In particular, in the occipital (O1, O2, Oz) channels the wavelet power of responders was up to 84% higher than that of nonresponders. Using a novel classification algorithm we were able to correctly predict the outcome of treatment with 90% sensitivity and 100% specificity. The proposed biomarker requires only a single EEG measurement and consequently is intrinsically different from biomarkers which exploit the changes in prefrontal EEG induced by pharmacotherapy over a given time.