One Week Antidepressant Efficacy Revealed by EEG Analysis

565 

Abstract Submission 

Yueheng Peng¹, Yan Peng², Guangying Wang¹, Fali Li³, Peng Xu³

¹University of Electronic Science and Technology of China, Chengdu, Sichuan, ²West China Second University Hospital, Chengdu, Sichuan, ³School of life Science and technology, University of Electronic Science and Technology of China, Chengdu, Sichuan

Yueheng Peng  
University of Electronic Science and Technology of China
Chengdu, Sichuan

Yan Peng  
West China Second University Hospital
Chengdu, Sichuan

Guangying Wang  
University of Electronic Science and Technology of China
Chengdu, Sichuan

Fali Li  
School of life Science and technology, University of Electronic Science and Technology of China
Chengdu, Sichuan

Peng Xu  
School of life Science and technology, University of Electronic Science and Technology of China
Chengdu, Sichuan

The treatment of major depressive disorder (MDD) was widely investigated, but their efficacy was evaluated by clinical scales, such as Hamilton Depression Scale and Patient Health Questionnaire-9, which were highly depended on patients' subjective feeling and required at least four weeks after baseline to reflect the medication efficacy. Under this situation, researchers began to look for electroencephalogram biomarkers to sensitively and objectively evaluate short-term efficacy, which potentially facilitates treatment selection and reduces time cost. In recent years, EEG analysis has been widely explored, utilizing multiple different tasks such as emotional face presentation and working memory, which might reflect differences in brain functioning between depression patients and healthy controls. In our current study, based on the eyes-open EEG recorded during tasks, the corresponding event-related potential (ERP) at baseline and week1 was investigated to reveal the efficacy of one-week medication. Related brain networks were constructed to further demonstrate the brain changes after one-week medication.

In this study, there were twenty-five medication-free depressed patients received medicine for two months. At the last week, all patients were divided into responders and non-responders based on Hamilton Depression Scale. 21-channel eyes-open EEG datasets were collected during tasks at baseline and after one-week treatment. Concretely, the binaural acoustic stimuli was of 40-ms duration. The tones were presented in five different intensity levels: 60, 70, 80, 90 and 100 dB. 100 times of each intensity level were presented in a random order with a randomized inter-stimulus interval of 1600-2100 ms. Thereafter, the EEG datasets collected during the tasks were further pre-processed into artifact-free 1-s-length segments by adopting procedures, including the reference electrode standardization technique re-referencing, [1, 10] Hz offline-bandpass filtering, baseline correction and artifact removal, etc. Subsequently, the corresponding ERPs were statistically compared under different phases. Eventually, corresponding EEG networks were constructed by using Phase Locking Value (PLV) and then statistically compared between the two stages.

On one hand, Fig. 1 illustrated the ERP amplitude of N100 and P200 between baseline and week1 in responder arm. Obvious reduction after taking medicine for one week could be observed. Specifically, the amplitudes of both N100 and P200 reduced. Conversely, no significant difference in ERP amplitude between week0 and week1 in non-responder group could be found.
On the other hand, Fig. 2 showed the topological differences of related EEG networks (derived from 100 dB N100) between week0 and week1 for the responder cohort, where the blue solid line indicates enhancement (week0<week1) and the red line indicates reduction (week0>week1). As we can see, there was a significant reduction after one week treatment, illustrating that the medication alleviated the activity of related brain regions. In contrast, after one-week medication, no significant changes in network patterns could be observed in non-responders.

Our present study revealed that after one-week medication, the ERP amplitude of responders decreased. And the corresponding brain network difference between week0 and week1 further demonstrated that the short-term treatment attenuated the activity of frontal-parietal lobes, which might help doctors to adjust therapeutic regimen at the initial stage of a certain long-term treatment.

Psychiatric (eg. Depression, Anxiety, Schizophrenia) ¹

Connectivity (eg. functional, effective, structural)

EEG ²

Affective Disorders

Electroencephaolography (EEG)

Emotions

Therapy