Download Frontal EEG asymmetry and symptom response to cognitive

Biological Psychology 87 (2011) 379–385
Contents lists available at ScienceDirect
Biological Psychology
journal homepage: www.elsevier.com/locate/biopsycho
Frontal EEG asymmetry and symptom response to cognitive behavioral
therapy in patients with social anxiety disorder
David A. Moscovitch a,∗ , Diane L. Santesso b , Vladimir Miskovic c , Randi E. McCabe d ,
Martin M. Antony e , Louis A. Schmidt b,c
a
Canada Research Chair in Mental Health Research, Department of Psychology, University of Waterloo, 200 University Avenue West, Waterloo, Ontario, Canada N2L 3G1
Department of Psychology, Neuroscience & Behaviour, McMaster University, Canada
c
McMaster Integrative Neuroscience Discovery and Study, McMaster University, Canada
d
Anxiety Treatment and Research Centre, St. Joseph’s Healthcare Hamilton, and Department of Psychiatry and Behavioural Neurosciences, McMaster University, Canada
e
Department of Psychology, Ryerson University, Canada
b
a r t i c l e
i n f o
Article history:
Received 17 November 2010
Accepted 25 April 2011
Available online 13 May 2011
Keywords:
EEG asymmetry
Social anxiety
Social phobia
Cognitive behavioral therapy
Neurobiology
Frontal cortex
a b s t r a c t
Although previous studies have shown that socially anxious individuals exhibit greater relative right
frontal electroencephalogram (EEG) activity at rest, no studies have investigated whether improvements
in symptoms as a result of treatment are associated with concomitant changes in resting brain activity. Regional EEG activity was measured at rest in 23 patients with social anxiety disorder (SAD) before
and after cognitive behavioral therapy (CBT). Results indicated that patients shifted significantly from
greater relative right to greater relative left resting frontal brain activity from pre- to posttreatment.
Greater left frontal EEG activity at pretreatment predicted greater reduction in social anxiety from preto posttreatment and lower posttreatment social anxiety after accounting for pretreatment symptoms.
These relations were specific to the frontal alpha EEG asymmetry metric. These preliminary findings suggest that resting frontal EEG asymmetry may be a predictor of symptom change and endstate functioning
in SAD patients who undergo efficacious psychological treatment.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
Social anxiety disorder (SAD) is characterized by extreme fear
and avoidance of social and performance situations (American
Psychiatric Association, 2000). With a lifetime prevalence estimated to be as high as 12%, SAD is the third most common
psychological disorder (Kessler et al., 2005). SAD has an early
onset, a chronic course, and a debilitating impact on quality of
life in interpersonal, emotional, and occupational domains. Numerous controlled trials have demonstrated that cognitive behavioral
therapy (CBT) is efficacious in the treatment of SAD (Moscovitch
et al., 2009; Smits & Hofmann, 2009), with approximately 50–70%
of patients experiencing significant symptom reduction after 12
weeks of therapy (Swinson et al., 2006).
Despite the proven efficacy of CBT in reducing symptoms of
social anxiety, little is known about the impact of CBT interventions on the neural mechanisms underlying social anxiety. Imaging
studies on the neurobiological effects of CBT among patients with
depression as well as those with anxiety disorders have shown significant treatment-related changes across broadly distributed brain
∗ Corresponding author. Tel.: +1 519 888 4567x32549; fax: +1 519 746 8631.
E-mail address: [email protected] (D.A. Moscovitch).
0301-0511/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.biopsycho.2011.04.009
regions (Porto et al., 2009; Roffman et al., 2005). However, only
one previous study has investigated the association between symptom improvement during CBT and corresponding brain changes in
patients with a principal diagnosis of SAD (Furmark et al., 2002),
and this study used positron emission tomography (PET).
The search for neural mechanisms underlying CBT interventions
for SAD is of considerable theoretical and practical importance.
Intervention studies provide us with a means by which to examine
how environmental influences impact brain–behavior relations.
Studies may also be useful in terms of identifying aspects of
neuronal function at pretreatment that may predict endstate functioning, thereby leading to improvements in the prognosis and
treatment of SAD.
Of particular interest to the present study are models that link
frontal brain asymmetry to individual differences in affective style
and emotion processing. Davidson and Fox have theorized that the
frontal lobes are differentially involved in positive versus negative
affective states and corresponding motivated behaviors (Davidson,
2000; Fox, 1991), with left frontal areas of the brain mediating
the experience of positive emotions (e.g., joy, happiness, etc.) and
approach behaviors, and right frontal areas of the brain mediating the experience of negative emotions (e.g., fear, sadness, etc.)
and withdrawal behaviors. Moreover, patterns of frontal electroencephalogram (EEG) asymmetry may serve as an index of risk for
380
D.A. Moscovitch et al. / Biological Psychology 87 (2011) 379–385
a variety of emotion-related disorders, including depression and
anxiety. A number of studies over the past three decades have supported these theoretical predictions (see Coan and Allen, 2004).
Non-clinical samples of adults selected for high levels of shyness
(Schmidt, 1999) and social anxiety (Beaton et al., 2008), or clinically
diagnosed with SAD (Davidson et al., 2000) have been shown to
exhibit significant relative elevations in right frontal brain activity
when assessed during resting states or periods of acute emotional
provocation.
Although no studies to date have investigated patterns of change
in frontal EEG activity among individuals with SAD before and
after CBT, preliminary research has demonstrated that frontal EEG
asymmetry can be balanced following therapeutic interventions.
For example, in a recent randomized controlled study, patients
with posttraumatic stress disorder who received CBT exhibited
decreased right frontal EEG activity from pre- to posttreatment
relative to those assigned to a waitlist control condition (Rabe et al.,
2008).1 Across a number of other studies, significant modifications
of EEG alpha asymmetry (increased relative left or decreased relative right frontal EEG activity) have been observed in depressed
adolescents who received massage and music therapy (Jones and
Field, 1999), in healthy individuals (Davidson et al., 2003) as well as
those with a history of suicidal depression (Barnhofer et al., 2007)
exposed to an 8-week mindfulness meditation program, and in
depressed individuals undergoing repetitive transcranial magnetic
stimulation treatment (Funk and George, 2008; but see Spronk
et al., 2008). Moreover, the administration of acute cortisol and
prednisone, which generate anxiogenic effects, has been shown to
increase right frontal EEG alpha activity among healthy participants
(Schmidt et al., 1999; Tops et al., 2005). Finally, a study that examined frontal asymmetry in young adults at two time points 1 year
apart (Blackhart et al., 2006) found that greater relative right frontal
EEG resting activity at time 1 was significantly associated with
higher levels of trait anxiety (controlling for symptoms of depression) at time 2, suggesting that EEG asymmetry may predict the
development of future anxiety symptoms. Similarly, Beaton et al.
(2008) found that a nonclinical sample of socially anxious undergraduate students demonstrated greater relative right frontal EEG
activity at rest, but only after accounting for symptoms of depressed
mood.
We investigated whether CBT was associated with changes
in frontal EEG asymmetry in individuals with SAD and whether
pretreatment EEG had utility in terms of predicting posttreatment functioning. Patients with a principal diagnosis of generalized
SAD completed 12 sessions of standardized group CBT. At preand posttreatment, resting regional EEG activity was recorded as
part of a larger study (see Miskovic et al., 2011). We hypothesized that: (i) at pretreatment, SAD patients would exhibit
greater right than left frontal resting EEG activity; (ii) patients
would shift significantly from greater right to greater relative left frontal brain activity from pre- to posttreatment; and
(iii) greater pretreatment left frontal EEG asymmetry would
uniquely predict greater treatment-related decreases in symptoms of social anxiety. Given the documented links between
depression and frontal EEG asymmetry (e.g., Henriques and
Davidson, 1990; Schaffer et al., 1983), the high rate of comorbidity between SAD and depressive disorders (Brown et al., 2001),
and the significant degree of overlap observed in clinical samples
between symptoms of social anxiety and depression (Brown and
1
Other studies have reported a variety of treatment-related electrocortical effects
in patients with anxiety disorders other than SAD, including changes in eventrelated potentials during CBT for spider phobia (Leutgeb et al., 2009) and changes
in EEG gamma activity during CBT for generalized anxiety disorder (Oathes et al.,
2008).
Barlow, 2009), we accounted for symptoms of depression in our
analyses.
2. Method
2.1. Participants
Thirty three outpatients were recruited for this study. Eight of these individuals
discontinued treatment and their participation in the study. Two participants had
extreme scores on EEG ␣ spectral power (±3 SD from mean) and were thus excluded
from the analysis, leaving 23 (12 male) eligible Caucasian, right-handed participants
for the current analysis. All participants were assessed and treated at the Anxiety
Treatment and Research Centre (ATRC), an urban Canadian anxiety treatment clinic
(Hamilton, Ontario), and each received a principal DSM-IV-TR (1) diagnosis of SAD
(generalized subtype), as determined by trained clinicians on the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders – 4th
edition (SCID-IV; First et al., 2001). Clinical severity ratings (CSRs) ranged from 4 to
7 (M = 5.57, SD = .90), with CSRs of 4 and above representing significant interference
and distress associated with the principal diagnosis. Sixteen participants (69.6% of
the sample) received at least one current comorbid DSM-IV diagnosis secondary to
their SAD, with major depressive disorder (single episode or recurrent) and generalized anxiety disorder the most common comorbid diagnoses (n = 7 each), followed
by specific phobia (n = 6), and obsessive compulsive disorder (n = 5). Participants
ranged in age from 19 to 73 years (M = 35.74 years, SD = 15.0). Of the 23 patients, 15
had never been married, 5 were currently married, and 3 were previously married;
19 had a high school education or greater; 11 earned an income of $40,000 (CDN)
or more per year.
Exclusionary criteria included current mania, psychosis, significant suicidality, and/or organic brain disorders. Individuals with current substance
abuse/dependence were excluded if the patients could not agree to refrain from
using substances prior to and during treatment and experimental protocols; and/or
were deemed by the clinical team to be unsuitable for group CBT targeting SAD (2
participants were included in the present study with secondary diagnoses of alcohol
and cannabis dependence). Finally, participants were excluded if they consumed ␤blockers within 3 days of the EEG testing; and alcohol, marijuana or antihistamines
within 12 h of EEG testing. Patients were requested to maintain stable medication type and dosage throughout the study, from 1 month prior to their first EEG
assessment until after their final testing session and were required to report their
medication adherence at each testing and treatment session. As participants were
recruited from a community outpatient psychiatric clinic situated within a hospital,
65% of them were taking psychotropic medications at the beginning of treatment,
with an average of 2.79 (SD = 2.3) medications per patient. Some patients (39%) were
taking selective serotonin reuptake inhibitors (SSRI) alone or in combination with
another medication such as benzodiazepines and antipsychotic medications. Participants who reported taking psychotropic medications pro re nata (PRN, or “as
needed”) were also required to maintain dose stability throughout the course of
the study. Among all participants taking psychotropic medication, reported duration of medication treatment at their stable dose prior to their first study session
ranged from 12 to 56 weeks (M = 28.7 weeks). Four participants reported on their
weekly adherence forms that they had either begun a new psychotropic medication or increased their dose of a pre-existing prescription during the previous
week. Our data analytic approach with respect to these participants is described
below.
2.1.1. Clinical treatment procedures
Participants completed a 12-session standardized course of group CBT for SAD
at the ATRC in St. Joseph’s Healthcare in Hamilton, Ontario. Sessions lasted approximately 2 hours each and were administered by 2–3 qualified therapists, with 7–9
patients per group. Therapists followed a manualized protocol based on Antony and
Swinson (2008) in their delivery of treatment, which included psychoeducation, cognitive restructuring, in-session and between-session exposure exercises and social
skills training.
2.1.2. Experimental procedures
Participants were tested in the Child Emotion Laboratory at McMaster University. All laboratory procedures were approved by the participating university
and hospital ethics committees and written consent was obtained from participants prior to testing. This study was part of a larger one (Miskovic et al., 2011),
in which patients were assessed across four separate visits: two before, one during,
and one after treatment. For the purpose of the present study, we examined asymmetry at pretreatment (initial visit) and posttreatment only in order to determine
(a) whether CBT was associated with pre-to-post changes in frontal EEG asymmetry
and (b) whether pretreatment EEG had utility in terms of predicting posttreatment
symptoms and endstate functioning. Regional EEG was recorded for a 6-min resting period (in alternating 1-min eyes-open/eyes-closed segments) at both pre- and
posttreatment. Each participant’s pre- and posttreatment assessments were scheduled to occur within a window of approximately 2 weeks before and after the first
and final group CBT session.
D.A. Moscovitch et al. / Biological Psychology 87 (2011) 379–385
381
2.1.3. Measures
Clinical measures. At pre- and posttreatment, both an independent clinician blind
to the purpose of the study and patients’ treatment status and one of the group
therapists rated each participant’s illness severity on the Clinical Global Impression
Severity Scale (CGI; Guy, 1976). The CGI Severity Scale has been used in previous
research as a valid and reliable measure to evaluate the efficacy of CBT for SAD
(Zaider et al., 2003).
Self-report measures. At the end of each EEG session, we examined self-report
responses on the Social Phobia Inventory (SPIN; Connor et al., 2000) and the Beck
Depression Inventory – Second Edition (BDI-II; Beck et al., 1996) at pre- and posttreatment. The SPIN is a 17-item inventory measuring symptoms of social fear, avoidance
and arousal. The BDI-II is a 21-question self-report inventory for measuring the
severity of depression, and is composed of items relating to cognitive, emotional
and physical symptoms of depression. Change scores for the SPIN and BDI-II were
computed by subtracting posttreatment from pretreatment total scores. Using the
regression method, residual SPIN scores were also computed separately for pre- and
posttreatment by partialling out BDI-II scores from SPIN scores at each time point.
2.1.4. EEG recording and data reduction
EEG recording. EEG was recorded using a lycra stretch cap (Electro-Cap, Inc.,
Eaton, OH) with electrodes positioned according to the International 10/20 Electrode
Placement System (Jasper, 1958). Each electrode site was filled with a small amount
of electrolyte gel (Omni-prep) and gently abraded. Electrode impedances below
10 k per site were considered acceptable, as were impedances of up to 500 between homologous sites. EEG was recorded from F3/F4, C3/C4, P3/P4, O1/O2, but
only those from the left and right mid-frontal (F3, F4), central (C3, C4), and parietal
(P3, P4) regions are presented here. All electrodes were referenced to the central
vertex (Cz). EEG was amplified by individual SA Instrumentation Bioamplifiers. The
filters were set at 1 Hz (high pass) to 100 Hz (low pass), with a 60 Hz notch filter.
The data were digitized on-line at a sampling rate of 512 Hz.
EEG data reduction and analysis. The EEG data were visually scored for artifact
due to eye blinks, eye movements, and other motor movements when amplitudes
exceeded ±50 ␮V, using the software developed by James Long Company (EEG Analysis Program; Caroga Lake, NY). This program removes data from all channels if
artifact is present on any one channel. All artifact-free EEG data were analyzed using
a discrete Fourier transform (DFT), with a Hanning window of 1-s width and 50%
overlap of epochs. Power (microvolts squared) was derived from the DFT output
in three alpha frequency bands (alpha-I: 8–10 Hz; alpha-II: 10–13 Hz; full alpha:
8–13 Hz). Spectral power was estimated for the different subbands because previous work has demonstrated both the lower and higher Hz components of the alpha
band may have separate factor structures (Goncharova and Davidson, 1995) and
that each of these alpha bands may be related to shy and socially anxious behaviors
(Schmidt and Fox, 1994). A natural log (ln) transformation was performed on the EEG
power data to normalize the distributions. Resting EEG was collapsed across eyesopen and eyes-closed. Resting data from eyes-open are highly related to those from
eyes-closed and are typically averaged (Hagemann and Naumann, 2001). Indeed,
in the present study, pretreatment, r-values > .96, p-values < .001, and posttreatment, r-values > 0.82, p-values < .001, frontal alpha EEG data for the eyes open and
eyes-closed conditions were highly correlated. Separate EEG asymmetry scores were
computed for the mid-frontal, central, and parietal regions [e.g., ln(right) − ln(left)].
Because EEG alpha power is inversely related to scalp-recorded cortical activity,
positive asymmetry scores are thought to reflect greater relative left EEG cortical
activity (Allen et al., 2004a; Davidson, 1988).
3. Results
3.1. Effectiveness of CBT on symptom change
Paired t-tests (pre- versus posttreatment) on patients’ CGI
severity scores indicated that illness severity decreased significantly from pretreatment to posttreatment as rated by both the
independent, blind clinician, t(18) = 3.98, p = .001 (M = 5.05 ± 0.78
versus 4.37 ± 1.11), and the group therapist, t(22) = 6.10, p < .001
(M = 5.30 ± .82 versus 3.83 ± 1.15). As expected, patients’ selfreported social anxiety and depression symptoms at pre- versus
posttreatment also indicated a significant decrease in both the SPIN,
t(22) = 5.68, p < .001 (M = 45.74 ± 12.2 versus 30.30 ± 15.3) and BDIII, t(22) = 3.34, p = .003 (M = 24.30 ± 9.0 versus 17.26 ± 11.8) total
scores after receiving treatment.
Fig. 1. Mean alpha EEG asymmetry scores for each region during the (A) resting
condition (10–13 Hz). Error bars represent standard error of the mean. Negative
scores reflect greater relative right EEG activity. Note: * p < .05.
Table 1
Pearson correlations between resting frontal alpha (10–13 Hz) EEG asymmetry,
SPIN, and BDI-II scores at pre- and posttreatment.
Pretreatment SPIN
Pretreatment BDI-II
Residual pretreatment SPIN
Posttreatment SPIN
Posttreatment BDI-II
Residual pretreatment SPIN
Pre- minus posttreatment SPIN
Pre- minus posttreatment BDI-II
Pretreatment
resting asymmetry
Posttreatment
resting asymmetry
−.13
−.20
−.07
−.48**
−.47**
−.27
.44**
.40*
.18
−.02
.22
−.03
.11
−.10
.21
−.16
Note: BDI-II, Beck Depression Inventory, 2nd edition; SPIN, Social Phobia Inventory.
*
p = .06 (two-tailed).
**
p < .05 (two-tailed).
alpha II, and full alpha. For alpha II (i.e., 10–13 Hz), a significant Time
by Region interaction, F(2,44) = 3.85, p = .03, 2p = .15, indicated that
there was a change in EEG asymmetry from pre- to posttreatment
in the frontal region, t(22) = 2.60, p = .02, but not in the central,
t(22) = 0.10, p = .92, or parietal, [t(22) = 0.16, p = .13], regions (see
Fig. 1). As predicted, individuals displayed a shift from relatively
greater right to left frontal EEG asymmetry with treatment.2
3.3. Correlations between frontal EEG asymmetry and self-report
measures
Table 1 presents the results of two-tailed Pearson correlations
between resting EEG asymmetry and the self-report measures. As
predicted, pretreatment resting frontal alpha EEG asymmetry was
negatively related to posttreatment SPIN and BDI-II scores such
that greater relative left frontal resting EEG asymmetry at pretreatment was associated with lower social anxiety and depression
symptoms after treatment, as shown in Fig. 2A and B. Of particular
interest, results demonstrated that greater relative left frontal EEG
3.2. Resting EEG asymmetry
An ANOVA with Time (pretreatment, posttreatment) and Region
(frontal, central, parietal) was performed to examine changes in
resting EEG asymmetry before and after CBT separately for alpha I,
2
There was a Time by Region interaction on resting frontal EEG asymmetry in
the 8–13 Hz frequency band, F(2,42) = 6.95, p = .005, 2p = .25, but a follow-up paired
t-test indicated there was no significant change in frontal EEG asymmetry from preto posttreatment (p = .13).
382
D.A. Moscovitch et al. / Biological Psychology 87 (2011) 379–385
Fig. 2. Scatterplots for the Pearson correlations between pretreatment resting frontal alpha (10–13 Hz) EEG asymmetry and (A) posttreatment Social Phobia Inventory (SPIN)
total score, (B) posttreatment Beck Depression Inventory II (BDI-II) total score, (C) pre- minus posttreatment SPIN total score, and (D) pre- minus posttreatment BDI-II total
score. Negative scores reflect greater relative right EEG activity.
asymmetry at pretreatment was associated with greater decrease
(pre- minus posttreatment scores) in social anxiety and depression
symptoms during treatment, as shown in Fig. 2C and D.3
A hierarchical regression was performed to determine whether
pretreatment SPIN scores or resting frontal alpha EEG asymmetry
was the best predictor of posttreatment social anxiety symptoms.
Posttreatment SPIN total scores were entered as the dependent
variable, pretreatment BDI-II scores were entered in the first step,
and pretreatment SPIN scores and pretreatment resting frontal
alpha EEG asymmetry values were entered simultaneously in the
second step.4 Overall, the model accounted for 50.2% of the variance
in posttreatment SPIN scores, F(3,19) = 6.39, p = .004. Pretreatment
BDI-II scores did not account for a significant amount of variance
in posttreatment SPIN scores, R2 = 6.2, F(1,21) = 1.38, p = .25. After
accounting for depression, pretreatment SPIN scores and resting
frontal alpha EEG asymmetry accounted for a significant amount of
3
Excluding the four participants who reported increasing their psychotropic
medication during the study led to marginal reduction in the observed magnitude of
changes in frontal asymmetry from pre- to posttreatment, t(19) = 1.96, p = .06. Pretreatment frontal alpha asymmetry values were still significantly correlated with
posttreatment SPIN, r = −.51, p < .05, and BDI-II, r = −.46, p < .05, scores, and with
pre- minus posttreatment SPIN scores, r = .49, p < .05, but not pre- minus posttreatment BDI-II scores, r = .36, p = .12. All non-significant correlations reported in Table 1
remained non-significant.
4
The interaction between SPIN and BDI-II scores was not significant. Thus, it was
excluded from the final regression model.
variance in posttreatment SPIN scores, R2 = 44.1, F(2,19) = 8.40,
p = .002, with each variable accounting for a significant amount
variance (pretreatment SPIN: partial r = .58, semi-partial r = .50,
p = .01; pretreatment EEG asymmetry: partial r = −.51, semi-partial
r = −.42, p = .02).
We also examined the potential predictive utility of pretreatment resting frontal EEG asymmetry over and above patients’
pretreatment social anxiety symptoms. Accordingly, a similar
regression analysis was performed with pretreatment BDI-II scores
entered on the first step, pretreatment SPIN scores entered second
step, and pretreatment frontal alpha EEG asymmetry entered on the
third step. Pretreatment BDI-II scores did not significantly predict
posttreatment social anxiety symptoms, partial/semi-partial r = .25,
p = .25. Both pretreatment SPIN scores, partial r = .53, semi-partial
r = .52, p = .01, and pretreatment frontal alpha EEG asymmetry,
partial r = −.51, semi-partial r = −.42, p = .02, were still significant
predictors of posttreatment social anxiety symptoms.
4. Discussion
We investigated changes in the pattern of frontal EEG asymmetry among patients with SAD who completed 12 sessions of group
CBT. In conjunction with symptom amelioration from pre- to posttreatment, significant electrocortical changes occurred that were
characterized by a shift from greater relative right to left frontal
alpha EEG asymmetry at rest. These findings are consistent with
the growing literature demonstrating that frontal EEG asymmetry
D.A. Moscovitch et al. / Biological Psychology 87 (2011) 379–385
is sensitive to the effects of therapeutic interventions across various types of samples (e.g., Barnhofer et al., 2007; Davidson et al.,
2003; Jones and Field, 1999; Rabe et al., 2008).
At pretreatment, SAD patients exhibited greater relative right
frontal alpha EEG asymmetry at rest. This finding is consistent
with the predictions of existing EEG asymmetry–emotion models
(Davidson, 2000; Fox, 1991) and with empirical reports of elevated
right frontal brain activity during rest or emotional provocation
among nonclinical and clinical samples of socially anxious individuals (Beaton et al., 2008; Davidson et al., 2000; Schmidt, 1999).
Of particular interest, the present study demonstrated that the
nature and degree of symptom change that occurred as a result of
treatment could be predicted by pretreatment resting frontal EEG
asymmetry. Pretreatment frontal alpha EEG asymmetry (less rightsided or more left-sided activity) at rest was associated with greater
reductions in social anxiety from pre- to posttreatment and lower
posttreatment symptoms of social anxiety, even after accounting
for pretreatment symptoms of social anxiety and depression. The
strongest changes in resting frontal EEG asymmetry from pre- to
posttreatment were observed in analyses of the high alpha range
(i.e., 10–13 Hz). Although some have argued that alpha subbands
are structurally (Goncharova and Davidson, 1995) and functionally
(Klimesch, 1999) heterogeneous, most recent evidence does not
support the suggestion that the subbands are differentially sensitive to predicting variation in temperamental traits (Shackman
et al., 2010).
The frontal EEG asymmetry–emotion model would predict
that greater relative right frontal asymmetry would be related to
increased social anxiety symptoms at pretreatment among patients
with SAD. However, pretreatment resting frontal EEG asymmetry
in the present study was not related to pretreatment symptoms
of social anxiety. There are several possible explanations for this
apparently puzzling finding. One important factor may have been
the restricted range of SPIN scores among SAD patients at pretreatment, with most participants’ scores clustering between 40 and 60.
Conversely, pretreatment BDI-II scores, which were significantly
correlated with greater pretreatment right frontal EEG asymmetry,
were distributed with greater variance in the sample (spread evenly
across participants between 10 and 40). However, this explanation is limited by the finding that posttreatment asymmetry was
also not significantly associated with posttreatment anxiety symptoms, despite the less restricted range of SPIN scores across the
study sample at the completion of therapy. A second possibility
is that although both the SPIN and BDI-II are considered valid
and reliable measures in so far as they adequately capture participants’ core negative emotional experiences associated with social
anxiety and depression, respectively, the SPIN (unlike the BDI-II)
might not effectively measure the type of avoidance or withdrawal
behavior that is purportedly related to EEG measures of frontal
asymmetry (see Coan and Allen, 2004). Indeed, previous studies
have also failed to support hypotheses and find significant associations between frontal EEG asymmetry and self-report measures
of behavioral inhibition (Coan and Allen, 2003; Harmon-Jones and
Allen, 1997). Investigators of these studies have concluded that the
relation between EEG asymmetry, withdrawal behavior, and negative affectivity is likely to be multifaceted and complex (see Coan
and Allen, 2003).
It is possible that the observed predictive relation between
increased relative left hemisphere activity at pretreatment and
enhanced CBT response might reflect superior left-sided verbal
processing abilities that better support the core verbal therapeutic techniques used in CBT, such as cognitive restructuring (see
for example Bruder et al., 1997). Unfortunately, we did not assess
patients’ verbal skills in order to test this hypothesis directly, nor is
it clear why such effects would be confined to left frontal EEG activity and not extend to central and even parietal areas. An alternative
383
interpretation is the putative involvement of the left orbital frontal
cortex in evaluating the affective salience of stimuli as well as the
extinction of conditioned fear responses. Consistent with this suggestion, Brody et al. (1998) reported that patients with OCD who
had higher pretreatment levels of activity in their left orbital frontal
cortex ultimately responded best to behavioral therapy.
Several limitations of the present study should be noted. First
and foremost, the study design did not include a comparison
group of clinical or healthy controls, or an alternate treatment or
waitlist condition against which the effects observed in patients
with SAD who underwent CBT could be evaluated. Previous wellcontrolled studies have clearly established that individuals with
SAD experience significant clinical gains during CBT that are not
observed among patients assigned to corresponding waitlist conditions (Ponniah and Hollon, 2000), and that resting frontal EEG
alpha asymmetry in adult clinical samples remain relatively stable over significant time intervals in the absence of treatment (e.g.,
Allen et al., 2004b; Jetha et al., 2009; Vuga et al., 2006). Nevertheless, because we did not include a control group in our design,
we cannot conclude definitively that changes in alpha asymmetry
were due to the specific effects of CBT per se. Additional research
is required before it is possible to infer causality in the observed
relation between the effects of CBT and changes in frontal EEG
asymmetry.
Second, as is common in outpatient psychiatric settings, the
majority of CBT participants in the present study were also taking
medication. As a result, we were unable to disentangle whether
changes in EEG asymmetry were associated specifically with the
effects of CBT, the effects of medication, or the combined effects of
therapy and medication. It is possible that some combination of CBT
and the type and/or dosage of medication were related to treatment
outcome and/or the link between pre-treatment EEG and treatment
response. The literature provides no clear indication of the relative efficacy of CBT plus medication versus CBT alone in the acute
treatment of SAD, as findings from clinical trials examining this
question have been mixed (see Blanco et al., 2010; Davidson et al.,
2004; Haug et al., 2003; Pontoski and Heimberg, 2010). Although
well-controlled future studies are needed to determine whether
differential changes in frontal EEG asymmetry occur in patients
with SAD as a result of “pure” psychological versus pharmacological treatment, it is important to note that findings from studies
based on such highly selected samples of patients undergoing tailored treatment protocols may not generalize readily to the types
of patients typically treated in most clinical practice settings.
Third, it is important to note that our study was limited by
its reliance on a purely subjective measure of medication adherence. Using this measure, a small minority of participants (n = 4)
reported that they increased their medication during the study.
However, analyses that excluded these participants from the overall sample produced only slight attenuations in the strength of our
results, which remained relatively robust and in the same direction as the observed effects when all participants were included
in the analyses. The slight attenuation in the strength of our
results in the absence of these participants may suggest that the
observed changes in frontal asymmetry were driven by the combined therapeutic effects of both medication and CBT, or may
simply reflect a decrease in statistical power related to removing four participants from a relatively small sample.A prospect
for future research involves determining the extent to which the
present results are driven by characteristics that are unique to
social anxiety and CBT per se. An interesting question concerns
the extent to which similar effects might be observed in other
types of affective psychopathology and in response to other types
of interventions specifically designed to target and reduce negative
affect and behavioral avoidance. Given the established association
between individual differences in frontal EEG asymmetry, on one
384
D.A. Moscovitch et al. / Biological Psychology 87 (2011) 379–385
hand, and broadly measured indicators of negative versus positive
affect and approach–withdrawal behaviors, on the other (Coan and
Allen, 2004), it is likely that the pattern of results observed here
would generalize from SAD to related emotional disorders that are
also characterized by high levels of emotional distress and withdrawal behaviors (e.g., generalized anxiety disorder, depression,
etc.), and from CBT to any intervention that is designed specifically
to reduce these emotional and behavioral symptoms.
Notwithstanding its limitations and the need for future replication and extension of the preliminary findings, our study suggests
that the frontal alpha EEG asymmetry metric may be a useful assessment tool to help clinicians gauge how patients might respond
to treatment. While it is too early to conclude that frontal EEG
asymmetry is a biological marker of affective psychopathology and
has any prognostic value for its treatment, our findings are consistent with the view that the effects of successful treatment for
SAD can be measured biologically and that symptom reduction is
associated with reliable changes in EEG asymmetry. Future studies are needed to determine the sensitivity and specificity of EEG
asymmetry in predicting individual differences in treatment outcome. Such studies would represent a crucial next step toward
enhancing our understanding of how EEG asymmetry might be
related to symptom changes for individual patients, moving beyond
the group-level correlations reported here. With such additional
supporting evidence, future innovations in CBT treatment might
incorporate EEG asymmetry measures in the routine assessment of
emotional disorders, leading potentially to more precise diagnosis
and the development of interventions customized to the specific
clinical needs and pretreatment symptom profiles of individual
patients.
Acknowledgements
This research was supported by an Ontario Mental Health Foundation New Investigator Fellowship and funding from the Canada
Research Chairs Program awarded to David A. Moscovitch, and by
NSERC and SSHRC Operating Grants awarded to Louis A. Schmidt.
We are grateful to Jessica Senn, Sue McKee, and to the clinicians
and staff at the Anxiety Treatment and Research Centre, for their
invaluable contributions to this study.
References
Allen, J.J.B., Coan, J.A., Nazarian, M., 2004a. Issues and assumptions on the road from
raw signals to metrics of frontal asymmetry in emotion. Biological Psychology
67, 183–218.
Allen, J.J.B., Urry, H.L., Hitt, S.K., Coan, J.A., 2004b. The stability of resting frontal electroencephalographic asymmetry in depression. Psychophysiology 41, 269–280.
American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental
Disorders (fourth ed. Test–Revised). Author, Washington, DC.
Antony, M.M., Swinson, R.P., 2008. The Shyness and Social Anxiety Workbook:
Proven, Step-By-Step Techniques for Overcoming Your Fear, second ed. New
Harbinger, Oakland, CA.
Barnhofer, T., Duggan, D., Crane, C., Hepburn, S., Fennell, M., Williams, J., 2007. Effects
of meditation on frontal alpha-asymmetry in previously suicidal individuals.
Neuroreport 18, 709–712.
Beaton, E.A., Schmidt, L.A., Ashbaugh, A.R., Santesso, D.L., Antony, M.M., McCabe,
R.E., 2008. Resting and reactive frontal brain electrical activity (EEG) among a
non-clinical sample of socially anxious adults: does concurrent depressive mood
matter? Neuropsychiatric Disease and Treatment 4, 187–192.
Beck, A.T., Steer, R.A., Brown, G.K., 1996. Beck Depression Inventory: Manual, second
ed. Pearson Assessment, San Antonio, TX.
Blackhart, G.C., Minnix, J.A., Kline, J.P., 2006. Can EEG asymmetry patterns predict
future development of anxiety and depression? A preliminary study. Biological
Psychology 72, 46–50.
Blanco, C., Olfson, M., Stein, D.J., Simpson, H.B., Gameroff, M.J., Narrow, W.H., 2010.
Treatment of obsessive–compulsive disorder by U.S. psychiatrists. Journal of
Clinical Psychiatry 67, 946–951.
Brody, A.L., Saxena, S., Schwartz, J.M., Stoessel, P.W., Maidment, K., Phelps, M.E.,
Baxter, L.R., 1998. FDG-PET predictors of response to behavioral therapy and
pharmacotherapy in obsessive compulsive disorder. Psychiatry Research 84,
1–6.
Brown, T.A., Barlow, D.H., 2009. A proposal for a dimensional classification system based on the shared features of the DSM-IV anxiety and mood disorders:
Implications for assessment and treatment. Psychological Assessment 21, 256–
271.
Brown, T.A., Campbell, L.A., Lehman, C.L., Grisham, J.R., Mancill, R.B., 2001. Current
and lifetime comorbidity of the DSM-IV anxiety and mood disorders in a large
clinical sample. Journal of Abnormal Psychology 110, 585–599.
Bruder, G.E., Stewart, J.W., Mercier, M., Agosti, V., Leite, P., Donovan, S., Quitkin,
F.M., 1997. Outcome of cognitive–behavioral therapy for depression: relation to
hemispheric dominance for verbal processing. Journal of Abnormal Psychology
106, 138–144.
Coan, J.A., Allen, J.J.B., 2003. Frontal EEG asymmetry and the behavioral activation
and inhibition systems. Psychophysiology 40, 106–114.
Coan, J.A., Allen, J.J.B., 2004. Frontal EEG asymmetry as a moderator and mediator of
emotion. Biological Psychology 67, 7–49.
Connor, K.M., Davidson, J.R.T., Churchill, L.E., Sherwood, A., Foa, E., Wesler, R.H., 2000.
Psychometric properties of the Social Phobia Inventory (SPIN). British Journal of
Psychiatry 176, 379–386.
Davidson, R.J., 1988. EEG measures of cerebral asymmetry: conceptual and methodological issues. International Journal of Neuroscience 39, 71–89.
Davidson, R.J., 2000. Affective style, psychopathology, and resilience: brain mechanisms and plasticity. American Psychologist 55, 1196–1214.
Davidson, R.J., Kabat-Zinn, J., Schumacher, J., Rosenkranz, M., Muller, D., Santorelli,
S.F., Urbanowski, F., Harrington, A., Bonus, K., Sheridan, J.F., 2003. Alterations
in brain and immune function produced by mindfulness meditation. Psychosomatic Medicine 65, 564–570.
Davidson, R.J., Marshall, J.R., Tomarken, A.J., Henriques, J.B., 2000. While a phobic
waits: regional brain electrical and autonomic activity in social phobics during
anticipation of public speaking. Biological Psychiatry 47, 85–95.
Davidson, J.R.T., Foa, E.D., Huppert, J.D., Keefe, F.J., Franklin, M.E., Compton, J.S., Zhao,
N., Connor, K.M., Lynch, T.R., Gadde, K.M., 2004. Fluoxetine, comprehensive cognitive behavioral therapy, and placebo in generalized social phobia. Archives of
General Psychiatry 61, 1005–1013.
First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 2001. Structured Clinical Interview for DSM-IV-TR Axis I Disorders – Patient Edition (SCID-I/P, 2/2001 revision).
NY Biometrics Research, New York State Psychiatric Institute, New York,
NY.
Fox, N.A., 1991. If it’s not left, it’s right: electroencephalograph asymmetry and the
development of emotion. American Psychologist 46, 863–872.
Funk, A.P., George, M.S., 2008. Prefrontal EEG asymmetry as a potential biomarker
of antidepressant treatment response with transcranial magnetic stimulation
(TMS): a case series. Clinical EEG and Neuroscience 39, 125–130.
Furmark, T., Tillfors, M., Marteinsdottir, I., Fischer, H., Langstrom, B., Fredrikson, M.,
2002. Common changes in cerebral blood flow in patients with social phobia
treated with citalopram or cognitive–behavioral therapy. Archives of General
Psychiatry 59, 425–433.
Goncharova, I.I., Davidson, R.J., 1995. The factor structure of EEG-differential validity
of low, high alpha power asymmetry in predicting affective style. Psychophysiology 32, S35.
Guy, W. (Ed.), 1976. ECDEU Assessment for Psychopharmacology. , Revised edition.
NIMH Publication, Rockville, MD.
Hagemann, D., Naumann, E., 2001. The effects of ocular artifacts on (lateralized)
broadband power in the EEG. Clinical Neurophysiology 112, 215–231.
Harmon-Jones, E., Allen, J.J.B., 1997. Behavioral activation sensitivity and resting
frontal EEG asymmetry: covariation of putative indicators related to risk for
mood disorders. Journal of Abnormal Psychology 106, 159–163.
Haug, T.T., Blomhoff, S., Hellstrom, K., Holme, I., Humble, M., Madsbu, H.P., Wold,
J.E., 2003. Exposure therapy and sertraline in social phobia: 1-year follow-up of
a randomised controlled trial. British Journal of Psychiatry 182, 312–318.
Henriques, J.B., Davidson, R.J., 1990. Regional brain electrical asymmetries discriminate between previously depressed and healthy control subjects. Journal of
Abnormal Psychology 99, 22–31.
Jasper, H.H., 1958. The ten–twenty electrode system of the International Federation.
Electroencephalography and Clinical Neurophysiology 10, 371–375.
Jetha, M.K., Schmidt, L.A., Goldberg, J.O., 2009. Long-term stability of resting frontal
EEG alpha asymmetry and power in a sample of stable community outpatients with schizophrenia. International Journal of Psychophysiology 72, 228–
233.
Jones, N.A., Field, T., 1999. Massage and music therapies attenuate frontal EEG asymmetry in depressed adolescents. Adolescence 34, 529–534.
Kessler, R.C., Berglund, P., Demler, O., Jin, R., Merikangas, K.R., Walters, E.E., 2005.
Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the
National Comorbidity Survey Replication. Archives of General Psychiatry 62,
593–602.
Klimesch, W., 1999. EEG alpha and theta oscillations reflect cognitive and memory
performance: a review and analysis. Brain Research Reviews 29, 169–195.
Leutgeb, V., Schäfer, A., Schienle, A., 2009. An event-related potential study on exposure therapy for patients suffering from spider phobia. Biological Psychology 82,
293–300.
Miskovic, V., Moscovitch, D.A., Santesso, D.L., McCabe, R.E., Antony, M.M., Schmidt,
L.A., 2011. Changes in EEG cross-frequency coupling during cognitive behavioral
therapy for social anxiety disorder. Psychological Science 22, 507–516.
Moscovitch, D.A., Antony, M.M., Swinson, R.P., 2009. Exposure-based treatments
for anxiety disorders: theory and process. In: Antony, M.M., Stein, M.B. (Eds.),
Handbook of Anxiety and Related Disorders. Oxford University Press, New York,
NY, pp. 461–475.
D.A. Moscovitch et al. / Biological Psychology 87 (2011) 379–385
Oathes, D.J., Ray, W.J., Yamasaki, A.S., Borkovec, T.D., Castonguay, L.G., Newman, M.G.,
Nitschke, J., 2008. Worry, generalized anxiety disorder, and emotion: evidence
from the EEG gamma band. Biological Psychology 79, 165–170.
Ponniah, K., Hollon, S.D., 2000. Empirically supported psychological interventions
for social phobia in adults: a qualitative review of randomized control trials.
Psychological Medicine 38, 3–14.
Pontoski, K.E., Heimberg, R.G., 2010. The myth of the superiority of concurrent combined treatments for anxiety disorders. Clinical Psychology: Science and Practice
17, 107–111.
Porto, P.R., Oliveira, L., Mari, J., Volchan, E., Figueira, I., Ventura, P., 2009. Does cognitive behavioral therapy change the brain? A systematic review of neuroimaging
in anxiety disorders. Journal of Neuropsychiatry and Clinical Neurosciences 21,
114–125.
Rabe, S., Zoellner, T., Beauducel, A., Maercker, A., Karl, A., 2008. Changes in brain
electrical activity after cognitive behavioral therapy for posttraumatic stress
disorder in patients injured in motor vehicle accidents. Psychosomatic Medicine
70, 13–19.
Roffman, J.L., Marci, C.D., Glick, D.M., Dougherty, D.D., Rauch, S.L., 2005. Neuroimaging and the functional neuroanatomy of psychotherapy. Psychological Medicine
35, 1385–1398.
Schaffer, C.E., Davidson, R.J., Saron, C., 1983. Frontal and parietal electroencephalogram asymmetry in depressed and non-depressed subjects. Biological
Psychiatry 18, 753–759.
Schmidt, L.A., 1999. Frontal brain electrical activity in shyness and sociability. Psychological Science 10, 316–320.
Schmidt, L.A., Fox, N.A., 1994. Patterns of cortical electrophysiology and autonomic
activity in adults’ shyness and sociability. Biological Psychology 38, 183–198.
385
Schmidt, L.A., Fox, N.A., Goldberg, M.C., Smith, C.C., Schulkin, J., 1999. Effects of acute
prednisone administration on memory, attention and emotion in healthy human
adults. Psychoneuroendocrinology 24, 461–483.
Shackman, A.J., McMenamin, B.W., Maxwell, J.S., Greischar, L.L., Davidson, R.J., 2010.
Identifying robust and sensitive frequency bands for interrogating neural oscillations. NeuroImage 51, 1319–1333.
Smits, J.A.J., Hofmann, S.G., 2009. A meta-analytic review of the effects of psychotherapy control conditions for anxiety disorders. Psychological Medicine 39,
229–239.
Spronk, D., Arns, M., Bootsma, A., van Ruth, R., Fitzgerald, P.B., 2008. Long-term
effects of left frontal rTMS on EEG and ERPs in patients with depression. Clinical
EEG and Neuroscience 39, 118–124.
Swinson, R.P., Antony, M.M., Bleau, P., Chokka, P., Craven, M., Fallu, A., Kjernisted, K.,
Lanius, R., Manassis, K., McIntosh, D., Plamondon, J., Rabheru, K., Van Ameringen, M., Walker, J.R., 2006. Clinical practice guidelines: management of anxiety
disorders. Canadian Journal of Psychiatry 51 (Suppl. 2), 1S–92S.
Tops, M., Wijers, A.A., van Staveren, A.S.J., Burin, K.J., Den Boer, J.A., Miejman, T.F., Korf, J., 2005. Acute cortisol administration modulates EEG alpha
asymmetry in volunteers: relevance to depression. Biological Psychology 69,
181–193.
Vuga, M., Fox, N.A., Cohn, J.F., George, C.J., Levenstein, R.M., Kovacs, M., 2006. Longterm stability of frontal electroencephalographic asymmetry in adults with a
history of depression and controls. International Journal of Psychophysiology
59, 107–115.
Zaider, T.I., Heimberg, R.G., Fresco, D.M., Schneier, F.R., Liebowitz, M.R., 2003. Evaluation of the Clinical Global Impression Scale among individuals with social
anxiety disorder. Psychological Medicine 33, 611–622.