EEG power, cordance and coherence differences between unipolar and bipolar depression

Abstract

Introduction

Understanding the biological underpinnings of unipolar (UD) and bipolar depression (BD) is vital for avoiding inappropriate treatment through the misdiagnosis of bipolar patients in their first depressive episode. One plausible way to distinguish between UD and BD is to compare EEG brain dynamics to identify potential neurophysiological biomarkers. Here we aimed to test group differences in EEG power, cordance and coherence values between UD and BD.

Methods

Twenty-five bipolar and 56 unipolar depression patients were recruited. Sociodemographic and clinical variables were collected in addition to resting state EEG. Data was analyzed with multivariate and repeated analyses of variance where parametric assumptions were met.

Results

Accordingly, we did not find any differences in the EEG absolute power and frontal asymmetry indexes between UD and BD. Regarding cordance, significant group differences were observed in the right theta cordance values (p=0.031). Regarding coherence, BD patients (as compared to UD) exhibited greater central–temporal theta (p=0.003), and parietal–temporal alpha (p=0.007) and theta (p=0.001) coherence. Lastly, less alpha coherence in BD was present at right frontal–central (p=0.007) and central inter-hemispheric (p=0.019) regions.

Conclusions

Our results demonstrate that EEG cordance and coherence values have potential to discriminate between UD and BD. The loss of temporal synchronization in the frontal interhemispheric and right sided frontolimbic neuronal networks may be a unique feature that distinguishes between BD and UD.

Introduction

Although bipolar depression (BD) and unipolar depression (UD) have been regarded as distinct clinical entities, treated with distinct therapeutic strategies, previous studies demonstrated that 60% of BD cases were incorrectly diagnosed as UD, and were consequently treated inappropriately (Dunner, 2003, Goodwin and Jamison, 2007, Hirschfeld et al., 2003). According to the literature, some well-known clinical symptoms, such as psychomotor retardation and severe anhedonia, were more pronounced in BD patients (Brockington et al., 1982, Mitchell and Malhi, 2004). Some of these symptoms have been linked to the specific biological underpinnings of BD as a result of electrophysiological investigations (Dewan et al., 1988). The DSM-4 and 5 argue however, that BD can only be differentiated from UD by the occurrence of hypomanic episodes throughout the life cycle of affected individuals (APA, 2013). Given this, the identification of potential biomarkers could be crucial in helping to distinguish between BD and UD, especially as the question of whether a biological overlap between BD and UD exists or not is still unresolved.

Comparing BD and UD with the use of brain oscillations could be one plausible way of dissociating the two diagnoses, though this has only been attempted by a few studies. For instance, Lieber and Newbury (1988) showed a greater reduction of alpha as well as excessive beta power activity in BD as compared to UD. In addition, they found that deficient left hemisphere alpha power in BD and decreased interhemispheric theta coherence in UD could discriminate these two groups. Recently, a magnetoencephalogram (MEG) study demonstrated that, compared to UD, BD patients had a greater alpha activity in bilateral temporo-parieto-occipital regions (Lee et al., 2010). More specifically, patients with BD had greater activation in the alpha band for right inferior/superior temporal, left middle occipital areas, and right precentral gyrus. In addition, resting state fMRI studies have also improved our insights into discriminating BD from UD. For instance, Liang et al. (2013) found many similarities between UD and BD but also demonstrated significant decreases in the bilateral precentral gyrus and left cingulate in UD, whereby significant decreases were observed in the right precentral gyrus, right cingulate, and left inferior frontal gyrus in the BD group. Given the limited number of comparative studies, and different methods, there appears to be little consensus on the biological differences between UD and BD. One possible way to overcome this issue may be the use of a combination of several EEG biomarkers, which eventually was the main purpose of the current study.

Numerous past EEG studies have attempted to evaluate the distinct features of BD and UD as compared to other clinical and non-clinical populations. One well replicated finding in UD is that, compared to healthy subjects, an interhemispheric frontal alpha asymmetry has been found due to an increased left frontal alpha power as it is well-known indicator of idling activity on that side (Henriques and Davidson, 1991, Noonan et al., 2009). Similarly, greater right-sided hyperactivity has been demonstrated in BD patients (Clementz et al., 1994, David and Cutting, 1990 Apr). Partly as a result of these findings, frontal alpha asymmetry has consequently been suggested to be a trait marker for depression in many studies (Coan and Allen, 2003, Debener et al., 2000, Gotlib et al., 1998, Henriques and Davidson, 1991), though some controversy still exists (Reid et al., 1998). Nevertheless, in a longitudinal study, frontotemporal asymmetry has predicted the occurrence of hypomanic and depressive episodes of BD patients (Clementz et al., 1994), with decreased alpha and increased theta power in the frontocentral regions being the most common findings in BD patients (Clementz et al., 1994, Degabriele and Lagopoulos, 2009). With regards to coherence, an EEG index for brain connectivity, BD was associated with a lack of inter-hemispheric synchronization, whereas UD was found to be related to lower anterior and posterior interhemispheric slow-wave coherence in another study (Lieber and Newbury, 1988). Lastly, a recent EEG study reported a global decrease in alpha synchronization in BD patients compared to healthy participants, where the decrease was most prominent in the right fronto-central and centro-parietal connections (Kim et al., 2013).

Given all, this study aimed to explore the resting state EEG differences between UD and BD by conjointly using well-replicated EEG biomarkers such as absolute power, coherence and cordance, which have been shown to be an index of cerebral local perfusion in previous studies (Hughes and John, 1999, Niedermeyer and Lopes da Silva, 2004, Shaw et al., 1978). Specifically, this study was based on the following three-fold hypothesis. We first attempted to support the findings in the common literature and predicted that BD patients would show a greater global increase in theta and a decrease in the alpha band activity. Secondly, regarding cordance, as theta cordance values were related to the local cerebral perfusion in previous studies, we expected group differences between UD and BD patients. Lastly, we hypothesized that the BD group would display a more drastic decrease in EEG coherence as compared to the UD group.