Research paperRemission of depression is associated with asymmetric hemispheric variation in EEG complexity before antidepressant treatment
Introduction
Depressive disorder is one of 12 leading causes of the global burden of disability worldwide, according to the report from the World Health Organization (WHO) in 2017. The adjusted global point prevalence of depressive disorder is 4.7% (Ferrari et al., 2013), affecting 9.5% of the American general population with mood disorders and 6.7% with major depressive disorder (Kessler et al., 2005). It not only affects quality of life but also causes an adverse socioeconomic impact. Depressive disorder, especially major depressive disorder, typically exhibits significant loss of pleasure, depressed mood, changed appetite, sleep patterns, and suicidal intentions. These symptoms occur almost daily and persist for at least two weeks (American Psychiatric Association, 2013).
The current available treatments for depression involve pharmaceutical approaches (selective serotonin reuptake inhibitors, tricyclic antidepressants, etc.), neuronal modulation with electrical techniques (electroconvulsive treatment, repetitive transcranial magnetic stimulation, deep brain stimulation), and psychosocial therapies (Qaseem et al., 2016). Antidepressants are frequently used to manage major depressive disorder in the acute phase, but limited efficacy to achieve remission and lack of promising response rates has been reported with this approach (Cipriani et al., 2018).
There has been increasing interest in identifying the biomarkers of depression and its treatment efficacy using electroencephalogram (EEG). Recently, studies have focused on the dynamic nature of the brain and its relationship with mental illness (Yang and Tsai, 2013). Novel analytic approaches have been applied to explore the association between nonlinear EEG characteristics and depression. Complexity is different from variability since the former is a quantitative measure of the dynamics (i.e., regularity) of a signal while the latter is a statistical measure of variance. For instance, a highly cyclic signal may have high variability (i.e., large variation) but low complexity (regular cyclic dynamics) (Yang et al., 2018); such difference can be found in the case of cardiac activity resulting from atrial fibrillation (van den Berg Maarten P. et al., 1997; Costa et al., 2002, 2005). Measuring the complexity of physiological signals therefore can reflect the pathophysiology of diseases that either increased (i.e., tending toward randomness) or decreased in complexity (i.e., tending toward regularity), which is not only associated with medical illness but also with mental disorders, including schizophrenia, bipolar disorder, and depressive disorder (Yang et al., 2015; Hager et al., 2017; Li et al., 2008; Méndez et al., 2012; Tang et al., 2009). Through such differences, it has been found that depressive patients have a greater predictable brain wave pattern than healthy individuals, using nonlinear forecasting to measure the regularity of patients’ EEG signals (Nandrino et al., 1994). These patients also exhibited a higher EEG complexity during the resting-state in the anterior brain (Méndez et al., 2012) but had a stronger suppression with left-hemisphere lateralization in EEG complexity during the task (Li et al., 2008), compared to healthy individuals or patients with schizophrenia. However, this abnormal EEG complexity pattern can change after treatment (Thomasson et al., 2000; Okazaki et al., 2013).
To further understand how EEG complexity interacts with depression, researchers have employed analyses that have focused on specific EEG frequency bands or regions, specifically the alpha band. One study applied the conventional band-pass filter to capture alpha amplitudes and calculated the alpha EEG complexity for 20 patients with depression and 20 healthy controls; they found that depressive patients exhibited a greater alpha complexity during the resting state compared with the control group (Tang et al., 2009). However, another study showed that there was no significant difference in average alpha complexity between depressive patients who responded to repetitive transcranial magnetic stimulation versus those who had a poor treatment response (Arns et al., 2014). Since increased frontal alpha activity (Jaworska et al., 2012) or greater left frontal alpha asymmetry is evident among depressive patients (Henriques and Davidson, 1991; Chang et al., 2012), it is surprising that little is known about the nonlinear complexity profile of alpha rhythmicity from the anterior area of the brain in patients of major depressive disorder, and whether complexity measures of EEG signals vary across the frontal region between patients with or without remission in depression.
The power spectral analysis has been widely used to address the relationship between alpha power and depression. This conventional analysis provides a general alpha profile, which lacks an adequate time-frequency resolution. To better identify the role of alpha oscillation on the nonlinear EEG matrix among the patients with or without remission of depression, the method for alpha rhythm extraction is important. The Hilbert-Huang transform (HHT) (Huang Norden E. et al., 1998) has been developed and widely used to capture the intrinsic components of the physiological signals and applied to EEG signals (Kortelainen and Väyrynen, 2015; Zhu et al., 2015). Unlike Fourier-based analysis, this method holds no priori assumptions as to the underlying structures of the time series and is therefore suitable for analyzing time series which consist of multiple periodic components. This decomposition is based on a simple assumption that any information consists of a finite number of intrinsic components or oscillations. Each oscillation component, termed an intrinsic mode function (IMF), is sequentially decomposed from the original time series by a sifting process. Therefore, it is expected to isolate confounding factors and measurement noises, which is a better approach to analyze noisy oscillations typically observed in EEG data.
The aim of the current study was to explore the nonlinear profile of alpha rhythmicity from the anterior area of the brain among patients with depressive disorder thoroughly. We have not only measured alpha power by using power spectral analysis but also applied HHT to extract an individual's instantaneous alpha amplitudes from patients who achieved the remission in depression or those who had poor antidepressant treatment response. The overall regional complexity of patients' instantaneous alpha amplitudes on the frontal, frontal-central, or bilateral frontal area was measured with the multiscale entropy (MSE), and the homogeneity of alpha-amplitude oscillation within the region during the resting-state was also tested. We hypothesized that antidepressant treatment responders would differ from poor responders in terms of alternations in anterior brain alpha-amplitude complexity and the homogeneity of regional complexity with a reversed left-frontal lateralization after treatment during resting-state; poor responders showed an opposite trend associated with treatment response.
Section snippets
Participants and experimental procedure
The present study was a retrospective study that was a part of a clinical trial conducted between May 2007 and February 2010. All the patients agreed and signed the written informed consent form, and the procedure was approved by Kai-Syuan Psychiatric Hospital's institutional review board. The study was also registered on Clinical.trials.gov (identifier number: NCT01075529).
The details of the recruitment of participants are described in a separate study (Yang et al., 2017). In short, patients
Participants’ demography
No differences in sex (p = .28), age (t = -0.20, p = 0.85), age of first or major depressive episode onset (t = -1.01, p = 0.32), number of major depressive episodes (t = 0.83, p = 0.41) (Table 1), or years of education (t = -1.38, p = 0.18). For poor responders, 18.18%, 9.09%, 54.54%, 13.64%, and 4.55% hold their highest academic degree in primary school, junior high school, high school, college and university, respectively; of remitters, 18.18%, 9.09%, 45.45%, 4.55%, 18.18% and 4.55% have
Discussion
The main findings of the current study are the following. 1) No differences were found between remitters and poor responders at baseline, including alpha power from the spectral analysis, averaged instantaneous alpha amplitudes, or the averaged complexity of alpha-amplitudes. 2) An increased heterogeneity of regional alpha complexity within frontal-central area after treatment among patients with poor treatment response was observed. 3) Contrary to expectations, the remitters showed a different
Contributors
H. J., Tsai wrote the manuscript and carried out the analysis with support from Albert C. Yang; C. H., Lin and W. C. Yang organized the clinical trial and collected the data; Albert C. Yang supervised the analytical approaches; Albert C. Yang and S. J. Tsai offer clinical perspectives on the findings and critically reviewed the final manuscript.
Role of the funding sources
The Ministry of Science and Technology of Taiwan provided funding only and had no other role in study design, data collection, analysis or interpretations.
Ethical statement
The authors assert that all procedures contributing to this work comply with the ethical standards of both Good Clinical Practice procedures and the most recent revision of Helsinki Declaration. All the patients agreed and signed the written informed consent, the procedure was approved by Kai-Syuan Psychiatric Hospital's institutional review board and was also registered on Clinical.trials.gov (identifier number: NCT01075529).
Declaration of Competing Interest
None of the authors has any potential conflict of interest to be disclosed.
Acknowledgments
Dr. Albert C. Yang and S. J. Tsai were supported by grants [109-2628-B-010-011; 109-2321-B-010-006; and 109-2634-F-075-001] from the Ministry of Science and Technology of Taiwan. Dr. Albert C. Yang was also supported by Mt. Jade Young Scholarship Award from Ministry of Education, Taiwan as well as Brain Research Center, National Yang-Ming University and the Ministry of Education (Aim for the Top University Plan), Taipei, Taiwan. Dr. Wei-Cheng Yang and Dr. Ching-Hua Lin were supported by grants
References (39)
- et al.
Non-linear EEG analyses predict non-response to rTMS treatment in major depressive disorder
Clin. Neurophysiol.
(2014) - et al.
Electroencephalographic alpha measures predict therapeutic response to a selective serotonin reuptake inhibitor antidepressant: pre- and post-treatment findings
Biol. Psychiatry
(2008) - et al.
Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis
Lancet
(2018) - et al.
Neural complexity as a potential translational biomarker for psychosis
J. Affect. Disord.
(2017) - et al.
Alpha power, alpha asymmetry and anterior cingulate cortex activity in depressed males and females
J. Psychiatry Res.
(2012) - et al.
Microstates in resting-state EEG: current status and future directions
Neurosci. Biobehav. Rev.
(2015) - et al.
Abnormal EEG complexity in patients with schizophrenia and depression
Clin. Neurophysiol.
(2008) - et al.
Event-related dynamics of cortical rhythms: frequency-specific features and functional correlates
Int. J. Psychophysiol.
(2001) - et al.
Effects of electroconvulsive therapy on neural complexity in patients with depression: report of three cases
J. Affect. Disord.
(2013) Functional brain imaging based on ERD/ERS
Vis. Res.
(2001)
Current source density measures of electroencephalographic alpha predict antidepressant treatment response
Biol. Psychiatry
Is mental illness complex? From behavior to brain
Prog. Neuropsychopharmacol. Biol. Psychiatry
Reduced physiologic complexity is associated with poor sleep in patients with major depression and primary insomnia
J. Affect. Disord.
Cognitive and neuropsychiatric correlates of EEG dynamic complexity in patients with Alzheimer's disease
Prog. Neuropsychopharmacol. Biol. Psychiatry
Analysis of spike waves in epilepsy using Hilbert-Huang transform
J. Med. Syst.
Severity classification on the Hamilton Depression Rating Scale
J. Affect. Disord.
Diagnostic and Statistical Manual Of Mental Disorders
An integrative assessment of the psychophysiologic alterations in young women with recurrent major depressive disorder
Psychosom. Med.
Multiscale entropy analysis of biological signals
Phys. Rev. E Stat. Nonlin. Soft Matter Phys.
Cited by (2)
EEG microstate in first-episode drug-naive adolescents with depression
2022, Journal of Neural Engineering