Cortical mechanisms of the symptomatology in major depressive disorder: A resting EEG study

Journal of Affective Disorders 131 (2011) 243–250

Contents lists available at ScienceDirect

Journal of Affective Disorders j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j a d

Research Report

Cortical mechanisms of the symptomatology in major depressive disorder: A resting EEG study Tien-Wen Lee a,b, Younger W.-Y. Yu c, Ming-Chao Chen d, Tai-Jui Chen e,f,⁎ a

Department of Psychiatry, Chang Gung Memorial Hospital, Taoyuan County, Taiwan College of Medicine, Chang Gung University, Taoyuan County, Taiwan Yu's Psychiatric Clinic, Kaohsiung, Taiwan d Kai-Suan Psychiatric Hospital, Kaohsiung, Taiwan e Department of Psychiatry, E-DA Hospital, Kaohsiung County, Taiwan f Department of Occupational Therapy, I-Shou University, Kaohsiung County, Taiwan b c

a r t i c l e

i n f o

Article history: Received 26 June 2010 Received in revised form 15 December 2010 Accepted 15 December 2010 Available online 22 January 2011

Keywords: Major depression Electroencephalography (EEG) Coherence Spectrum Hamilton Depression Rating Scale (HDRS) Symptomatology

a b s t r a c t Background: Diagnosis and treatment rely on symptom criteria in modern psychiatry. However, the cortical mechanisms of symptomatology in major depressive disorder (MDD) are still not clear. This study examined neural correlates of symptom clusters of MDD by electroencephalography (EEG). Methods: Resting state eye-closed EEG signals were recorded in 196 depressive patients. Quantitative EEG (qEEG) of regional power, coherence and power series correlation across delta, theta, alpha and beta frequencies were used to correlate with overall depression severity evaluated by the Hamilton Depression Rating Scale (HDRS). Further, statistical comparisons between patients with high vs. low qEEG indices (median-split) were undertaken regarding symptom severity of core depression, sleep, activity, psychic anxiety, somatic anxiety, and delusion. Results: None of the qEEG indices significantly correlated with overall depression severity or differentiated symptom severity of core depression, sleep, activity and psychic anxiety. A higher symptom severity of somatic anxiety was associated with higher regional power over widespread cortical regions and lower strengths at bi-temporal, temporo-parietal and fronto-parietal connections. A higher symptom severity of delusion was associated with higher regional power in the frontal and temporal regions, and lower strengths at inter-hemispheric (frontal, temporal and parietal) and fronto–temporo-parietal connections. Limitations: Our EEG recording with sampling rate of 128 Hz and 20 electrodes may provide restricted spatial and temporal precision. Conclusions: Our results suggest that cortical mechanisms play important roles in the symptom manifestation of cognitive distortion (sub-score of delusion) and somatic anxiety in MDD. Our findings further imply that psychic anxiety and somatic anxiety are distinct entities. © 2010 Elsevier B.V. All rights reserved.

1. Introduction Major depressive disorder (MDD) is regarded as a polymorphic syndrome involving broad-range aberrations. ⁎ Corresponding author. Department of Psychiatry, E-DA Hospital, No. 1, E-Da Rd., Chiao-Su Village, Yen-Chao Township, Kaohsiung County, Taiwan, ROC. Tel.: +886 7 6150011x2650; fax: +886 7 6155352. E-mail address: [email protected] (T.-J. Chen). 0165-0327/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2010.12.015

It is characterized by very heterogeneous abnormalities in the central nervous system, involving widespread cortical and subcortical structures and diverse neurotransmitter and endocrine systems in the brain (Konarski et al., 2008; Maletic et al., 2007; Meyer, 2008; Stahl et al., 2003). Correspondingly, MDD manifestation covers diverse symptomatic, cognitive, affective, executive, autonomic, perceptual and attentional domains. Although direct comparison of MDD and matched controls has yielded fruitful results about the associated

244

T.-W. Lee et al. / Journal of Affective Disorders 131 (2011) 243–250

neuropathology, an inevitable ambiguity exists as to what neuropsychological dysfunction that comparison stands for. The ambiguity originates from the heterogeneity of MDD, and in turn interferes with attributing biological discoveries to specific phenomenological or neuropsychological constructs. To extend the understanding of MDD, a useful strategy is to isolate each component and then clarify its contribution to MDD. This approach has gained substantial success in many research topics, for example, in identifying the neural underpinnings of suicidality in MDD by comparing suicide attempters with high and low lethality (Oquendo et al., 2003). Up to now, the diagnosis and treatment of MDD are largely based on symptom criteria (American Psychiatric Association, 1994); however, the neural correlates of MDD symptomatology are still not clear. Factor analysis has revealed that the numerous clinical manifestations of MDD can indeed be categorized into fewer symptom clusters (Serretti et al., 1998). Conceptually, dividing clinical manifestation into summarized symptomatic constructs is similar to dissecting mental phenomena (e.g. IQ or temperament) into several psychological constituents. Factorization allows linking biological features to each symptom construct instead of to the overall depression syndrome or to a single symptom, is equipped with higher sensitivity and plausibility, and has been proven to be useful in genetic research on MDD (Serretti et al., 1999; Yu et al., 2002). This study planned to investigate the cortical neural correlates underlying depression symptomatology by combining a symptom-cluster-oriented approach and quantitative electroencephalography (qEEG). The symptoms were assessed by the 21-item Hamilton Depression Rating Scale and were divided into six clusters, namely core, sleep, activity, psychic anxiety, somatic anxiety, and delusion (Serretti et al., 1999). The qEEG indices were comprised of spectrum power and connectivity strengths, and were related to each symptom cluster of MDD. The rationale to probe the interaction between neural modules, in addition to the traditional analyses which emphasize regional electrophysiological profiles, was based on the fact that execution of brain function relies on efficient interactions among its components. Accordingly, neural dysfunction may arise from abnormal interactions between neural nodes, in contrast to the consequence of regional defects. Compatible with this perspective, recent affective neuroscience has discovered that the strengths of corticolimbic connections possess potential as a biomarker of MDD (Anand et al., 2005; Bae et al., 2006; Brody et al., 2001; Fingelkurts et al., 2007; Hamilton and Gotlib, 2008). Evidence has also implied widespread cortical disturbances, both focal and inter-regional, in MDD (Fingelkurts et al., 2007; Goodwin, 1997; Vasic et al., 2008). Among the six symptom clusters, the roles of deep cortical and subcortical mechanisms have been addressed for core depression, sleep, activity and psychic anxiety (Bishop, 2007; Cermakian and Boivin, 2009; Drevets et al., 2008; Kalia, 2005; Knab et al., 2009; Quirk and Mueller, 2008). We thus did not expect prominent cortical involvement of the above four symptom clusters. As to the symptom of delusion, we predicted that frontal and temporal power may differentiate the severity of cognitive distortion (Kunert et al., 2007). Further, the inter-hemispheric and fronto-temporal discon-

nections, frequently reported in schizophrenia, may also present in the MDD patients vulnerable to cognitive distortion or psychosis (David, 1994; Lawrie et al., 2002; Mitelman et al., 2005). As to the symptom cluster of somatic anxiety, we predicted that widespread cortical regions may distinguish the severity of somatic anxiety, resembling the cortical neural dysfunction of somatization or somatoform disorders (GarciaCampayo et al., 2001; Tokunaga et al., 1997). 2. Materials and methods 2.1. Subjects A total of 196 patients who met the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria for major depressive disorder (MDD) were enrolled in this study (American Psychiatric Association, 1994); these participants were recruited from Yu's Psychiatric Clinic, Kai-Suan Psychiatric Hospital, Kaohsiung City, and E-DA Hospital, Kaohsiung County. The Hamilton Depression Rating Scale (HDRS-21) was evaluated for each patient and those who scored less than 18 were excluded. The patients had no additional diagnoses on Axis I of the DSM-IV (including schizophrenia, substance abuse, generalized anxiety, panic and obsessive compulsive disorders) or major medical and/or neurological disorders. All the participants were evaluated by licensed psychiatrists following a semi-structural interview process, while neurological and physical examinations were performed by licensed medical doctors. Medical checkup included thyroid hormone level and careful scrutiny of medical records. Only the patients who had been medicationfree for at least two weeks were enrolled. This project was approved by the institutional ethics committee, with the standard conforming to Helsinki Declaration. Informed consent was obtained from all participants prior to the commencement of the investigation. To evaluate the specific clusters of depressive symptoms, the HDRS items were grouped according to the following factors: core (Items 1, 2, 7, 8, 10, and 13), sleep (Items 4, 5, and 6), activity (Items 7 and 8), psychic anxiety (Items 9 and 10), somatic anxiety (Items 11, 12, and 13), and delusion (Items 2, 15, and 20), as described by Serretti et al. (1999). 2.2. EEG recordings and analyses All participants received a 3-minute conventional, eyes closed, awake, digital EEG after 5-minute habituation to the experimental environment before formal treatment (Brain Atlas III computer, Biologic System Company, Mundelein, IL, USA). Recordings were in accord with the international 10–20 system with linked ear reference, 128 Hz sampling rate, high pass filter 0.05 Hz, low pass filter 70 Hz, notch filter 60 Hz and impedance below 3 kΩ. We recorded 20 electrodes (Fp1, Fp2, F7, F3, Fz, F4, F8, T3, C3, Cz, C4, T4, T5, P3, Pz, P4, T6, O1, Oz, and O2). Vigilance was monitored by the EEG technician who would alert subjects when signs of drowsiness appeared in the tracings. The vertical eyeball movement was detected from electrodes placed above and below the right eye, with the horizontal analog detected from electrodes placed at the left outer canthus. EEG traces with artifacts were deleted by a semi-automated module provided by the software EEGLAB

T.-W. Lee et al. / Journal of Affective Disorders 131 (2011) 243–250

(http://sccn.ucsd.edu/eeglab). In brief, the artefact segments were first deleted via visual inspection by experienced EEG technician and then the signal quality was examined by channel statistics and QQ-plot. The EEG channels that did not pass Kolmogorov–Smirnov test (P b 0.05) were re-examined and trimmed until the EEG signals behaving like Gaussian distribution. Since the functional connectivity was our main interest, we adopted Hjorth's method of surface Laplacian to re-reference the EEG data to reduce false-positive results due to reference dependency (Hjorth, 1975; Pfurtscheller and Andrew, 1999). We defined the frequency bands as follows: delta 1 to 4 Hz, theta 4 to 8 Hz, alpha 8 to 12 Hz, beta 12 to 24 Hz, beta1 12 to 18 Hz, and beta2 18 to 24 Hz (since the relationship between beta spectrum and the 6 depressive symptom clusters is still not clear, beta, beta1 and beta2 frequency bands were all included in the analyses (Aftanas and Pavlov, 2005; Korb et al., 2008; Pollock and Schneider, 1990; Strelets et al., 1997; Yamada et al., 1995)). To investigate connectivity strength between separate cortical regions, we applied two complementary approaches of coherence analyses and time–frequency power correlation (TFPC), with the former addressing phase consistency and the latter addressing synchronized oscillation of power (Laufs et al., 2003; Rusalova and Kostyunina, 2004). The implementation of coherence analysis and time–frequency power series correlation both start with short-time Fast Fourier Transform (FFT) with time window width 1000 ms and sliding step 500 ms. The averaged power spectrum and phase were calculated for every sliding window and frequency band. TFPC was calculated by Pearson correlation to the betweenchannel (between-region) power series. The method of between-channel coherence can be referred to elsewhere (Pfurtscheller and Andrew, 1999). The regional mean power was calculated by FFT and was log-transformed to approach Gaussian distribution. We made a systemic approach to examine the functional connectivity of 37 electrode pairs: (1) six symmetrical connections of F7-F8, F3-F4, C3-C4, T3–T4, T5–T6 and P3–P4; (2) fifteen ipsilateral channel pairs of the left hemisphere from all the possible combinations of F3, F7, C3, P3, T3 and T5; (3) fifteen ipsilateral channel pairs of the right hemisphere from all the possible combinations of F4, F8, C4, P4, T4 and T6; (4) one midline anterior–posterior connection of Fz–Pz. Given that reduction of the connectivity strengths has been known to be accompanied with aging process (Charlton et al., 2010; Damoiseaux et al., 2008; Sambataro et al., 2010), the age effect was regressed out for the above qEEG indices using linear regression model, and the residuals were incorporated into further analyses of depression symptom scores.

245

null hypothesis assumed that if a certain qEEG index was irrelevant to a specific depressive symptom cluster, the two subgroups divided by the qEEG index would not demonstrate a significant difference in that symptom sub-score. For each of the test sets, the criterion for significance was set at P b 0.05, two-tailed. For each qEEG index, we reported both P b 0.01 and Bonferroni correction according to P = 1 − (1 − 0.05)1/n, where n equals the number of comparisons in case the multiple comparison correction is too stringent. 3. Results 3.1. Demographic information and qEEG indices Among the 196 recruited patients, 112 were females and 84 were males. The mean age was 40.51 (SD 15.28) and the mean HDRS score was 27.6 (SD 4.1), with the mean sub-score of core 10.8 (SD 1.6), sleep 4.4 (SD 1.9), activity 2.2 (SD 0.9), psychic anxiety 3.2 (SD 1.0), somatic anxiety 5.7 (SD 1.4) and delusion 2.9 (SD 1.4). 3.2. Regional power and depressive symptomatology We examined the regional powers and the depressive symptomatology. The statistical threshold was set at 0.0030 according to Bonferroni correction (α = 0.05, two sided, n = 17—equal to the number of analyzed electrodes, Fp1, Fp2 and Oz not included). No correlation was found between total HDRS score and regional power. The median-split analyses did not show significant differences regarding all the HDRS sub-scores except for somatic anxiety and delusion. It is interesting that the same pattern was also present in the connectivity analyses (described in the next two sections). In other words, our results suggested that cortical mechanisms, regional and inter-regional, contribute to the MDD symptoms of somatic anxiety and delusion but not to the symptoms of depressive core, sleep, activity, and psychic anxiety, and are not related to the overall depression severity. We noticed that the electrode-frequency couples differentiating somatic anxiety involved broad cortical regions, whereas the regional power at frontal and temporal regions differentiated the severity of delusion. Most of the significant findings centered on the beta/beta1/beta2 frequencies at threshold P b 0.01, see Table 1. Without exception, the patients with higher regional power had higher scores of somatic anxiety and delusion at the electrode-frequency pairs showing significant between-group differences. 3.3. Between-region time-frequency power series correlation (TFPC) and depressive symptomatology

2.3. Statistical analyses To examine the neural correlates reflecting global depression severity, the derived qEEG indices were correlated with total score of the HDRS. In addition, each qEEG index served to median-split the patients into two subgroups, i.e. high vs. low qEEG. Since the distribution of the sub-scores did not comply with Gaussian, we performed Mann–Whitney U-test to examine whether the two subgroups (high qEEG vs. low qEEG) differed in their symptom sub-scores of the HDRS. Our

We median-split the patients based on connectivity strengths into two subgroups of high connectivity vs. low connectivity. The statistical threshold was set at 0.0014 according to Bonferroni correction (α = 0.05, two sided, n = 37—equal to the number of channel pairs of interest). Our TFPC analyses showed that the connectivity strengths of a widely distributed network across several frequency bands differentiated the severity of somatic (not psychic) anxiety and delusion, see Table 2.

246

T.-W. Lee et al. / Journal of Affective Disorders 131 (2011) 243–250

Table 1 Comparison of the sub-scores of somatic anxiety and delusion in the depressive patients with high vs. low regional power across 17 channels at 6 frequency bands. δ F7 F3 Fz F4 F8 T3 C3 Cz C4 T4 T5 P3 Pz P4 T6 O1 O2

θ

α

β

β1

x

x+

x

x

β2

x x

+ + x

x+

x x x

x + x

x x+ + + x x x x+ x x x

x + x + x x x + x x +

Table 2 Comparison of the sub-scores of somatic anxiety and delusion in the depressive patients with high vs. low connectivity strengths across 37 connections at 6 frequency bands.

x+ x+ x+ x x+ x+ x+ x + x

x: somatic anxiety, +: delusion, with a P value of the Mann–Whitney U-test b 0.01. The threshold after Bonferroni correction was 0.0030, with the P value of the comparisons b 0.0030 marked in bold.

δ F7–F8 F3–F4 C3–C4 T3–T4 T5–T6 P3–P4 F7–T3 F7–T5 F3–T3 F3–T5 F4–T4 F4–T6 F8–T4 F8–T6 F7–C3 F7–P3 F3–C3 F3–P3 F4–C4

θ

+

α

β

β1 β2

x + x x

+ + +

x

x+ + x

+ + x

+

+

x x x x+ + x x x

+ +

δ F4–P4 F8–C4 F8–P4 F3–F7 F4–F8 T3–C3 T3–P3 T5–C3 T5–P3 T4–C4 T4–P4 T6–C4 T6–P4 T3–T5 T4–T6 P3–C3 P4–C4 Fz–Pz

θ

α

β

β1

x

β2 +

+ x + + + x+ x + + x

x x x

x x x

+ +

x+ x

+ + + x x x + x+

+

x+

+ x+

+

x: somatic anxiety, +: delusion, with a P value of the Mann–Whitney U-test b 0.01. The threshold after Bonferroni correction was 0.0014, with the P value of the comparisons b 0.0014 marked in bold.

4. Discussion None of the between-regional connectivity strengths derived by TFPC differentiated the symptom scores of core, sleep, activity and psychic anxiety, nor were significantly correlated with total HDRS scores. Without exception, the patients with lower TFPC values at the connection-frequency pairs with significant between-group differences showed higher scores of somatic anxiety and delusion.

3.4. Between-region coherence and depressive symptomatology As in the regional power and TFPC analyses, none of the between-regional coherence values significantly correlated with the total HDRS score nor distinguished the sub-scores of the symptoms of core, sleep, activity and psychic anxiety. As in the TFPC analyses, the patients with lower coherence values at the connection-frequency pairs with significant betweengroup difference showed higher sub-scores of somatic anxiety and delusion. The summarized topography of the significant networks relevant to the severity of somatic anxiety and delusion is illustrated in Fig. 1. The supplementary material summarized several points which might be interesting to the readers (http://www.websdj. idv.tw/kiki/Symptom_Suppl.pdf): (1) the details of regional power TFPC and coherence analyses, (2) the correlation analyses of qEEG indices and HDRS scores, (3) the averaged values of the regional power and the connectivity strengths are illustrated in Figure S1; (4) the negative influence of age effect on the qEEG indices which has been reported before (Charlton et al., 2010; Damoiseaux et al., 2008; Sambataro et al., 2010), (5) the consistency of connectivity analyses derived by TFPC and coherence methods was poor at lower frequency and fair at higher frequency range, supporting the potential application of TFPC in functional connectivity analysis at resting state.

Major depressive disorder (MDD) is heterogeneous in many perspectives. This study adopted a symptom-wise approach to elucidate the neural correlates of depression symptomatology. Concordant with our prediction, we found that cortical qEEG indices of regional power and connectivity strengths differentiated the severity of somatic anxiety and delusion. Cortical mechanisms, either regional or interregional, did not show a significant relevance with overall depression severity or the symptoms of core, sleep, activity and psychic anxiety. The sub-score of somatic anxiety is defined by summing HDRS item 11 (anxiety somatic), 12 (gastrointestinal) and 13 (somatic symptoms general). The somatic anxiety is conceptually akin to somatoform disorder in DSM-IV where the physical suffering is not explained by general medical conditions and the somatosensory disturbances, if present, reveal non-dermatomal distribution. Despite the long-observed co-morbidity of somatization, somatoform and depression, the exploration of neurodysfunctional patterns in functional somatic syndrome is in its infancy, and the diagnosis still relies on the presence of subjective distress in the absence of objective findings. Nevertheless, recent neuroimaging advancement has provided the important insight that "functional" somatic distress does have an "organic" physiological basis. For example, Hakala et al. (2004; Hakala et al., 2006) found that somatization disorder was associated with bilateral enlargement of caudate nuclei volumes and low glucose metabolism in the caudate and putamen. Our study revealed that higher regional powers across the brain, centering on beta frequencies were associated with higher severity of somatic anxiety in MDD patients. In contrast, stronger neural interactions indicated by the functional connectivity analyses were associated with lower somatic anxiety, with the connectivity strengths at bi-temporal,

T.-W. Lee et al. / Journal of Affective Disorders 131 (2011) 243–250

247

Fig. 1. Summarized topography of connection-frequency pairs that distinguished the severity of somatic anxiety (Left) and delusion (Right).

temporo-parietal and fronto-parietal connections surpassing conservative threshold with Bonferroni correction. It is interesting that our analyses did not notice any cortical neural correlates significantly relevant to psychic anxiety, suggesting that psychic anxiety and somatic anxiety are distinct entities. In addition to the striatal structures described above, functional imaging studies have revealed dysfunctional cortical circuits in somatization and somatoform disorders. Concordant with the concept that functional somatic syndrome is not caused by a focal neurological deficit; the reported cortical neural pattern of somatization is widespread, covering frontal, temporo-parietal, cerebellar and even the whole hemispheric regions (Garcia-Campayo et al., 2001). Further, the binding potential of benzodiazepine receptors is decreased in the superior frontal, temporal, and parietal cortices of the patients with somatoform disorders (Tokunaga et al., 1997). Our findings of regional powers over broad regions were thus compatible with extant knowledge and provided electrophysiological insight. The underlying mechanism of the diffuse cortical involvement in somatoform disorders is not clear, but it is probably mediated by the immune system and inflammation-associated processes (Dimsdale and Dantzer, 2007). There have not been any reports to date regarding the cortical connectivity pattern in the somatic anxiety of depression. The demonstrated inverse relationship between the connectivity strengths and the severity of somatic anxiety imply that the reduction in cortical information exchange can contribute to functional somatic discomfort in a depressive state. Together, diffuse higher regional powers and less interaction among neural modules could be the neuropathological features of somatic anxiety in depression. The sub-score of delusion is defined by summing HDRS item 2 (feelings of guilt), 15 (hypochondriasis) and 20 (paranoid symptoms), which, although not equivalent to real “delusion”, quantifies the degree of cognitive distortion and maladaptive beliefs about oneself and the world, held

with subjective certainty by the patient. The development of cognitive distortion has been suggested to involve various aberrant mental processes largely associated with cortical dysfunctions, such as attributional and attentional biases, misinterpretation of social signals, defective associative learning, impaired reasoning, and abnormal language-, perception- or emotion-related processes (Blackwood et al., 2000; Dawson et al., 2000; Kunert et al., 2007; Smets et al., 1992). As expected from the wide range of contributors to delusion, clinical observation has revealed very heterogeneous brain structure changes related to the delusion of organic causes, comprising the frontal, temporal, and parietal association cortices, and some subcortical structures (review) (Kunert et al., 2007). Our results showed a parallel relationship of the sub-scores of delusion and the values of mean power in the bilateral frontal and temporal regions, i.e. the higher the regional power, the stronger the cognitive distortion. Our connectivity analyses further revealed an opposite relationship of the sub-scores of delusion and the values of connectivity strengths of inter-hemispheric (frontal, temporal and parietal) and fronto–temporo-parietal interactions, with the connectivity strengths at inter-hemispheric, fronto-temporal (F4–T6), fronto-parietal (F7–C3) and temporoparietal (T5–P3) connections surpassing multiple comparisons correction. The noticed association of frontal/temporal power and sub-score of delusion is compatible with previous reports in that the severity of reality distortion in psychosis is related to abnormalities in the superior temporal gyrus and prefrontal cortex (Kim et al., 1999; MacDonald and Carter, 2003; Pae et al., 2004; Ragland et al., 2007; Rajarethinam et al., 2000; Simpson et al., 1999; Wright et al., 1995). Our finding that higher sub-scores of delusion were associated with lower connectivity strengths over a widely distributed network, i.e. inter-hemispheric and fronto–temporo-parietal, was akin to “disconnection” in schizophrenia. Disconnection theory explains psychosis as a consequence of failure in integrating information from several neural modules which may also

248

T.-W. Lee et al. / Journal of Affective Disorders 131 (2011) 243–250

cause a disruption in control and coordination processes. Abnormal inter-hemispheric transmission in schizophrenia has been supported by abundant evidence of experimental psychology and measurement of corpus callosum (review) (David, 1994). Previous neuroimaging studies have shown that a reduction in fronto-temporal functional connectivity is associated with an increase in psychotic symptoms (Higashima et al., 2006; Lawrie et al., 2002; Mitelman et al., 2005). Our findings of reduced fronto-parietal/temporo-parietal interaction and increased cognitive distortion have been less addressed in previous reports of psychotic disorders. Given that the parietal and temporal lobes are part of the network of language, self/other coordinate transformation and heteromodal perceptual integration (Rajarethinam et al., 2000; Wright et al., 1995; Zacks and Michelon, 2005), the reduced connection might contribute to the development of delusions via aberrant semantic/conceptual processing or abnormal selfother distinction. As in our analyses, a recent report also demonstrated that the level of coherence in some pairs of EEG signals was inversely related to psychosis in schizophrenia (Bob et al., 2008). Overall, the topography illustrated in this study highlights the brain circuits indicating vulnerability of psychosis in a depressive state, with both commonality and uniqueness present when compared with other psychotic disorders. It is noteworthy that our symptom-cluster approach, in contrast to that of a single symptom, may provide better correspondence with neurological/physiological entity. In addition, our analyses demonstrated a one-to-many relationship between psychosis/cognitive distortion and inter-hemispheric/frontotemporo-parietal network. In summary, the defect of a single neural circuit might contribute to several symptoms, and a single symptom manifestation might engage several different neural circuits. Our negative results regarding core depression and psychic anxiety are compatible with extant literature which has addressed the roles of subcortical and deep cortical structures in a depressed mood and psychological fear/anxiety, such as the amygdala, peri-amygdaloid complex, hippocampal structure, orbitofrontal cortex and subgenual cingulate cortex (Bishop, 2007; Drevets et al., 2008; Kalia, 2005; Quirk and Mueller, 2008). A cortical mechanism, especially the prefrontal cortex, has been implied to regulate emotion through a cortico-limbic pathway or attentional modulation, but not to host the mood per se, which might explain the observed correlation between prefrontal activity and depression severity (Grimm et al., 2008; Taylor and Liberzon, 2007). Our negative results that cortical qEEG indices failed to differentiate sub-scores of activity and of sleep are also not surprising. Circadian rhythm is believed to be controlled by the hypothalamus (especially the suprachiasmatic nuclei) which receives efferent projection from the raphe nucleus (Cermakian and Boivin, 2009; Kalman and Kalman, 2009). The biological regulating factors of activity level are still not well understood, but evidence has suggested that the activity level is influenced by striatal dopamine level (Knab et al., 2009). Together with previous literature and our study, we propose a distinction of cortical and subcortical mechanisms of the manifestation of MDD: subcortical and deep cortical (notably limbic structures) structures are responsible for psychological depression/anxiety and vegetative dysfunctions, whereas both cortical and subcortical mechanisms are

relevant to the level of somatization and cognitive distortion/ delusion in a depressive state (Garcia-Campayo et al., 2001; Kunert et al., 2007; Tokunaga et al., 1997). 5. Conclusions and limitations To the best of our knowledge, this resting EEG study is the first to adopt a systemic investigation to clarify the cortical mechanisms of the main symptom constructs in major depression. This study delineates the cortical topography of vulnerability to somatic anxiety and cognitive distortion in a depressive state. Our findings suggest that psychic and somatic anxieties are distinct entities, with the latter associated with broad-range cortical involvement. It is thus worthwhile to explore somatic and psychic anxiety separately in the future. Since the number of imaging studies of somatoform disorders is limited, the neural patterns of somatic anxiety and of somatoform disorders warrant further studies to clarify. We suggest combining imaging facilities and evaluative tools for each symptom cluster to further our understanding of MDD. Our EEG recording with 128 Hz sampling and 20 electrodes may provide restricted spatial and temporal information. Higher sampling rate and electrode number are recommended to enable the assessment of higher frequency range and better spatial precision. In addition, our analysis at higher frequency range may be contaminated by electromyographical noise even though Hjorth's reference method has been applied to reduce the noise effect which might cause reference dependency (Pfurtscheller and Andrew, 1999; Whitham et al., 2007). Recent research suggested a discrepancy between structural and functional connectivity (Ponten et al., 2010). It is also worthwhile to attempt structural connectivity approach to elucidate its implication in depression symptoms. Role of funding source This work was supported by grant 95-2314-B-214-016-MY2 (EDPJ96011) from the E-DA Hospital, Kaohsiung County and grant KS92015 from the Kai-Suan Psychiatric Hospital-Kaohsiung, Taiwan, ROC. The above funding bodies had no further role in the study design, in the collection, analysis and interpretation of data, in the writing of the report, and in the decision to submit the paper for publication.

Conflict of interest We have no professional, personal or financial affiliations that may be perceived to have biased the presentation.

Acknowledgements We thank Mr. Higgins who assisted with the preparation and proof-reading of the manuscript. We are also grateful to the helpful comments by two anonymous reviewers. References Aftanas, L.I., Pavlov, S.V., 2005. Characteristics of interhemispheric EEG band power distribution in high anxiety individuals under emotionally neutral and aversive arousal conditions. Zh. Vyssh. Nerv. Deiat. Im. I. P. Pavlova 55, 322–328. Anand, A., Li, Y., Wang, Y., Wu, J., Gao, S., Bukhari, L., Mathews, V.P., Kalnin, A., Lowe, M.J., 2005. Activity and connectivity of brain mood regulating circuit in depression: a functional magnetic resonance study. Biol. Psychiatry 57, 1079–1088.

T.-W. Lee et al. / Journal of Affective Disorders 131 (2011) 243–250 American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV). American Psychiatric Association, Washington, DC. Bae, J.N., MacFall, J.R., Krishnan, K.R., Payne, M.E., Steffens, D.C., Taylor, W.D., 2006. Dorsolateral prefrontal cortex and anterior cingulate cortex white matter alterations in late-life depression. Biol. Psychiatry 60, 1356–1363. Bishop, S.J., 2007. Neurocognitive mechanisms of anxiety: an integrative account. Trends Cogn. Sci. 11, 307–316. Blackwood, N.J., Howard, R.J., ffytche, D.H., Simmons, A., Bentall, R.P., Murray, R.M., 2000. Imaging attentional and attributional bias: an fMRI approach to the paranoid delusion. Psychol. Med. 30, 873–883. Bob, P., Palus, M., Susta, M., Glaslova, K., 2008. EEG phase synchronization in patients with paranoid schizophrenia. Neurosci. Lett. 447, 73–77. Brody, A.L., Barsom, M.W., Bota, R.G., Saxena, S., 2001. Prefrontal–subcortical and limbic circuit mediation of major depressive disorder. Semin. Clin. Neuropsychiatry 6, 102–112. Cermakian, N., Boivin, D.B., 2009. The regulation of central and peripheral circadian clocks in humans. Obes. Rev. 10 (Suppl 2), 25–36. Charlton, R.A., Barrick, T.R., Lawes, I.N., Markus, H.S., Morris, R.G., 2010. White matter pathways associated with working memory in normal aging. Cortex 46, 474–489. Damoiseaux, J.S., Beckmann, C.F., Arigita, E.J., Barkhof, F., Scheltens, P., Stam, C.J., Smith, S.M., Rombouts, S.A., 2008. Reduced resting-state brain activity in the "default network" in normal aging. Cereb. Cortex 18, 1856–1864. David, A.S., 1994. Schizophrenia and the corpus callosum: developmental, structural and functional relationships. Behav. Brain Res. 64, 203–211. Dawson, M.E., Schell, A.M., Hazlett, E.A., Nuechterlein, K.H., Filion, D.L., 2000. On the clinical and cognitive meaning of impaired sensorimotor gating in schizophrenia. Psychiatry Res. 96, 187–197. Dimsdale, J.E., Dantzer, R., 2007. A biological substrate for somatoform disorders: importance of pathophysiology. Psychosom. Med. 69, 850–854. Drevets, W.C., Price, J.L., Furey, M.L., 2008. Brain structural and functional abnormalities in mood disorders: implications for neurocircuitry models of depression. Brain Struct. Funct. 213, 93–118. Fingelkurts, A.A., Fingelkurts, A.A., Rytsala, H., Suominen, K., Isometsa, E., Kahkonen, S., 2007. Impaired functional connectivity at EEG alpha and theta frequency bands in major depression. Hum. Brain Mapp. 28, 247–261. Garcia-Campayo, J., Sanz-Carrillo, C., Baringo, T., Ceballos, C., 2001. SPECT scan in somatization disorder patients: an exploratory study of eleven cases. Aust. N. Z. J. Psychiatry 35, 359–363. Goodwin, G.M., 1997. Neuropsychological and neuroimaging evidence for the involvement of the frontal lobes in depression. J. Psychopharmacol. 11, 115–122. Grimm, S., Beck, J., Schuepbach, D., Hell, D., Boesiger, P., Bermpohl, F., Niehaus, L., Boeker, H., Northoff, G., 2008. Imbalance between left and right dorsolateral prefrontal cortex in major depression is linked to negative emotional judgment: an fMRI study in severe major depressive disorder. Biol. Psychiatry 63, 369–376. Hakala, M., Karlsson, H., Kurki, T., Aalto, S., Koponen, S., Vahlberg, T., Niemi, P.M., 2004. Volumes of the caudate nuclei in women with somatization disorder and healthy women. Psychiatry Res. 131, 71–78. Hakala, M., Vahlberg, T., Niemi, P.M., Karlsson, H., 2006. Brain glucose metabolism and temperament in relation to severe somatization. Psychiatry Clin. Neurosci. 60, 669–675. Hamilton, J.P., Gotlib, I.H., 2008. Neural substrates of increased memory sensitivity for negative stimuli in major depression. Biol. Psychiatry 63, 1155–1162. Higashima, M., Takeda, T., Kikuchi, M., Nagasawa, T., Koshino, Y., 2006. Functional connectivity between hemispheres and schizophrenic symptoms: a longitudinal study of interhemispheric EEG coherence in patients with acute exacerbations of schizophrenia. Clin. EEG Neurosci. 37, 10–15. Hjorth, B., 1975. An on-line transformation of EEG scalp potentials into orthogonal source derivations. Electroencephalogr. Clin. Neurophysiol. 39, 526–530. Kalia, M., 2005. Neurobiological basis of depression: an update. Metabolism 54, 24–27. Kalman, J., Kalman, S., 2009. Depression as chronobiological illness. Neuropsychopharmacol. Hung. 11, 69–81. Kim, D.K., Kim, B.L., Sohn, S.E., Lim, S.W., Na, D.G., Paik, C.H., Krishnan, K.R., Carroll, B.J., 1999. Candidate neuroanatomic substrates of psychosis in old-aged depression. Prog. Neuropsychopharmacol. Biol. Psychiatry 23, 793–807. Knab, A.M., Bowen, R.S., Hamilton, A.T., Gulledge, A.A., Lightfoot, J.T., 2009. Altered dopaminergic profiles: implications for the regulation of voluntary physical activity. Behav. Brain Res. 204, 147–152.

249

Konarski, J.Z., McIntyre, R.S., Kennedy, S.H., Rafi-Tari, S., Soczynska, J.K., Ketter, T.A., 2008. Volumetric neuroimaging investigations in mood disorders: bipolar disorder versus major depressive disorder. Bipolar Disord. 10, 1–37. Korb, A.S., Cook, I.A., Hunter, A.M., Leuchter, A.F., 2008. Brain electrical source differences between depressed subjects and healthy controls. Brain Topogr. 21, 138–146. Kunert, H.J., Norra, C., Hoff, P., 2007. Theories of delusional disorders. update review. Psychopathol. 40, 191–202. Laufs, H., Kleinschmidt, A., Beyerle, A., Eger, E., Salek-Haddadi, A., Preibisch, C., Krakow, K., 2003. EEG-correlated fMRI of human alpha activity. Neuroimage 19, 1463–1476. Lawrie, S.M., Buechel, C., Whalley, H.C., Frith, C.D., Friston, K.J., Johnstone, E.C., 2002. Reduced frontotemporal functional connectivity in schizophrenia associated with auditory hallucinations. Biol. Psychiatry 51, 1008–1011. MacDonald 3rd, A.W., Carter, C.S., 2003. Event-related FMRI study of context processing in dorsolateral prefrontal cortex of patients with schizophrenia. J. Abnorm. Psychol. 112, 689–697. Maletic, V., Robinson, M., Oakes, T., Iyengar, S., Ball, S.G., Russell, J., 2007. Neurobiology of depression: an integrated view of key findings. Int. J. Clin. Pract. 61, 2030–2040. Meyer, J.H., 2008. Applying neuroimaging ligands to study major depressive disorder. Semin. Nucl. Med. 38, 287–304. Mitelman, S.A., Buchsbaum, M.S., Brickman, A.M., Shihabuddin, L., 2005. Cortical intercorrelations of frontal area volumes in schizophrenia. Neuroimage 27, 753–770. Oquendo, M.A., Placidi, G.P., Malone, K.M., Campbell, C., Keilp, J., Brodsky, B., Kegeles, L.S., Cooper, T.B., Parsey, R.V., van Heertum, R.L., Mann, J.J., 2003. Positron emission tomography of regional brain metabolic responses to a serotonergic challenge and lethality of suicide attempts in major depression. Arch. Gen. Psychiatry 60, 14–22. Pae, C.U., Choe, B.Y., Joo, R.H., Lim, H.K., Kim, T.S., Yoo, S.S., Choi, B.G., Kim, J.J., Lee, S.J., Lee, C., Paik, I.H., Lee, C.U., 2004. Neuronal dysfunction of the frontal lobe in schizophrenia. Neuropsychobiology 50, 211–215. Pfurtscheller, G., Andrew, C., 1999. Event-related changes of band power and coherence: methodology and interpretation. J. Clin. Neurophysiol. 16, 512–519. Pollock, V.E., Schneider, L.S., 1990. Quantitative, waking EEG research on depression. Biol. Psychiatry 27, 757–780. Ponten, S.C., Daffertshofer, A., Hillebrand, A., Stam, C.J., 2010. The relationship between structural and functional connectivity: graph theoretical analysis of an EEG neural mass model. Neuroimage 52, 985–994. Quirk, G.J., Mueller, D., 2008. Neural mechanisms of extinction learning and retrieval. Neuropsychopharmacology 33, 56–72. Ragland, J.D., Yoon, J., Minzenberg, M.J., Carter, C.S., 2007. Neuroimaging of cognitive disability in schizophrenia: search for a pathophysiological mechanism. Int. Rev. Psychiatry 19, 417–427. Rajarethinam, R.P., DeQuardo, J.R., Nalepa, R., Tandon, R., 2000. Superior temporal gyrus in schizophrenia: a volumetric magnetic resonance imaging study. Schizophr. Res. 41, 303–312. Rusalova, M.N., Kostyunina, M.B., 2004. Spectral correlation studies of emotional states in humans. Neurosci. Behav. Physiol. 34, 803–808. Sambataro, F., Murty, V.P., Callicott, J.H., Tan, H.Y., Das, S., Weinberger, D.R., Mattay, V.S., 2010. Age-related alterations in default mode network: impact on working memory performance. Neurobiol. Aging 31, 839–852. Serretti, A., Cusin, C., Lattuada, E., Di Bella, D., Catalano, M., Smeraldi, E., 1999. Serotonin transporter gene (5-HTTLPR) is not associated with depressive symptomatology in mood disorders. Mol. Psychiatry 4, 280–283. Serretti, A., Lattuada, E., Cusin, C., Macciardi, F., Smeraldi, E., 1998. Analysis of depressive symptomatology in mood disorders. Depress. Anxiety 8, 80–85. Simpson, S., Baldwin, R.C., Jackson, A., Burns, A., 1999. The differentiation of DSM-III-R psychotic depression in later life from nonpsychotic depression: comparisons of brain changes measured by multispectral analysis of magnetic resonance brain images, neuropsychological findings, and clinical features. Biol. Psychiatry 45, 193–204. Smets, M., Gallois, P., Forzy, G., Dereux, J.F., Parquet, P.J., 1992. Correlations between delayed auditory evoked potentials and clinical evaluation of schizophrenic symptoms. Ann. Med. Psychol. Paris 150, 215–219. Stahl, S.M., Zhang, L., Damatarca, C., Grady, M., 2003. Brain circuits determine destiny in depression: a novel approach to the psychopharmacology of wakefulness, fatigue, and executive dysfunction in major depressive disorder. J. Clin. Psychiatry 64 (Suppl 14), 6–17. Strelets, V.B., Danilova, N.N., Kornilova, I.V., 1997. The EEG rhythms and psychological indices of emotions in reactive depression. Zh. Vyssh. Nerv. Deiat. Im. I. P. Pavlova 47, 11–21. Taylor, S.F., Liberzon, I., 2007. Neural correlates of emotion regulation in psychopathology. Trends Cogn. Sci. 11, 413–418.

250

T.-W. Lee et al. / Journal of Affective Disorders 131 (2011) 243–250

Tokunaga, M., Ida, I., Higuchi, T., Mikuni, M., 1997. Alterations of benzodiazepine receptor binding potential in anxiety and somatoform disorders measured by 123I-iomazenil SPECT. Radiat. Med. 15, 163–169. Vasic, N., Walter, H., Hose, A., Wolf, R.C., 2008. Gray matter reduction associated with psychopathology and cognitive dysfunction in unipolar depression: a voxel-based morphometry study. J. Affect. Disord. 109, 107–116. Whitham, E.M., Pope, K.J., Fitzgibbon, S.P., Lewis, T., Clark, C.R., Loveless, S., Broberg, M., Wallace, A., DeLosAngeles, D., Lillie, P., Hardy, A., Fronsko, R., Pulbrook, A., Willoughby, J.O., 2007. Scalp electrical recording during paralysis: quantitative evidence that EEG frequencies above 20 Hz are contaminated by EMG. Clin. Neurophysiol. 118, 1877–1888.

Wright, I.C., McGuire, P.K., Poline, J.B., Travere, J.M., Murray, R.M., Frith, C.D., Frackowiak, R.S., Friston, K.J., 1995. A voxel-based method for the statistical analysis of gray and white matter density applied to schizophrenia. Neuroimage 2, 244–252. Yamada, M., Kimura, M., Mori, T., Endo, S., 1995. EEG power and coherence in presenile and senile depression. Characteristic findings related differences between anxiety type retardation type. Nippon Ika Daigaku Zasshi 62, 176–185. Yu, Y.W., Tsai, S.J., Chen, T.J., Lin, C.H., Hong, C.J., 2002. Association study of the serotonin transporter promoter polymorphism and symptomatology and antidepressant response in major depressive disorders. Mol. Psychiatry 7, 1115–1119. Zacks, J.M., Michelon, P., 2005. Transformations of visuospatial images. Behav. Cogn. Neurosci. Rev. 4, 96–118.