Content uploaded by Jürgen Kayser
Author content
All content in this area was uploaded by Jürgen Kayser on Jun 07, 2016
Content may be subject to copyright.
Content uploaded by Jürgen Kayser
Author content
All content in this area was uploaded by Jürgen Kayser on Jun 07, 2016
Content may be subject to copyright.
In S. J. Luck & E. S. Kappenman (Eds.), The Oxford Handbook of Event-Related Potential Components (pp. 563-592)
New York: Oxford University Press © 2012.
Event-Related Brain Potentials in Depression:
Clinical, Cognitive and Neurophysiologic Implications
Gerard E. Bruder *, Jürgen Kayser, and Craig E. Tenke
Division of Cognitive Neuroscience, New York State Psychiatric Institute
and
Department of Psychiatry, Columbia University College of Physicians & Surgeons
Revised 10 March 2009
Introduction
Individuals having a depressive disorder commonly
experience difficulties in concentration, attention and other
cognitive functions, such as memory and executive control
(Austin et al., 2001; Porter et al., 2003). The recording of
event-related brain potentials (ERPs) provides a noninva-
sive means for studying cognitive deficits in depressive
disorders and their underlying neurophysiologic mecha-
nisms. The precise temporal resolution of ERPs can reveal
unique information about the specific stage of processing
that may lead to disruption of performance on cognitive
tasks, e.g., early sensory/attentional processing as reflected
in the N1 potential or later cognitive evaluation as reflected
in the P3 potential. Moreover, ERPs can provide non-
invasive biological markers for assessing treatment effects
and, most promisingly, for determining who will benefit
from a particular course of treatment.
By far, the largest number of ERP studies of depression
have focused on the cognitive P3 potential during target
detection “oddball” tasks. We will review the findings of
these studies and focus on recent studies that examined P3
subcomponents, which provide new evidence concerning
specific cognitive operations that may be disturbed in
depression. After reviewing these findings, we will examine
ERP findings in depressed patients obtained during more
challenging cognitive paradigms, including more demand-
ing auditory or visual discrimination tasks. We will also
review studies that have recorded ERPs in depressed
patients during recognition memory tasks, which provide
information on ERP correlates of episodic memory. Surpri-
singly few studies have measured ERPs of depressed
patients during processing of emotional stimuli, and yet,
such data may have particular relevance to mood disorders
and will therefore be reviewed. A number of recent studies
in depressed patients have found abnormalities of negative
brain potentials associated with monitoring of cognitive
performance, e.g., error-related negativity (ERN). These
studies, as well as others measuring the intensity-depen-
dency of auditory N1-P2 potentials, will be highlighted
because they suggest the potential value of these ERP
measures for predicting clinical response to antidepressants.
One aim of this review is therefore to bring together the
findings of studies measuring ERPs in depressed patients
during a variety of sensory, cognitive and emotional tasks,
so as to contribute toward a better understanding of the
specific processes and neurophysiologic mechanisms that
are dysfunctional in depressive disorders. For instance,
evidence of ERP abnormalities related to attentional or
cognitive control processes are suggestive of deficits
involving frontal or anterior cingulate cortex. Another aim
is to highlight the clinical relevance of ERP findings in
depressed patients by pointing to the relation of the patients’
ERPs to their clinical features, most notably severity of
depressive symptoms, diagnostic subtype, and therapeutic
response to treatments. From a more methodological per-
spective, we will present new findings illustrating the power
of combining current-source density (CSD) and principal
components analysis (PCA) techniques, which take better
advantage of both the temporal resolution of ERPs and the
spatial resolution of dense electrode arrays than traditional
analysis methods of reference-dependent surface potentials
(Kayser & Tenke, 2006a,b).
P3 in Auditory and Visual Oddball Tasks
The P3 or P300 potential provides physiologic measures
associated with attentional and working memory operations
during cognitive task performance (see Polich, 2007; Chap-
ter 7, this volume). It has typically been measured during
oddball tasks, in which a subject responds to an infrequent
target stimulus in a series of frequent nontarget standard
stimuli. In the typical study, subjects hear a pseudorandom
sequence of 90% low-pitched and 10% high-pitched tones,
each presented for 50 ms at a rate of 1 per second, and the
subject’s task is to respond to the infrequent high-pitched
tone (e.g., by pressing a button or silently counting). With
* Address reprint requests to: Gerard E. Bruder, New York State
Psychiatric Institute, Division of Cognitive Neuroscience, Unit 50, 1051
Riverside Drive, New York, NY 10032, USA. Email: bruderj@pi.cpmc.
columbia.edu
2G.E. Bruder, J. Kayser, C.E. Tenke
Figure 1. Grand mean, nose-referenced ERP waveforms for 26
healthy adults comparing targets (solid lines) and nontargets
(dashed lines) in an auditory oddball task at frontal (Fz), central
(Cz), parietal (Pz) and occipital (Oz) midline electrode sites (data
from Kayser et al., 1998).
all common EEG recording reference schemes (nose, linked
mastoids, average reference), the classical P3 potential
(P3b) is maximal over midline parietal scalp sites and has a
peak latency ranging from 300-500 ms. Figure 1 illustrates
the average waveforms for healthy adults at midline frontal
(Fz), central (Cz), parietal (Pz) and occipital (Oz) electrode
sites (nose reference) to infrequent targets (solid line) and
frequent nontargets (dashed line) in an oddball task. The
waveforms are typical of those seen for auditory oddball
tasks consisting of early N1 and P2 peaks to both targets
and nontargets, followed by a negative peak and a late
positive peak occurring about 200 ms (N2) and 350 ms (P3)
relative to the onset of only the target stimuli. The P3b
component has its maximum at Pz.
Most studies in depressed patients have used an auditory
oddball task. Although specific procedures vary from study
to study (e.g., frequency of target and nontarget tones,
stimulus duration, interstimulus intervals, response mode),
the use of the same basic task facilitates the comparison and
summary of P3 findings across studies. However, despite
the use of largely comparable oddball tasks, there have been
conflicting findings as to whether depressed patients have
reduced P3 amplitude. A review of early studies (Roth et
al., 1986) using mostly oddball tasks found that only about
half showed reduced P3 amplitude in depressed patients
when compared to healthy controls. Table 1 summarizes the
findings of more recent studies published over the last 20
years that compared P3 amplitudes for depressed patients
and healthy controls in auditory oddball tasks. Sixty percent
(12 of 20) of the comparisons listed in Table 1 found
significantly smaller P3 amplitude in patients having a
major depressive disorder (MDD) as compared to healthy
controls (HC). These studies had moderate to large effect
sizes, which ranged widely from 0.52 to 2.25 (Cohen’s d).
Among studies that failed to find significant differences,
there were often trends for depressed patients to have
smaller P3 than controls, but with small effect sizes ranging
from 0.11 to 0.52. The mean effect size of studies reported
in Table 1 is 0.85 (SD = 0.75; Median = 0.79), indicative of
a moderate group difference. Thus, while there continue to
be conflicting findings, the overall trend is for most studies
using an auditory oddball paradigm to show at least some
reduction of P3 amplitude in depressed patients.
The large difference in effect size across studies does,
however, suggest that differences in the clinical character-
istics of the patients in these studies may have played a role.
Although differences in P3 amplitude among patients have
generally not been found to be related to their overall
severity of depression, there is evidence that some subtypes
of depression show the greatest reductions of P3 amplitude.
All three studies testing patients having a major depression
with melancholic features found reduced P3 in patients,
with large effect sizes of 0.85, 0.98 and 2.25 (Ancy et al.,
1996; Gangadhar et al., 1993; Urretavizcaya et al., 2003).
Melancholic features include profound loss of interest or
pleasure, lack of reactivity to usual pleasurable stimuli, and
associated symptoms, such as early morning awakening,
worse in the morning, psychomotor retardation, weight loss
and excessive guilt (American Psychiatric Association,
1994). Also, P3 has been found to be more reduced in
patients having a psychotic than non-psychotic depression
(Karaaslan et al., 2003; Kaustio et al., 2002) and patients
who have attempted suicide compared to those without
suicidal history (Hansenne et al., 1996). Smaller P3
ERPs in depression 3
amplitude was associated with higher scores on scales for
assessing suicidal risk (Hansenne et al., 1996) and psychotic
symptoms (Santosh et al., 1994). Greater P3 reduction in
psychotic depression is consistent with evidence that cogni-
tive deficits on neuropsychological tests are more severe in
psychotic than nonpsychotic depression (Castaneda et al.,
2008) and with the robust P3 reduction seen in schizophre-
nia (Jeon & Polich, 2003; Chapter 18, this volume).
The patients in all but two of the studies in Table 1 were
unmedicated at the time of testing. Although one of these
studies found no difference in P3 between medicated and
unmedicated patients (Sara et al., 1994), studies have gene-
rally found P3 amplitude to increase or normalize following
treatment with antidepressants or ECT (Blackwood et al.,
1987; Gangadhar et al., 1993; Nurminen et al., 2005; but see
negative findings of Vandoolaeghe et al., 1998). Following
vagus nerve stimulation, treatment responders but not
nonresponders showed an increase in P3 amplitude, but no
control group was tested and so it is not known whether this
treatment normalized P3 (Neuhaus et al., 2007). These fin-
dings indicate that reduced P3 in depressed patients during
an auditory oddball task is at least partially state-dependent
and may normalize with improvement of depression during
successful treatment. This is also supported by the finding
that young women with a history of a major depressive epi-
sode, but no current depressive disorder, did not differ from
normal controls in P3 amplitude during an auditory oddball
task (Houston et al., 2004).
Fewer studies have measured P3 in depressed patients
during visual oddball tasks, and as is case for auditory
modality, there have been conflicting findings. Diner et al.
(1985) conducted one of the first studies, in which 10
depressed patients and 10 controls were tested in a variant
of a 3-stimulus visual oddball task with an infrequent target
(letter string ‘DTM’), a frequent standard (‘RSC’), and
infrequent nontarget three-letter words. P3 amplitude to
targets was significantly smaller in depressed patients when
compared to controls, and greater severity of depression
was associated with smaller P3. In contrast, Bange and
Bathien (1998) found no difference in P3 amplitude
between patients having either unipolar major depressive
disorder (n = 12) or bipolar depressive disorder (n = 11) and
healthy controls (n = 20) in either single-stimulus or two-
stimulus visual oddball tasks. They did, however, report that
patients having a bipolar depressive disorder had signifi-
cantly longer latency of the P3 peak compared to controls,
and the depressed patients showed a reduction in P3 latency
in remission. Although some studies have not found a
difference in P3 latency between depressed patients and
healthy controls in visual or auditory oddball tasks (Diner
et al., 1985; Blackwood et al., 1987; Gangadhar et al.,
1993), longer P3 latency in bipolar depressed patients, but
not unipolar depressed patients, parallels the findings of
Muir et al. (1991) for the auditory modality. This suggests
Table 1. Auditory Oddball Studies Comparing Depressed Patients and Healthy Controls.
Study Sample a EEG
Montage
EEG
Reference
P3 Amplitude Effect
Size b
Blackwood et al. (1987) 16 MDD (med-free), 59 HC Cz Left Ear MDD < HC .79
Muir et al. (1991) 46 MDD (35 med-free), 212 HC Cz Left Ear MDD < HC .52
Gangadhar et al. (1993) 17 MDD (med-free), 22 HC Cz Mastoids MDD < HC .98
Sara et al. (1994) 14 MDD (med-free), 27 HC
13 MDD (medicated)
Fz, Cz, Pz Linked Ears MDD = HC
MDD = HC
.18
.31
Hansenne et al. (1996) 10 MDDwS (med-free), 20 HC
10 MDDwoS (med-free)
Cz Left Ear MDDwS<HC
MDDwoS=HC
1.72
-.12
Ancy et al. (1996) 17 MDD (15 med-free), 15 HC Cz Mastoids MDD < HC .85
Yanai et al. (1997) 16 MDD (med-free), 17 HC Pz Linked Ears MDD < HC 2.18
Wagner et al. (1997) 11 MDD (med-free), 10 HC Fz, Cz Right Mastoid MDD < HC -
Bruder et al. (1998) 40 MDD/DYS (med-free), 22 HC 12 sites Nose MDD/DYS = HC -
Vandoolacghe et al. (1998) 35 MDD (med-free), 11 HC Cz Mastoids MDD = HC .52
Kaustio et al. (2002) 22 MDD/DYS (med-free), 22 HC 16 sites Right Mastoid MDD/DYS = HC -
Anderer et al. (2002) 60 MDD (med-free), 29 HC 19 sites Average Mastoids MDD < HC -
Röschke & Wagner (2003) 21 MDD (med-free), 21 HC Cz, Pz Right Mastoid MDD < HC -
Urretavizcaya et al. (2003) 50 MDD (med-free), 31 HC C3, Cz, C4 Linked Ears MDD < HC 2.25
Kaiser et al. (2003) 16 MDD (medicated), 16 HC 18 sites Average MDD = HC .11
Karaaslan et al. (2003) 16 MDDwP (med-free), 20 HC
20 MDDwoP (med-free)
Cz Linked Mastoids MDDwP < HC
MDDwoP = HC
1.43
.08
Kawasaki et al. (2004) 22 MDD (med-free), 22 HC 16 sites Linked Ears MDD < HC .90
aMDD = major depressive disorder; HC = healthy controls; DYS = dysthymic disorder; MDDwS = MDD with suicide attempt;
MDDwoS = MDD without suicide attempt; MDDwP = MDD with psychotic features; MDDwoP = MDD without psychotic features
bCohen’s d effect size
4G.E. Bruder, J. Kayser, C.E. Tenke
that patients who typically display psychomotor retardation,
e.g., those having bipolar or melancholic depressions, may
be most likely to show longer P3 latency suggestive of a
slowing of cognitive processing. Schlegel et al. (1991) also
found that longer P3 latency for depressed patients (n = 36)
in an auditory task was correlated with their total score on
the Bech-Rafaelsen Melancholia Scale and the four retarda-
tion items on this scale.
P3 in Cognitively Challenging Auditory and Visual Tasks
The conflicting findings for P3 amplitude in depressed
patients may in part be due to the use of simple oddball
tasks that are not cognitively challenging enough to elicit
robust P3 reductions in patients having subtle cognitive
deficits. We have argued that it would be more fruitful to
measure P3 in depressed patients during cognitively de-
manding tasks (Bruder, 1992). Given evidence from neuro-
psychological and dichotic listening tests suggestive of right
parietotemporal dysfunction in depression (Bruder et al.,
1989; Heller et al., 1995), we reasoned that depressed
patients might show greater P3 deficits in tasks that tap right
hemispheric processing, e.g., those involving spatial or
complex tonal processing. ERPs were measured in 25
unmedicated depressed patients and 27 healthy controls
during spatial and temporal discrimination tasks in the
auditory modality (Bruder et al., 1991). The spatial task
used a dichotic paradigm to manipulate the apparent
location of a click and the subject’s task was to discriminate
a difference in the location of standard and test stimuli. The
temporal task required discriminating a difference in dura-
tion of a standard click train and a test click train. A titration
procedure was used to determine the difference between
standard and test stimuli in each task that would yield 75%
correct responses for each subject, and these threshold
values were used during the ERP measurements. To eva-
luate differences between subtypes of depression, patients
were divided into those having either a typical, melancholic
form of depression or an atypical depression. Patients
meeting criteria for atypical depression showed symptoms
that are in some respects opposite of those seen for
melancholia, i.e., reactivity of mood with preserved plea-
sure capacity and one or more associated features – hyper-
somnia, overeating, rejection sensitivity or bodily inertia.
There was no difference among the patient subgroups and
healthy controls in behavioral thresholds for discriminating
stimuli in the spatial or temporal tasks, and no difference in
their P3 amplitudes. However, patients having a typical,
melancholic depression had considerably longer P3 latency
in the spatial task when compared to patients having
atypical depression and healthy controls. In contrast, there
was no difference among groups in P3 latency during the
temporal discrimination task, which indicates that the
cognitive task in which the P3 is measured is an important
factor. The melancholic subgroup showed evidence of a
slowing of cognitive processing only in the spatial task that
involves predominantly right hemisphere processing.
Moreover, it supports findings from oddball tasks suggest-
ing that longer P3 latency is most evident in specific diag-
nostic subtypes, i.e., melancholic and bipolar depression.
Given evidence of right hemisphere dysfunction in
depression, a subsequent study measured ERPs of 44
unmedicated depressed patients and 19 healthy controls
during a complex tone test (Bruder et al., 1995). This is a
cognitively demanding dichotic listening task that yields a
left ear (right hemisphere) advantage in healthy adults for
perceiving complex tones (Sidtis, 1981; Tenke et al., 1993).
Depressed patients had significantly smaller P3 amplitude
compared to controls and also failed to show either the
behavioral left ear (right hemisphere) advantage or the
hemispheric asymmetry of P3 seen for controls. The ab-
sence of any difference in early sensory potentials (e.g., N1)
between depressed patients and controls supports the con-
clusion that the lack of a right hemisphere advantage for
perceiving complex tones is related to a relatively late stage
of cognitive processing reflected in the P3.
Arguing that binaural oddball tasks are too simple to
consistently reveal cognitive dysfunction in depression,
Tenke et al. (2008) developed a dichotic oddball task that
increases the cognitive challenge. ERPs of 38 unmedicated
depressed patients and 26 healthy controls were measured
in tonal and phonetic tasks with dichotic presentation of
stimuli. Tonal nontargets were pairs of complex tones (cor-
responding to musical notes G and B above middle C)
presented simultaneously to each ear (L/R) in an alternating
series (G/B or B/G). A different target tone (note A)
replaced one of the pair on 20% of the trials. Phonetic
nontargets were pairs of syllables (/ba/, /da/) presented
simultaneously to each ear (L/R) in an alternating series and
the target was a different syllable (/ta/). The subject’s task
was to respond to the target with a button press. Target
detection was poorer in depressed patients than controls for
both tones and syllables. Patients also showed reductions of
current source density (CSD) for parietal and temporal lobe
sources corresponding to P3. While reduction of the parietal
source was related to the patients’ poorer performance,
temporal lobe source reductions were not. Given the
involvement of primary and secondary auditory cortex in
tonal and phonetic processing (Zatorre et al., 1992), these
findings support evidence of temporoparietal dysfunctions
in depression (e.g., Bruder et al., 1995; Deldin et al., 2000;
Heller et al., 1995; Post et al., 1987). The P3 source reduc-
tion in depressed patients was not lateralized to one hemi-
sphere, and the tonal and phonetic tasks did not yield
consistent behavioral ear advantages in healthy adults. The
above findings indicate that cognitively challenging dichotic
listening tasks yield consistently smaller P3 amplitudes in
ERPs in depression 5
depressed patients when compared to healthy adults.
Two studies measuring ERPs during cognitively deman-
ding visual tasks agreed in showing that individuals “at
risk” for later development of depressive disorders had
reduced P3 amplitude. Houston et al. (2003) used a visuo-
spatial oddball task that challenged attention and a complex
cognitive skill (i.e., mental rotation). Young women with a
history of a major depression episode but no current
depressive disorder (n = 29) had smaller P3 amplitude when
compared to those with no history of depression (n = 101).
Moreover, topographic maps of CSD measures correspon-
ding to P3 indicated that the difference between the pre-
viously depressed and non-depressed groups was maximal
over the right prefrontal region. Similarly, Zhang et al.
(2007) measured ERPs of healthy adults with or without a
family history of depression (n = 14 per group). The task
was a visual go/no-go task, in which large or small letters H
and O were presented on a monitor and subjects were
required to respond with the right hand to a large H or with
the left hand to a large O, and no response was required for
smaller letters. Subjects with a family history of depression,
who are at increased risk for developing a depressive dis-
order, showed smaller P3 amplitudes over temporoparietal
regions when compared to low risk subjects. LORETA
source localization methods pointed to decreased activation
of the left middle temporal gyrus in the high risk subjects.
The authors suggested that P3 decrement in visual tasks
reflects a vulnerability marker for developing depression.
This contrasts with the P3 findings for simple auditory
oddball tasks, where subjects at high risk for depression did
not differ from low risk subjects in P3 amplitude (Houston
et al., 2004) and where P3 increase following remission of
depression was suggestive of a more state-dependent effect.
Thus, while P3 reductions in a simple auditory oddball task
appear to reflect the patient’s current clinical state, P3
reductions in more demanding visual tasks appear to reflect
underlying vulnerability for a depressive disorder. It is
interesting to note in this regard that we have found EEG
evidence of reduced right posterior activity in offspring at
risk for depressive disorders (Bruder et al., 2007), which
implicates cortical regions known to mediate visual atten-
tion and perception as possible vulnerability indicators for
depression.
P3 Subcomponents
P3 is not a unitary phenomenon but consists of two or
more subcomponents associated with different cognitive
operations and neural generators (see Chapter 7, this
volume). Although the focus of most studies in depressed
patients has been on the parietal maximum P3b, this compo-
nent is often preceded by a component with a more fronto-
central topography, i.e., P3a. This frontal aspect of P3 is
prominent to novel distracter stimuli (e.g., environmental
sounds) that are interspersed along with target and standard
stimuli in a three-stimulus oddball task (Polich & Criado,
2006; Simons et al., 2001; Spencer et al., 1999). The impor-
tance of differentiating between P3 subcomponents is that
the novelty P3 or P3a is thought to reflect frontal attention
or orienting mechanisms, whereas P3b reflects temporo-
parietal mechanisms associated with context updating and
memory processing (Polich, 2007). Studies examining P3
subcomponents in depressed patients could therefore pro-
vide new information concerning the nature of their cogni-
tive deficit and underlying neurophysiologic mechanisms.
The first studies to examine P3a and P3b subcomponents
in depressed patients used go/no-go reaction time tasks
(Pierson et al., 1996) and divided patients into two sub-
groups to deal with the issue of clinical heterogeneity of
depression. The authors referred to an initial study, in which
they recorded ERPs during a simple forewarned reaction-
time task and reported that a subgroup of anxious-agitated-
impulsive patients had greater amplitude of the frontal P3a
when compared to a subgroup of patients having retarded-
blunted affect. In their subsequent study, they used a com-
plex forewarned choice reaction-time task so as to measure
P3 subcomponents in a more effortful and cognitively
demanding task. Although they reported finding no diffe-
rence in P3a amplitude among groups, peak-to-peak mea-
sures of N2b-P3a amplitudes were smaller in depressed
patients than controls, and retarded-blunted affect patients
had smaller N2b-P3a than the anxious-agitated-impulsive
patients, with the same tendency when compared to
controls. Also, the anxious-agitated-impulsive subgroup had
larger P3b amplitudes when compared to either the
retarded-blunted-affect subgroup or controls. These findings
supported the importance of differentiating between P3
subcomponents and patients with different symptom
features in studies of depressed patients.
In a study measuring ERPs during tonal or phonetic two-
stimulus oddball tasks (Bruder et al., 2002), we used princi-
pal components analysis (PCA) to identify and measure
overlapping P3 subcomponents in patients having a depres-
sive disorder alone (n = 58), an anxiety disorder alone (n =
22), comorbidity of these disorders (n = 18), and healthy
controls (n = 49). An early P3 subcomponent (peak latency
315 ms) was larger in patients having an anxiety disorder
alone (primarily social phobia or panic disorder) when
compared to depressed patients or healthy controls.
Depressed patients having a comorbid anxiety disorder
tended to have a smaller early P3 than healthy controls, but
those having a depressive disorder alone did not. The timing
and frontocentral topography of this early P3 subcomponent
resembles that seen for P3a. It should be noted, however,
that our study used a nose recording reference, as opposed
to linked ears in Pierson et al. (1996), which has implica-
tions for P3 morphology and topography. Nevertheless,
6G.E. Bruder, J. Kayser, C.E. Tenke
Figure 2. Grand average, nose-referenced ERP waveforms for 20
depressed patients and 20 healthy controls to nontarget, target and
novel stimuli at midline sites (Fz, Cz, Pz, Oz).
Figure 3. Mean integrated amplitude (± SEM) of novelty P3 and
target P3b in 20 depressed patients and 20 healthy controls at
frontal (Fz), central (Cz), parietal (Pz), and occipital (Oz) midline
sites. Significant simple group effects at each site are indicated
by: * <.05, ** <.01.
these findings appear to be in agreement with other evi-
dence that P3a or novelty P3 is heightened in patients
having an anxiety disorder. Thus, patients having a post
traumatic stress disorder were reported to have a larger
novelty P3 at frontal sites when compared to normal
controls (Kimble et al., 2000). We also found that a later
positive subcomponent (peak latency 400 ms) with a
parietal maximum typical of P3b did not differ between
patients having a depressive disorder alone and controls, but
was larger in depressed patients having a comorbid anxiety
disorder when compared to the other groups. The above
findings suggest that patients having a depressive disorder,
an anxiety disorder, or comorbidity of these disorders differ
in the amplitude of P3 subcomponents.
A limitation of the above studies is that P3 subcompo-
nents in depressed patients were measured in paradigms that
are not ideal for measuring P3a or novelty P3. In a recent
study (Bruder et al., in press), ERPs of 20 unmedicated
depressed patients and 20 healthy controls were recorded
from a 30-channel montage (nose reference) during a
novelty oddball task (Friedman et al., 1993) with three
stimuli: infrequent target tones (p = .12), frequent nontarget
tones (p = .76), and infrequent novel stimuli (e.g., animal or
environmental sounds; p = .12). Subjects responded as
quickly as possible to target tones only. There was no diffe-
rence between patients and controls in accuracy or reaction
time. Figure 2 shows the grand average waveforms at mid-
line electrode sites for patients and controls. The waveforms
show the expected N1 and N2 potentials, which are most
evident at vertex (Cz) and a novelty P3 that is also evident
at this central site. The P3 to targets is largest at the mid-
parietal site (Pz), which is typical of the P3b component.
The greater P3 to novels than targets at Fz and Cz reflects
the more frontocentral distribution of the novelty P3. As can
be seen in Figure 2, both the novelty P3 and target P3 were
reduced in depressed patients when compared to controls.
Average ERP waveforms for each stimulus condition and
for each subject were carefully inspected to select time
windows that bracketed the peaks and optimized the
measurement of mean integrated amplitude of the novelty
P3 (220-375 ms) and target P3 (280-470 ms). Amplitude of
the novelty P3 was significantly smaller in patients than
controls at frontal (p < .05), central (p < .05) and parietal (p
< .01) sites (see top portion of Figure 3). Patients also
tended to have smaller P3 amplitude to targets at central (p
< .05) and parietal (p < .10) sites (see lower portion of
Figure 3). The difference between patients and controls at
the parietal site had a large effect size for the novelty P3
(1.0) and a smaller effect size for the target P3 (0.61). There
was no significant difference in the mean integrated
amplitude between patients and controls in the N1 (70-145
ms) and N2 (150-240 ms) windows, which indicates that the
reduced novelty P3 in patients was likely not due to an
earlier deficit in detection of the deviant novel sounds.
The novelty P3 reduction in depressed patients is
suggestive of a deficit in automatic shifting of attention
(orienting) and evaluation of novel environmental sounds
(Friedman et al., 2001; Polich, 2007). There are, however,
two issues that needed further study. First, the novelty P3
component overlaps with the P3b component to targets,
which leaves open the possible contribution of P3b to the
ERPs in depression 7
Figure 4. A: Grand mean CSD waveforms for novel, target and nontarget stimuli illustrating
the difference between depressed patients and controls in novelty P3 source at vertex (Cz) and
the lack of a difference in source (Pz). CSD-PCA factor loadings (orange inset) and
topographies separate an early vertex source unique to novels (factor 241) from a subsequent
temporoparietal P3 source common to novels and targets (343) and from later, target-specific
centroparietal source (542). B: Selected means (± SEM) of novelty vertex source (241), which
was significantly reduced in depressed patients at midcentral sites (Cz and CPz).
group differences in mean amplitude in the novelty P3
window. The use of multivariate statistics, such as principal
components analysis (PCA), could aide in identifying and
measuring these separate P3 subcomponents. Second, both
neuroimaging and ERP studies have found evidence that
prefrontal, anterior cingulate, and hippocampal regions are
involved in novelty processing (Halgren et al., 1995; Knight
et al., 1998; Polich, 2007), but the neural generators under-
lying the novelty P3 reductions in depressed patients remain
unknown. An independent replication and extension of the
above study was therefore performed, in which ERPs of a
larger sample of depressed patients (n = 49) and healthy
controls (n = 49) were recorded from 67 channels during the
same novelty oddball task (Tenke et al., in press). Most
importantly, we applied a combined CSD-PCA approach to
help identify neural sources corresponding to P3 subcom-
ponents (see Kayser & Tenke, 2006a,b for details).
In the initial step for this approach, all averaged ERP
waveforms are transformed into reference-free current
source density (CSD) estimates using spherical spline sur-
face Laplacian algorithm suggestged by Perrin et al. (1989).
CSD is a mathematical transformation (2nd spatial deriva-
tive), which provides a representation of the direction, loca-
tion, and intensity of current generators that underlie an
ERP topography. CSD maps represent the magnitude of
radial (transcranial) current flow entering (sinks) and
leaving (sources) the scalp (Nunez, 1981; Nunez, & Srini
vasan, 2006). CSD is a reference-free technique that pro-
vides topographies with more sharply localized peaks than
those of scalp potentials and eliminate volume-conducted
activity from distant regions (Tenke & Kayser, 2005). In the
next step, the averaged CSD waveforms are submitted to an
unrestricted te mporal principal components analysis (PCA),
followed by Varimax rotation of covariance loadings
(Kayser & Tenke, 2006a). This approach yields distinctive
PCA components (factor loadings) and corresponding
weighting coefficients (factor scores), which provide a
concise, effient simplification of the temporal and spatial
distribution of neuronal generators (Kayser & Tenke, 2003,
2006c). Temporal PCA not only aids in determining the
relevant statistically independent components within a data
set, it also generates efficient measurements of these over-
lapping components. The combined CSD-PCA method
overcomes two critical limitations of ERP research: (1) the
dependence of ERP surface potentials on a reference loca-
tion (e.g., linked mastoids or nose)1 and (2) the definition
and measurement of ERP components (e.g., peak or integra-
ted amplitudes in specified time windows). The use of
reference-free CSD measures sharpens topographies related
to underlying neuronal generators, and PCA allows identifi-
cation and quantification of statistically independent factors
(sources and sinks) corresponding to ERP/CSD compo-
nents. The CSD-PCA technique provides a conservative
source localization method that avoids any biophysical
assumptions, unlike other popular tools (e.g., BESA or
LORETA).
1 No recording reference anywhere on the human body can be considered
neutral or inactive (e.g., sternum, neck, mastoid, nose, ear lobe) and any
site will be differentially affected by a given combination of neuronal
generators through volume-conducted activity (see Kayser & Tenke,
2006a). The choice of the reference is, therefore, essential for identifying
both spatial and temporal information in ERP recordings, as the reference
will invariably affect the spatio-temporal activation of ERP generator
patterns. Although some reference choices may enhance or reduce a
particular generator topography, all reference schemes, including a
montage-dependent average reference, are subject to the same reference
problem. Using multiple reference schemes may help for recognition of
distinct ERP components, but will not solve the reference problem.
8G.E. Bruder, J. Kayser, C.E. Tenke
Figure 4A shows the grand mean CSD waveforms for
novel, target and nontarget stimuli for depressed patients
and controls. In addition to the expected N1 and N2 sinks to
target stimuli at vertex (Cz), there was an early source to
novel stimuli, but not targets. There was also a prominent
source over mid-parietal sites (Pz), which corresponds to
the late-positive, parietal-maximum P3b component. The
extracted CSD-PCA factor loadings (orange inset in Figure
4A) and topographies separate the early vertex source
unique to novels (241 ms peak latency of factor loadings)
from parietotemporal P3 source activity common to targets
and novels (343 ms loadings peak) and from later target-
specific centroparietal source activity (542 ms loadings
peak). The novelty vertex source (factor 241) was markedly
reduced in depressed patients when compared to controls (p
= .01), with the largest group difference at the midline
central site (Group by Electrode interaction, p < .001; see
Figure 4B). Group differences were less evident for the later
P3 sources to targets (factors 343 and 542). Thus, a vertex
source that was only present to novel distracter stimuli, and
had a shorter latency (241 ms) than the source correspon-
ding to the parietal P3b, was markedly reduced in depressed
patients when compared to healthy controls.
Our studies indicate that the novelty P3 is reduced in
depressed patients. The findings using the CSD-PCA tech-
nique are remarkable for two reasons. First, the novelty ver-
tex source that discriminated between patients and controls
had a shorter peak latency, i.e., 241 ms from onset of novel
stimuli, than the sources corresponding to the parietal P3b
component. This suggests that the novelty P3 reduction in
depressed patients is indicative of a deficit in early shifting
of attention (i.e., orienting) to novel distracter stimuli and
not to later cognitive evaluation of these stimuli. Second,
the novelty vertex source was localizable to the fronto-
central region within and along the longitudinal fissure.
Studies using other source localization techniques have
localized generators of novelty P3 to the region of the
anterior cingulate cortex (ACC), whereas P3b to target
stimuli has prominent sources in the region of the temporal-
parietal junction (Dien et al., 2003; Mecklinger & Ullsper-
ger, 1995). Contributions of other cortical areas (including
frontal gyrus, insula, posterior cingulate) to these compo-
nents are, however, also known (Kiehl et al., 2001). Despite
convergent evidence for involvement of ACC in the novelty
P3, anatomical and biophysical considerations demand
caution when interpreting putative generators of midline
ERPs. The radial orientation of an equivalent dipole within
the longitudinal fissure that is typical of inverse solutions is
not normal to the surface of the cingulate gyrus, but rather
is tangential to the local alignment of cortical neurons. This
paradoxical alignment requires additional assumptions
before the generator can be considered to be physiologically
plausible (Tenke & Kayser, 2005; Kayser & Tenke, 2006a;
Kayser et al., 2007; Tenke & Kayser, 2008). Frontal cortex,
including the anterior cingulate, is known to be of key
importance for attention, and has been found to be
dysfunctional in depressed patients (Bremmer et al., 2004;
Drevets et al., 1997; Siegle et al., 2004). Although this may
point to the frontal cortex – and in particular the ACC – as
being responsible for novelty P3 reductions in depression,
studies indicate that the hippocampus and other cortical
structures are also involved in the generation of the novelty
P3 (Halgren et al., 1995; Knight, 1996; Kiehl et al., 2001),
and therefore further research is needed to pinpoint the
origins of this deficit in depressed patients.
ERPs During Processing of Emotional Words or Pictures
Studies in healthy adults have consistently found that
emotionally arousing words or pictures elicit a late positive
potential (i.e., beyond 300 ms) extending into a slow wave,
and the amplitude of this potential is greater for negative or
positive emotional stimuli when c ompared to neutral st imuli
(Johnston et al., 1986; Kayser et al., 1997; Naumann et al.,
1992; Palomba et al., 1997). Several studies in depressed
patients have reported abnormalities of P3 to visually-
presented emotional stimuli. In one of the first studies,
Blackburn et al. (1990) recorded ERPs from 3 midline sites
referenced to the left ear and found that depressed patients
(n = 15) had smaller P3 amplitude to negative than neutral
or positive words, whereas healthy controls (n = 15) showed
larger P3 amplitude to negative words than neutral or
positive words. Inspection of their data also suggests that
depressed patients had smaller P3 amplitude than controls
to negative words, but not to neutral or positive words, but
no statistics were presented to support the significance of
group differences in P3 amplitude. All the patients in the
Blackburn et al. study were taking antidepressant medica-
tions and its impact on their findings is unknown. Using a
30-channel montage, Kayser et al. (2000) measured nose-
referenced ERPs of 30 unmedicated depressed patients and
16 healthy controls during passive viewing of negative
pictures of patients with dermatological diseases or neutral
control pictures of these patients after surgical treatment. As
shown in Figure 5A, depressed patients had significantly
smaller amplitude of a late P3 potential (460 ms peak
latency of a surface potential factor derived from unrestric-
ted temporal PCA followed by Varimax rotation) when
compared to controls. As in prior studies (Cacioppo et al.,
1993, 1996; Kayser et al., 1997), healthy controls showed
enhanced late P3 (P460) amplitude to negative compared to
neutral stimuli and this enhancement was greatest over the
right parietal region (see Figure 5B). In contrast, depressed
patients did not show this increase in late P3 to negative as
compared to neutral stimuli over either hemisphere.
Interestingly, the PCA-based ERP decomposition also
revealed an early P3 subcomponent (330 ms peak latency of
ERPs in depression 9
Figure 5. A: Grand mean, nose-referenced ERP waveforms at lateral-parietal (P7/8) sites comparing neutral and negative stimuli for
healthy and depressed participants. Distinct ERP components (P1, N1, N2, early and late P3) are labeled for healthy adults at site P8.
Data-driven ERP measures of P3 subcomponents were determined by means of PCA (factors P330 and P460; cf. Kayser et al., 2000).
B: Topographies of PCA factor scores for factor P330 (early P3 rising phase) and factor P460 (classic parietal P3) comparing negative
and neutral stimuli and their differences (pooled across visual fields of lateralized presentations) for healthy and depressed participants.
Unlike healthy adults, depressed patients showed no effect of emotional content for factor P460, but showed instead an emotional content
effect for factor P330 with comparable right-posterior lateralization.
factor loadings) consisting of a right parietal and frontal
positivity, which showed a right-lateralized, negative-
larger-than-neutral emotion effect in patients, suggesting
intact early classification but impaired late evaluation of
affective significance in depression. There was also no
difference between depressed patients and healthy controls
in valence and arousal self-report measures to these stimuli,
which further suggest preserved (cognitive) classification of
emotional stimuli in depression.
While subjects in the Kayser et al. (2000) study passive-
ly viewed the emotional pictures to reduce the impact of
cognitive processing resulting from specific task demands
(e.g., target detection, matching paradigm), Deldin et al.
(2000) recorded ERPs (linked mastoids) from 9 sites to
positive, neutral and negative face and word stimuli during
a recognition memory task. They found a lateralized
abnormality of the N2 potential in depressed patients (n =
19) when compared to healthy controls (n = 15). N2
amplitude over right parietal region was reduced in the
depressed patients and this reduction was most evident
during the processing of pleasant faces. If one assumes that
the recording location of referenced surface potentials
reflects differential activation of the underlying cortical
regions, the findings of both Kayser et al. (2000) and Deldin
et al. (2000), involving different ERP components and
methods, appear to be consistent with the hypothesis that
depressed patients have impaired activation of right parietal
regions during the processing of emotional stimuli (Heller,
1990, 1993). Of course, further study of the neural genera-
tors of these effects is needed before this conclusion can be
drawn with confidence. Additional evidence of reduced P3
amplitude to emotional stimuli in depressed subjects was
obtained by Cavanagh and Geisler (2006), but it was
present only for midline electrode sites. They recorded
ERPs (linked ears) from 7 sites during a visual oddball task
in which neutral faces served as standards and happy or
fearful faces were targets. Depressed subjects (n = 36) had
reduced P3 amplitude to happy faces when compared to
non-depressed controls (n = 18). In summary, the most
consistent finding across studies was reduced late P3 (P3b)
amplitude to emotional stimuli in depressed subjects, but the
valence of stimuli to which this occurs and the laterality of
10 G.E. Bruder, J. Kayser, C.E. Tenke
this P3 deficit are less clear. However, only Kayser et al.
(2000) used a sufficiently dense EEG montage to evaluate
lateralized P3 activity over inferior temporal and parietal
regions.
Given evidence that depressed patients have a negative
bias for processing information during memory tasks,
several studies have measured ERPs of depressed patients
in memory tasks with stimuli of different valence to assess
processing bias. Studies recording ERPs during recognition
memory tasks have found that the influence of emotional
valence of stimuli on late positive or P3 potentials in
healthy adults was less evident for depressed subjects
(Deldin et al., 2001b; Dietrich et al., 2000a). Although the
specific findings differ across studies using word or face
stimuli, evidence of mood congruent biases in depressed
patients have been reported in studies measuring slow wave
amplitudes during sustained processing of positive, neutral
or negative stimuli in working memory tasks (Deldin et al.,
2001a; Deveney & Deldin, 2004; Shestyuk et al., 2005).
ERPs to Olfactory Stimuli
Given the overlapping cortical and limbic systems
involved in olfaction, emotional processing, and depression
(most notably, the amygdala and orbitofrontal cortex), the
study of ERPs to olfactory stimuli may hold particular
promise for elucidating neurophysiologic dysfunctions
responsible for abnormalities of emotional reactivity in
depressed patients. Odors appear to be powerful emotional
stimuli with distinctive hedonic valence (Pause et al., 2003).
Importantly, the emotional content of odors can be
perceived with little cognitive mediation (Ehrlichman &
Bastone, 1992), allowing a more direct assessment of
emotional processing in depression. Although there is
evidence of differences in emotional evaluation of odors
between depressed patients and controls (Pause et al., 2000;
Steiner et al., 1993), we know of only one study that used
ERPs to study olfactory processing in depressed patients.
Pause et al. (2003) measured olfactory ERPs at 30 scalp
locations (linked ears reference) in 22 patients having a
MDD and 22 healthy controls in a task requiring discrimi-
nation of pleasant (phenyl-ethyl alcohol = rose) and unplea-
sant (isobutyraldehyde = rotten butter) odors presented
using a constant-flow olfactometer. Control tasks measured
visual ERPs to colors or emotional pictures (Lang et al.,
1999). Although patients performed as well as controls, they
showed reduced amplitude of P2 and early P3 potentials at
frontal sites. In contrast, only visual ERPs reflecting later
cognitive processing (P3b and slow wave) were reduced in
depressed patients to colors or emotional slides. The authors
attributed the reduction of the olfactory P2 potential in
depressed patients to a deficit in the ability to preattentively
encode the pleasantness of odors. Reduced early P3 at
frontal sites in depressed patients was thought to reflect a
reduction in early cognitive evaluative processes. They
further proposed that reduced olfactory P2 and P3 in de-
pressed patients may be related to specific alterations in the
amygdala and orbitofrontal cortex, respectively. When 14
of the 22 patients were retested after successful antidepres-
sant treatment, these patients no longer showed smaller
olfactory ERPs. However, reduced sample power and
possibility of selective patient dropout or repeated testing
may have contributed to these null findings. In addition to
significant problems regarding olfactory ERP component
definition and measurement, this study did not control for
medication and no information was given about the relation
of the olfactory ERP deficits to severity of depressive
symptoms. Further studies recording ERPs to odors of
positive and negative valence are needed to replicate and
expand on the encouraging findings of Pause et al. (2003).
During Recognition Memory Tasks
A meta analysis indicated that depression is associated
with memory impairments for tests of both recall and
recognition (Burt et al., 1995). This memory loss is not
universal but appears to depend on patient characteristics,
such as diagnostic subtype, severity of depression and age
(Purcell et al., 1997). Unmedicated outpatients having a
major depressive disorder demonstrated a deficit in verbal
episodic memory on the California Verbal Learning Test
(Otto et al., 1994). Impaired verbal episodic memory in
depressed patients may stem from left prefrontal and medial
temporal deficits, in particular involving the hippocampus.
Thus, Sapolsky (2000) reviewed evidence from volumetric
MRI studies in patients having severe, repeated depressive
episodes and found evidence of hippocampal atrophy,
which was greater on the left side. These hippocampal
deficits in depressed patients have been linked to explicit
memory impairments (Sapolsky, 2000; Shah et al., 1998).
However, studies have rarely measured neurophysiologic
functioning of depressed patients while they were engaged
in a memory task.
ERP correlates of memory processes have been
examined during a continuous word recognition memory
task (Friedman, 1990). Subjects viewed a series of words,
some of which were repeated after a number of intervening
words, and their task was to decide whether each word was
new (not previously presented) or old (previously presen-
ted). A robust, replicable finding in healthy adults has been
a more positive-going potential for correctly recognized old
than new words about 250 to 800 ms after word onset,
referred to as the “old-new effect” (see Chapter 14, this
volume). Intracranial recordings in and around medial
temporal structures of epilepsy patients have shown similar
old-new effects, suggesting generators in the hippocampus,
parahippocampal gyrus or amygdala (Elger et al., 1997;
Smith et al., 1986). Moreover, patients with left anterior
ERPs in depression 11
Figure 6. P3 source activity during auditory (A) and visual (B) word recognition memory
tasks in 37 depressed patients and 40 healthy controls. Grand mean CSD waveforms at
selected left parietal sites illustrate group differences at site P7 and old/new effects for each
group at site P3. Significant topographic old-new effects (max T2 randomization tests) are
shown for each modality and group for CSD factors corresponding to the modality-specific
P3 source (cf. CSD factor loadings in green inset). C: P3 source means (± SEM) for old and
new items (across modality) at lateral parietal sites of the left (LH: P7, P3, CP5) and right
(RH: P8, P4, CP6) hemisphere.
temporal lobectomy showed a dramatic reduction of the old-
new effect for word recognition when compared to patients
with right temporal lobectomy or controls (Johnson, 1995;
Smith & Halgren, 1989; Rugg et al., 1991).
Given evidence of memory impairments and hippo-
campal deficits in depressed patients, one would predict that
they will show a reduced old-new effect during a word
recognition task. One published study reported findings
consistent with this prediction (Dietrich et al., 2000a). They
measured ERPs of 11 unmedicated depressed patients and
11 healthy controls during a continuous word recognition
memory task. The depressed patients were significantly
poorer in recognizing the repeated (old) items and showed
smaller old-new effect when compared to controls. More-
over, the reduction of the old-new effect for words persisted
following clinical improvement of depression (Dietrich et
al., 2000b). This study, however, had several methodo-
logical weaknesses, such as the use of a right mastoid
reference (which is problematic for examining laterality
effects), a small sample size, and a younger control group
showing unusually large P3 amplitudes (group mean > 20
µV), which limit the impact of the findings.
In a recent study (Kayser et al., in preparation), we mea-
sured 31-channel ERPs from 37 right-handed, unmedicated
depressed patients (21 men) and 40 right-handed, healthy
controls (19 men) during continuous recognition memory
tasks, in which a series of words were presented in either
the visual or auditory modality. Subjects indicated for each
word whether it was “new” or “old” by pressing one of two
buttons (for procedural details see Kayser et al., 2007).
Although all subjects had adequate, above-chance
performance (86.4% overall correct recognition of repeated
words; SD = 12.7), depressed women showed poorer
recognition memory than healthy women, but there was no
group difference in men (group by gender interaction, p <
.05). There was, however, no significant group difference in
12 G.E. Bruder, J. Kayser, C.E. Tenke
response latency for visual or auditory word presentations.
For improved spatial and temporal characterization of the
ERP old-new effects, the data were analyzed using our
newly developed CSD-PCA technique (Kayser & Tenke,
2006a,b; Kayser et al., 2007).
Both patients and controls showed the expected old-new
effects, with greater late source activity (positivity) at
posterior sites to correctly recognized old words for both
auditory (Figure 6A) and visual (Figure 6B) modalities.
This source activity, corresponding to the late P3b potential,
was identified in separate PCAs for each modality by the
auditory CSD factor 650 (peak latency of factor loadings in
ms) and the visual CSD factor 475 (peak latency in ms; see
green factor loading waveforms in Figure 6A,B). Based on
response-locked averages, these P3 source factors peaked
about 170 ms and 140 ms, respectively, prior to response
onset in either modality (cf. Kayser et al., 2007). As evident
for both groups in the CSD factor topographies, increased
lateral-parietal P3 sources (warm colors) for old as
compared to new auditory stimuli were accompanied by
increased lateral-frontal sinks (right portion of Figure 6A),
and similar old-new effects with mid-parietal and mid-
frontal P3 sources were found for visual stimuli (right
portion of Figure 6B). These old-new effects were present
for both groups, as indicated by the significant pairwise max
(T2) randomization tests (Maris, 2004) for each group and
modality (last column in Figure 6A,B). However, there were
notable topographic group differences in the visual P3 old-
new effect, which was shifted towards occipital sites in
depressed patients. A repeated measures ANOVA of P3
source factor scores from both modalities was computed at
homologous left and right lateral-parietal sites (P3/4, P7/8,
CP5/6), where the P3 source was prominent. This analysis
revealed a significant Group main effect (p < .001) and
interactions of Group by Hemisphere (p = .001) and Group
by Hemisphere by Condition (new, old) (p < .01). Healthy
adults had overall greater P3 source activity at lateral-
parietal sites when compared to depressed patients, particu-
larly over the left hemisphere (Figure 6C). Although the
condition main effects, indicative of the old-new effects,
were highly significant for both groups (p < .0001), the old-
new effect was larger over the left than right hemisphere in
controls (p = .01), but not in patients, and there was a
significant simple Group by Condition interaction at the left
(p < .05) but not right hemisphere. An analogous ANOVA
for the accompanying sink activity at lateral-frontal sites
(FC5/6, F7/8, FT9/10) revealed only a marginally
significant Group main effect (p = .07), but a significant
Group by Gender interaction (p < .05), stemming from
reduced sinks in depressed compared to healthy women, but
no group difference in men.
In summary, although the findings show only small
behavioral impairment of recognition memory for words in
depressed women and none in depressed men, they indicate
that the ERP correlate of conscious episodic memory
retrieval is reduced in depressed patients over the left
parietal region and this reduction is largely independent of
processing modality, which suggests a deficit in accessing
semantic (i.e., lexicon) information during continuous word
recognition. Event-related fMRI studies in healthy adults
have found that recognition of “old” words involves a left-
lateralized network including frontal, lateral parietal,
posterior cingulate and the precuneous (Henson et al.,
2000). Also, given evidence of left medial temporal lobe
involvement in the old-new effect for words (Johnson,
1995; Smith & Halgren, 1989; Rugg et al., 1991), a
distributed network including the hippocampus or other
medial temporal lobe structures may also contribute to the
reduced old-new effect for depressed patients. Further
studies using ERP measures in conjunction with
neuroimaging techniques are needed to further resolve the
neural basis of the episodic memory deficit in depression.
N1 and Intensity Dependence of Auditory ERPs
Up to this point we have focused on late cognitive
potentials, but studies have also examined earlier negative
brain potentials (N1 or N2) in depressed subjects. The N1
potential is known to reflect early sensory processing of
stimuli and is also modulated by attention and arousal level
(see Chapters 4 and 11, this volume). However, unlike the
omnipresent mid-parietal P3b potential, the amplitude and
topography of N1 and N2 change considerably with the
recording reference, dependent on processing modality. For
example, for a visual N1 peaking at approximately 140 ms,
the nose-referenced ERP morphology will reveal a distinct
inferior parietal negativity, which will reverse into a distinct
positive deflection at mid-parietal sites when re-referencing
these ERPs to linked mastoids, leaving only a substantially
reduced negative deflection over lateral parietal regions
(e.g., Kayser et al., 2003, 2007). In contrast, the auditory N1
peaking at about 100 ms will maintain a central maximum
with most common reference schemes, because the
direction and location of the known underlying generator
within the primary auditory cortex will always result in a
mid-central negativity, unless a vertex reference is used.
The reason is that the reference location, like all other
electrodes included in the EEG montage, is an active site,
and the differential activity (i.e., the potential difference or
ERP) between any two recording sites will tend to be
smaller with closer proximity, or larger with increasing
distance (e.g., cf. chapter 3 in Luck, 2005).
There have been reports of reduced N1 amplitude in
depressed subjects when compared to non-depressed
controls (Burkhart & Thomas, 1993; Knott & Lapierre,
1987; El Massioui & Lesevre, 1988; Sandman et al., 1987).
These four studies recorded ERPs primarily at central sites
ERPs in depression 13
Figure 7. A: N1 sink (factor 120) topography for 51 healthy
controls, 22 treatment responders or 11 treatment nonresponders.
B: N1 sink means (± SEM) for subgroups of treatment responders
and nonresponders to mono or dual therapy.
(linked-ears reference ) in dichotic listening or tone counting
tasks, but one study used a visual RT task (Knott &
Lapierre, 1987). Twice as many studies did not, however,
find evidence of N1 reduction in depressed patients in
auditory tasks (Blackwood et al., 1987; Bruder et al., 1995;
Bruder et al., 1998; El Massioui et al., 1996; Knott et al.,
1991; Ogura et al., 1993; Sara et al., 1994; Tenke et al.,
2008). All four studies that recorded ERPs mainly at mid-
line sites (2 linked ears, 1 left ear, and 1 nose reference)
during binaural oddball tasks found no difference in N1
amplitude between depressed patients and controls (Black-
wood et al., 1987; Bruder et al., 1998; Ogura et al., 1993;
Sara et al., 1994). The remaining four studies that found no
N1 reduction in depressed patients recorded ERPs during
different dichotic listening tasks (2 linked ears and 2 nose
reference). The lack of a N1 reduction in depressed patients
was also evident in our findings for a novelty oddball task
(see Figures 2 and 4) and for auditory and visual word
recognition memory tasks (Figure 6). Although medication
differences across studies is not an issue (all patients were
off medication), the extent to which differences in clinical
characteristics of patients could account for the conflicting
findings is unclear.
Although our CSD-PCA study using the novelty oddball
task did not focus on N1 sink activity (Tenke et al., in
press), a subsequent analysis of a subgroup of depressed
patients who responded favorably to antidepressants showed
reduced amplitude of an N1 sink (120 ms loadings peak) to
novel sounds. The depressed patients were tested during a
pretreatment session and subsequently treated as part of
ongoing clinical trials in which they received 8-12 weeks of
monotherapy with escitalopram or other selective serotonin
reuptake inhibitor (SSRI), the noradrenaline/dopamine
reuptake inhibitor (NDRI) bupropion, or dual therapy with
both SSRI and NDRI antidepressants. Following treatment,
the Clinical Global Impression: Improvement (CGI-I) scale
was used by an independent clinician to rate the treatment
response of the patients. Responders (rated as being much
or very much improved) showed reduced N1 sink activity
(maximum anterior to Sylvian fissure) compared to either
nonresponders (p < .05) or healthy controls (p = .01),
whereas no difference was found between nonresponders
and controls (see blue regions in Figure 7A). Although
samples were small, it is interesting to note that responders
to monotherapy had the smallest N1 and nonresponders to
dual therapy had the largest N1 (see Figure 7B). The CSD-
PCA topographies (Figure 7A) indicate that N1 sinks were
coupled with sources posterior to the Sylvian fissure, and
are thereby consistent with tangentially-oriented generators
in or adjacent to primary auditory cortex. Given the high
serotonergic innervation of primary auditory cortex (Lewis
et al., 1986; Campbell et al., 1987), it is possible that
reduced pretreatment N1 to novel, distracter sounds may
reflect lower level of serotonin neuronal activity in respon-
ders. Interestingly, a study of the effects of tryptophan
depletion on mismatch negativity (MMN) in healthy adults
suggested that decreased serotonin may decrease involun-
tary attention shifting to task-irrelevant sounds (Ahveninen
et al., 2002). Further study should therefore examine
whether the reduced N1 sink activity in depressed patients
who respond favorably to antidepressants may be associated
with decreased automatic directing of attention to the task-
irrelevant novel sounds.
There is also evidence that intensity-dependence of early
auditory ERPs (N1-P2) may be of value for identifying a
subgroup of depressed patients with a serotonin deficit
responsive to treatment with antidepressants that act on the
serotonergic neural system. Increases in tone intensity from
60 to 100 dB are known to result in a linear increase in N1-
P2 amplitude in healthy adults. Hegerl and Juckel (1993)
reviewed findings from basic and clinical studies suggesting
that the slope of the function relating tone intensity and N1-
P2 amplitude provides a noninvasive indicator of central
serotonergic activity. Juckel et al. (1999) found direct evi-
dence of an inverse relationship between serotonergic
neural activity in the dorsal raphé and intensity dependence
of auditory ERPs recorded from primary auditory cortex in
cats. Hegerl and Juckel (2001) suggest that seotonergic
neurons modulate activity in primary auditory cortex by
providing a stable, tonic firing rate. A high firing rate of
serotonergic neurons is associated with a weak intensity
14 G.E. Bruder, J. Kayser, C.E. Tenke
dependence, i.e., only a small increase in N1/P2 amplitude
with increasing tone intensity, whereas a low tonic firing
rate is related to a strong intensity dependence, i.e., a large
increase in N1/P2 amplitude. Depressed patients having low
serotonergic activity, as evidenced by pronounced intensity
dependence of N1-P2 potentials before treatment, respon-
ded better to an SSRI antidepressant compared to patients
having evidence of high serotonergic activity (Gallinat et
al., 2000; Hegerl and Juckel, 2001; Paige et al., 1994).
Paige et al. (1995) also found that small samples of
responders (n = 4) and nonresponders (n = 4) to the NDRI
bupropion showed similar differences in intensity
dependence, which could raise questions about the
specificity of this finding to SSRI antidepressants. Three
studies do, however, suggest that the relation of intensity
dependence of auditory ERPs and clinical improvement
differs for serotonergic and noradrenergic antidepressants.
Linka et al. (2004) tested 16 inpatients having a MDD
episode before receiving 3-4 weeks of treatment with the
SSRI citalopram. Stronger intensity dependence of N1 was
associated with greater decrease in depression following
treatment, which is in accord with earlier findings for SSRIs
(Gallinat et al., 2000; Hegerl and Juckel, 2001; Paige et al.,
1994). In their next study (Linka et al., 2005), 14 inpatients
having a major depressive episode were tested before
receiving the selective noradrenaline reuptake inhibitor
(NARI) reboxetine. In contrast to findings for SSRIs,
smaller intensity dependence of N1 was associated with
greater improvement in depression following 3-4 weeks of
treatment with an NARI antidepressant. Patients were not,
however, randomly assigned to treatment, which weakens
the comparison of findings for the SSRI and NARI
antidepressants. More recently, Mulert et al. (2007)
measured the intensity dependence in depressed patients
who were randomly assigned to treatment with either the
SSRI citalopram or the noradrenergic antidepressant
reboxetine. Indices of intensity dependence were obtained
using LORETA analyses to measure the tomographic
current source distribution in primary auditory cortex for the
latency window 60-240 ms following stimulus onset. They
found a significant difference between citalopram
responders (n = 7) and nonresponders (n = 4), with
responders showing the expected stronger intensity
dependence. In contrast, reboxetine responders (n = 3) and
nonresponders (n = 6) did not show a significant difference
in intensity dependence. These are encouraging findings,
but given the small samples in these studies, further
research is needed to investigate the specificity of intensity
dependency as a predictor of SSRI treatment response.
If intensity dependence of auditory ERPs provides a
marker of central serotonergic activity, the slope of this
function would be expected to decrease following treatment
with an SSRI. Gallinat et al. (2000) retested 19 depressed
patients following 4 weeks of treatment with an SSRI and
found no change in the intensity dependence function,
which agrees with prior findings for two studies using SSRI
or other antidepressants (Paige et al., 1994, 1995). In
contrast, a double-blind placebo controlled study in healthy
adults did find a decrease in the slope of the N1-P2 function
during acute administration with a single dose of the SSRI
citalopram (Nathan et al., 2006). However, acute depletion
of serotonin in healthy adults after trytophan administration
did not affect intensity dependence (Debener et al., 2002;
Dierks et al., 1999).
Studies of intensity dependence of auditory ERPs as
predictors of response to antidepressants are of particular
interest because of the potential for clinical application, but
they have suffered from a number of limitations. Sample
sizes have generally been small, most studies used open
treatment with only a single antidepressant, and retest
intervals when patients were on an SSRI have been too
short to expect significant enhancement of serotonin. Also,
a variety of different methods have been used to measure
intensity dependence, including measures of scalp
potentials, LORETA, and dipole source analysis of N1, P2
or N1-P2 difference waveforms. Interestingly, reliabilities
(temporal stability, internal consistency) of intensity-
dependent ERP amplitude slope estimates can be substan-
tially improved by using PCA-based as opposed to peak-
based amplitude measures (Beauducel et al., 2000), which
suggests possible avenues of improvement for predicting
treatment response.
Nd, N2 and MMN
Studies have also used auditory ERPs to study selective
attention in depressed subjects. Negativity in the region of
the N1is known to be greater to attended than unattended
stimuli, which has allowed the measurement of “attention-
related” N1, and more specifically, the negative difference
(Nd) potential, i.e., the difference in ERP to attended and
ignored stimuli (see Chapter 11, this volume). Three studies
have agreed in finding no difference in attention-related N1
or Nd in depressed subjects and healthy controls (Burkhart
& Thomas, 1993; Massioui & Lesevre, 1988; Knott et al.,
1991). Thus, depression does not appear to involve a deficit
in voluntarily directing attention to specific stimuli.
There is, however, less agreement concerning the N2
potential in depression, with some studies finding increased
N2 amplitude in depressed or dysthymic subjects when
compared to non-depressed controls (Bruder et al., 1998;
Giese-Davis et al., 1993; Sandman et al., 1992), and others
finding no difference (Blackwood et al., 1987; Kaiser et al.,
2003) or reduced N2 in depressed subjects (Deldin et al.,
2000; Massioui et al., 1996; Massioui & Lesevre, 1988;
Sandman et al., 1987). While this difference in N2 findings
could in part stem from differences in clinical characteris-
ERPs in depression 15
tics of patients, Sara et al. (1994) found no evidence of a
relation between N2 amplitude and severity of depression.
They did find that drug-free patients having a major
depression showed greater N2 amplitude when compared to
medicated depressed patients and healthy controls, which
differs from the lack of medication effects on P3 in their
study. However, the subjects in most studies were off
medication and this is therefore not likely to be a factor.
Differences in tasks used in the above studies may well
have contributed to different findings. Specifically, the three
studies finding increased N2 amplitude in depressed
subjects used auditory oddball (Bruder et al., 1998), tone
discrimination (Giese-Davis et al., 1993) or tone counting
tasks (Sandman et al., 1992). In contrast, two studies
finding decreased N2 amplitude used auditory selective
attention tasks (Massioui et al., 1996; Massioui & Lesevre,
1988) and one used a visual recognition memory task
(Deldin et al., 2000). Moreover, the studies differed widely
in the methods used to compute N2 amplitude. Some studies
computed N2 amplitude based on difference waveforms
(e.g., target minus nontargets) to reduce the influence of
exogenous components (N1, P2), while others used
baseline-to-peak or peak-to-peak measures that may have
been more affected by these overlapping components.
Moreover, N2 identification, and accordingly the experi-
menter’s decision of how and where to measure it, is
considerably affected by the choice of ERP recording
reference.
As seen for P3, N2 is composed of two or more overlap-
ping subcomponents (Näätänen & Gaillard, 1983). Mis-
match negativity (MMN) or N2a is associated with
automatic detection of a mismatch between stimuli (see
chapter 6, this volume). This precedes and overlaps N2b,
which is associated with categorization and controlled
processing of target stimuli. Both are typically computed by
obtaining difference waveforms, subtracting the waveforms
for frequent from rare stimuli. The problem is that little
attention has been directed to obtaining separate measures
of MMN and N2b in depressed patients. In an auditory
oddball task, Ogura et al. (1993) measured mean integrated
amplitude of N2 from difference waveforms (rare minus
frequent stimuli; linked ears) in 36 unmedicated depressed
patients and 36 healthy controls. To obtain estimates of N2
subcomponents, they measured the mean amplitude of N2a
in the latency range of 120-165ms and N2b in the latency
range of 170-235ms. The mean amplitudes were smaller in
depressed patients in both the early and late windows.
While the N2a estimate for rare stimuli was reduced in
depressed patients compared to controls, negativity in the
N2b latency range was greater to frequent stimuli in
depressed patients. They concluded that the automatic
processing of mismatch was reduced in depressed patients,
whereas the later controlled processing of nontargets was
more activated in these patients. Another possibility not
considered by the authors is that N2 and P2 typically
overlap in auditory oddball tasks, such that P2 is present for
frequent nontarget tones but replaced (or overlapped) by N2
for infrequent target tones (cf. Kayser et al., 1998). In this
case, their findings for frequent stimuli could be interpreted
as a reduced nontarget P2 in depressed patients. A critical
limitation of this study, however, is that N2a was not
obtained in a standard MMN paradigm, where subjects do
not attend to tones and parameters are optimized for
measuring MMN. Giese-Davis et al. (1993) used the para-
digm of Sams et al. (1983) to provide separate measures of
N2a (150-250 ms) and N2b (150-350 ms) using difference
waveforms (linked mastoids). They found no difference
between dysthymic subjects and controls in N2a in an
“ignore” condition, but dysthymics had markedly greater
N2b than controls. Umbricht et al. (2003) also found no
difference in MMN (nose reference) between 22 depressed
patients and 25 healthy controls in a standard paradigm.
Sumich et al. (2006) compared the amplitude of N2
(linked mastoids) to target tones in 70 subclinically-
depressed subjects (i.e., those scoring 2 or more on the
Depression Anxiety and Stress Scale) and 70 subjects with
no signs of depression. While these groups did not differ in
overall level of N2, the nondepressed subjects showed
greater N2 amplitude over the right than left central sites,
but subclinically depressed subjects did not show a hemi-
spheric asymmetry of N2. Although in a different modality,
this parallels the finding of reduced N2 amplitude over right
parietal sites in depressed patients during the processing of
pleasant faces (Deldin et al., 2000). These findings are also
of interest given reports that depressed patients show
reduced P3 amplitude over right temporoparietal sites
(Kawasaki et al., 2004) or fail to show the right-greater-
than-left P3 asymmetry seen in healthy adults for tonal
stimuli (Bruder et al., 1998) or emotional pictures (Kayser
et al., 2000). The above findings support the hypothesis that
depression is associated with reduced activation of right
temporoparietal regions during the processing of tonal or
emotional stimuli.
N2 has also been measured in tasks designed to study
conflict processing or response inhibition in depressed
patients. In the visual modality, a negative potential, i.e.,
N270, was measured in 25 unmedicated depressed patients
and 25 matched controls at frontal (F3/4) and parietal (P3/4)
electrodes (linked ears reference) during an S1-S2 paradigm
(Mao et al., 2005). Subjects indicated whether the S2 stimu-
lus (colored dot) matched the S1 stimulus or was a mis-
match. The N270 potential was elicited to S2 stimuli that
differed from the S1 stimulus and was measured from its
peak amplitude in the difference waveform (mismatch
minus match conditions). Depressed patients had smaller
N270 amplitude in the difference waveforms compared to
16 G.E. Bruder, J. Kayser, C.E. Tenke
controls at frontal and parietal electrode sites. Mao et al.
interpreted the reduced N270 as evidence of impairment of
a “conflict processing system”, involving anterior cingulate
and dorsal lateral prefrontal cortex. This system, which is
active under mismatch or stimulus discrepancy conditions,
is thought to involve the same brain processes as response
conflict or error detection. In the auditory modality, Kaiser
et al. (2003) measured 61-channel ERPs (ave rage reference)
of 16 medicated depressed patients and 16 healthy controls
during a go/no-go task. The Go task was a modification of
an auditory oddball task, but the No-Go task required
inhibition of responses to rare tones. Depressed patients did
not differ from controls in performance or ERPs during the
Go task, but performed more poorly than controls in the No-
Go task. Also, the patients showed a reduction of inferior
frontotemporal positivity in the N2 latency range (i.e.,
polarity-inverted N2) during the No-Go task. They
interpreted this as suggesting a deficit in response inhibition
in depression, which is thought to involve a prefrontal
executive control system.
Error-Related Negativity and Post-Error Processing
Following errors in two-choice reaction-time tasks, such
as go/no-go or Eriksen flanker tasks, there is an increase in
response-locked frontocentral negativity referred to as error-
related negativity (ERN) or error negativity (Ne; see Chapter
10, this volume). This component peaks 50-150 following
an incorrect response and is maximum over midline
frontocentral sites. The ERN has been considered an
electrophysiologic index of a response-monitoring or
conflict detection with likely generators in the region of the
anterior cingulate (Dehaene et al., 1994; Ruchsow et al.,
2002; van Veen & Carter, 2002; see Falkenstein et al.,
2000, for a review). Given the substantial evidence for the
role of the ACC in depression (Drevets, 2000), it is not
surprising that studies have found abnormalities of ERN in
depressed subjects. Chiu and Deldin (2007) measured ERN
(linked mastoids) in 18 individuals having a current major
depressive episode and 17 nondepressed controls during an
arrow flanker task, in which a target arrow was flanked by
congruent, incongruent or neutral distracters and subjects
responded in the direction of the target arrow. Subjects were
also given accuracy feedback under reward, punishment or
neutral conditions. The amplitude of ERN was greatest at
frontal and frontocentral sites, and the depressed group
showed greater ERN amplitude than the controls,
particularly in the punishment condition. More recently,
Holmes and Pizzagalli (2008) measured the ERN (129-
channel montage, average reference) of 20 unmedicated
patients with MDD and 20 matched healthy controls during
a Stroop task. The depressed patients had significantly
larger ERN than controls. Using LORETA analyses they
found that depressed patients, relative to controls, showed
greater current density in rostral ACC and medial prefrontal
cortex at the time of maximal ERN (80 ms following
errors). Moreover, functional connectivity analyses revealed
that activity in these regions was correlated with subsequent
activity in left dorsolateral prefrontal cortex in healthy
controls but not in depressed patients. This supported their
hypothesis that exaggerated error processing (i.e., increased
ERN) in depressed patients is not followed by recruitment
of prefrontal-based cognitive control.
There is evidence that enhanced ERN depends on the
severity of depression or negative affect. First, in the study
by Chiu and Deldin (2007), the magnitude of ERN in their
neutral condition was larger in subjects with greater severity
of self-ratings of depression. Second, Tucker et al. (2003)
found evidence of larger feedback-related negativity in
subjects having a major depression when compared to non-
depressed controls, and this difference was greatest in
subjects with moderate depression, but less in those with
more severe depression. Third, two studies measured ERN
during an Eriksen flanker task or a Go/NoGo Task in
patients having a major depressive disorder “in remission”
and found no difference between the patients and controls
(Ruchsow et al., 2004, 2006). Moreover, remitted depressed
patients in both studies showed less ERN than controls for
error trials following another error.
Enhanced ERN is not specific to depression but is seen
in children and adults having an obsessive-compulsive
disorder (Gehring et al., 2000; Hajcak et al., 2008).
Moreover, college students with high scores on scales
measuring general “negative affect” had greater ERN
amplitude than those with lower negative affect (Hajcak et
al., 2004; Luu et al., 2000). Given that negative affect is
evident in both depression and anxiety, these findings are
consistent with the conclusion that enhanced ERN is present
in both depressive and anxiety disorders. This raises the
question as to whether increased ERN is associated with
depression per se or anxiety states that often accompany
depressive disorders. No study has directly compared ERN
in patients having a depressive disorder alone, an anxiety
disorder alone, or comorbidity of these disorders.
Importantly, two studies in elderly depressed patients
suggest that elevated ERN is associated with poor outcome
of treatment with an SSRI antidepressant. In their initial
study, Kalayam and Alexopoulos (2003) measured ERN
(linked mastoids) in 22 elderly depressed patients (over 60
years old) during a Stroop interference task. They compared
the ERN of 13 patients whose depression remitted during 6
weeks of treatment with the citalopram and 9 unremitted
patients. The unremitted patients had greater ERN than the
remitted patients, with the greatest difference between
groups at the left frontal site (F3). Greater ERN amplitude
at this site was correlated with less change in depressive
symptoms during treatment and abnormal initiation/
ERPs in depression 17
perseveration scores on the Mattis Dementia Rating Scale.
They hypothesized that anterior cingulate dysfunction
contributes to limited improvement in depression during
treatment. In their next study, Alexopoulos et al. (2007)
recorded ERPs (128-channel montage, average reference)
of 12 elderly depressed patients in an emotional Go/No-Go
task, citing neuroimaging evidence that this task activates
the rostral anterior cingulate. The 6 patients who remained
symptomatic after 8 weeks of treatment had larger ERN at
midline frontal and frontocentral sites when compared to the
6 patients who were remitters. The nonremitters also had
smaller amplitude of error-related positivity 150-350 ms
after an incorrect response. These findings are intriguing
given evidence from neuroimaging and electrophysiologic
studies linking increased rostral anterior cingulate activity
and clinical response to antidepressants (Mayberg et al.,
1997; Pizzagalli et al., 2001). The studies do, however, have
several limitations. The samples were extremely small, there
was no placebo control group, and the lack of a normal
control group makes it difficult to know whether the non-
remitters had abnormally large ERN or remitters abnormally
small ERN. Also, findings for geriatric depression may not
generalize to younger depressed patients.
There is also a question as to why both increased ERN
and dysfunction of rostral ACC should be related to poorer
response to antidepressants. A possible explanation is
provided by the EEG findings of Pizzagalli and his
associates. Pizzagalli et al. (2006) measured resting EEG
prior to subjects performing an Eriksen flanker task.
Subjects having high scores on the Beck Depression
Inventory, unlike subjects with low scores, showed lower
accuracy after incorrect than correct trials. Also, topo-
graphic analyses of resting EEG using LORETA indicated
that depressed subjects had reduced pre-task gamma band
activity localized to the region of the rostral ACC. Also,
higher pretask gamma was predictive of post-error adjust-
ment in behavioral performance. They interpreted these
findings as suggesting that depressed subjects have deficits
in pre-task tonic activity in rostral ACC and in making
behavioral adjustments after errors. Although increased
affective reactions to errors in depressed patients might be
expected to be associated with greater ERN during task
performance, no ERP data were reported in this study.
Pizzagalli et al. (2001) previously reported that greater
resting EEG theta activity, localized by LORETA to the
rostral region of the ACC, was predictive of more favorable
response to treatment with the antidepressant nortriptyline.
They suggested that in treatment responders, rostral ACC
hyperactivity prior to treatment may reflect increased sensi-
tivity to affective conflict or the ability to monitor the
outcome of actions and adjust behavior, e.g., by making
post-error behavioral adjustments. In treatment nonrespon-
ders, this adaptive action monitoring may be dysfunctional
leading to reduced post-error behavioral adjustments. Fol-
lowing suggestions that resting and task related theta at
anterior midline sites may reflect a common process (Tenke
& Kayser, 2005; Tzur & Berger, 2007; Wang et al., 2005),
it may be predicted that treatment nonresponders, as com-
pared to responders, will have smaller resting EEG theta
activity and greater ERN associated with failure to adjust
their behavior during task performance. Future studies
measuring both resting EEG and ERN in depressed adults
prior to treatment are needed to evaluate this hypothesis.
Further study should also examine the possible relation
of ERN abnormalities in depression and anxiety disorders
to neurotransmitter systems. In this regard, there is evidence
that the serotonin transporter gene (5-HTTLRP) is
associated with ERN amplitude in healthy subjects
(Fallgatter et al., 2004). Those with one or two copies of the
low-activity 5-HTTLRP short genotype showed greater
ERN compared to those homozygous for the long allele.
Individuals with the short allele are at increased risk for
developing depression in response to stress, which raises the
possibility that heightened ERN may be a risk marker for a
form of depression that responds poorly to treatment with
antidepressants.
Conclusions
Cognitive P3 Potential
The classical P3 potential with parietal maximum has
been extensively studied in depressed patients. Although
there have been conflicting findings, the general trend is for
depressed patients to show some reduction of P3 to target
stimuli in an auditory oddball task (mean effect size = 0.85).
Differences in P3 findings across studies could well stem
from differences in the clinical features of the patients, with
patients having melancholic or psychotic depressions
having the largest P3 reductions. Reduced P3 amplitude in
depressed patients is at least partially state-dependent, in
that clinical improvement during treatment is accompanied
by an increase in P3. Although there are fewer reports of
abnormal P3 latency in depression, there is evidence that P3
latency is longer in patients having a bipolar or melancholic
depression who typically show psychomotor retardation or
cognitive slowing.
The moderate size of the P3 reduction in depression may
also stem from the use of simple oddball tasks with
relatively little cognitive demand. Studies measuring P3 in
cognitively challenging auditory or visual tasks have more
consistently found a P3 reduction in depressed patients.
Also, depressed patients show an overall reduced P3 to
visually-presented affective stimuli and they fail to show
greater late P3 amplitude to negative as compared to neutral
stimuli, which is seen in healthy adults. Overall, the P3
decrement in depressed patients suggests a deficit in
18 G.E. Bruder, J. Kayser, C.E. Tenke
temporoparietal regions involved in context updating,
memory, and emotional processing, although frontal regions
may also play a role. P3 reduction is not, however, specific
to depression, but is seen in other neuropsychiatric disorders
that display cognitive deficits, such as schizophrenia,
alcoholism, Alzheimer’s disease or Parkinson’s disease
(Jeon & Polich, 2003; Polich & Herbst, 2000; see also
Chapters 17, 18, and 20, this volume).
One of the problems is that most studies in depressed
subjects have not differentiated P3 subcomponents. The P3a
or novelty P3 has a more frontocentral distribution than the
classical P3b potential and may contribute to the P3 reduc-
tion in depressed patients. P3a or novelty P3 is reduced in
depressed patients, but is increased in patients having an
anxiety disorder. This also underscores the importance of
taking the patients specific clinical features, and particularly
comorbidity of depression and anxiety, into account. The
reduced novelty P3 in depression suggests an orienting
deficit or dysfunction of automatic switching of attention to
task-irrelevant stimuli, which is thought to involve
prefrontal, anterior cingulate and hippocampal regions.
Old-New Effect During Recognition Memory Tasks
The increased late positive potential to correctly
recognized words, i.e., the old-new effect, is thought to be
a neurophysiologic correlate of conscious episodic memory
retrieval. Studies of continuous word recognition indicate
that this old-new effect is reduced in depressed patients.
Given evidence of left medial temporal involvement in the
old-new effect for words, the reduced old-new effect is
consistent with neuroimaging findings of reduced hippo-
campal volumes in depressed patients. Although the old-
new deficit appears to be greatest over the left parietal
region, further study is needed to determine whether this is
specific to recognition memory for words or occurs for
nonverbal stimuli as well.
N1 and Nd Potentials
Most studies have not found a reduction of N1 amplitude
in depressed patients. Also, studies have agreed in finding
no difference between depressed subjects and controls in
attention-related N1 or the Nd potential. These findings
argue against any deficit in early sensory processing or
voluntarily directing attention in depressed patients. A
subgroup of depressed patients who respond favorably to
antidepressants were found to have reduced auditory N1 to
novel distracters, which may be related to the extensive
serotonergic innervation of primary auditory cortex. There
is evidence that the intensity dependence of early auditory
potentials (N1-P2) provides an index of central serotonergic
activity. Depressed patients with pronounced intensity
dependence of N1-P2 prior to treatment have better
response to treatment with SSRI antidepressants when
compared to patients with less intensity dependence. Small
sample sizes and methodological weaknesses in studies of
intensity dependence do, however, limit the strength of the
conclusions from these studies. Also, further study is
needed to examine the extent to which findings are specific
to SSRI antidepressants or generalize to other classes of
antidepressants with different mechanisms of action.
N2 and MMN Potentials
N2 amplitude in depressed or dysthymic subjects has
been reported to be increased, decreased or no different
when compared to healthy controls. Differences in the tasks,
clinical characteristics of patients, or medication status may
have contribute d to these different findings. Methodological
difficulties in identifying and measuring N2, particularly
when using an EEG montage with a limited number of
recording channels, has also contributed to inconsistent
findings. Tasks involving simple discrimination or counting
of tones have shown increased N2 in depressed patients,
whereas those involving more complex decisions or
response inhibition (e.g., selective attention or recognition
memory tasks) were more likely to show decreased N2.
Studies have reported that depressed patients have reduced
amplitude of potentials in the N2 latency range during
visual S1-S2 or auditory go/no-go tasks under stimulus
mismatch or response conflict conditions. It was suggested
that executive control systems, involving prefrontal and
anterior cingulate cortex, may be responsible for these
deficits. As is the case for P3 studies in depressed subjects,
little attention has been directed to separating N2
subcomponents (i.e., MMN and N2b). Two studies that
measured MMN under standard “ignore” conditions found
no difference between dysthymic or depressed patients and
controls, while two studies found evidence of enhanced N2b
in dysthymic or depressed subjects. Further study of MMN
and N2b in depressed patients using conditions known to
maximize these potentials, as well as techniques for
separately measuring them, are needed to draw more
definitive conclusions.
Performance-monitoring Potentials
Healthy adults show an increase in frontocentral
negativity following errors in tw o-choice reaction time tasks
or following negative feedback. Studies have found that
error-related negativity (ERN) is heightened not only in
depressed subjects, but also children or adults having
anxiety disorders, and college students with general
“negative affect.” This is therefore not specific to
depressive disorders and is likely to be particularly high in
subjects having comorbidity of depression and anxiety, but
this has yet to be studied. The clinical relevance stems from
findings that ERN is greater in elderly patients whose
depression failed to remit following treatment with an SSRI
ERPs in depression 19
antidepressant when compared to remitters. Given evidence
that ERN is generated in medial frontal areas in or near the
anterior cingulate, these findings are consistent with EEG
and neuroimaging evidence that the rostral ACC is asso-
ciated with clinical response to antidepressants. However,
the sample sizes in these studies have been very small and
the studies lacked healthy adult or placebo control groups.
Heightened ERN has also been found in healthy adults with
the low-activity serotonin transporter allele (5-HTTLRP
short), suggesting that heightened ERN may be a risk
marker for developing a form of depression that responds
poorly to treatment.
Future Directions
Given the conflicting P3 findings for depressed patients
during a standard two-stimulus, auditory oddball task, con-
tinued use of this task in depressed patients would appear to
be of limited value. On the other hand, ERPs continue to be
a useful tool for studying the nature of cognitive deficits in
depression and for providing information about their neuro-
physiologic underpinnings. It is of value to measure ERPs
during more challenging cognitive tasks that allow evalua-
tion of hypotheses concerning the specific cognitive or
neurophysiologic deficits in depression. For instance, the
measurement of N2 and P3 during go/no-go tasks can be
used to test hypotheses about response inhibition or conflict
monitoring in depression (Donkers & van Boxtel, 2004;
Kaiser et al., 2003). ERN can also be measured during
go/no-go or flanker tasks to test hypotheses concerning per-
formance monitoring and ACC dysfunction in depression
(Chiu & Deldin, 2007; Ruchsow et al., 2006). Similarly,
measuring the “old-new” effect during recognition memory
tasks provides a means for evaluating hypotheses concer-
ning conscious memory retrieval deficits in depression and
their neurophysiolologic correlates (Kayser et al., 2007).
Future studies of the N2 and P3 potentials in depression
should also differentiate subcomponents that involve
different cognitive operations and neuronal generators. The
novelty oddball task can be of particular value for mea-
suring P3a or novelty P3 in depression (Bruder et al., in
press; Tenke et al., in press). It is also important to use
techniques that are capable of providing separate measures
of neuronal sources underlying these subcomponents, e.g.,
the use of combined CSD-PCA measures (Kayser & Tenke,
2006a,b). In general, the use of denser EEG montages and
appropriate methods to capture the temporal and spatial
dynamics of ERPs and define ERP components would be
advantageous (Kayser & Tenke, 2005).
Surprisingly few studies have used ERPs to study the
processing of emotional information in depressed patients.
ERPs can be used to measure neurophysiologic responses
to emotional stimuli without a cognitive task, so as to yield
a purer measure of affective processing in depressed pati-
ents (Kayser et al., 1997, 2000). In this regard, one area that
is ripe for exploration is the measurement of ERPs to olfac-
tory stimuli (Pause et al., 2003). Measurement of ERPs to
olfactory stimuli provides a window for studying neurophy-
siologic correlates of emotional processing because of the
direct projections to cortical and limbic structures that are
known to be involved in emotional processing and depres-
sion, in particular the amygdala, orbitofrontal cortex, and
medial temporal cortex. On the other hand, there is also
growing interest in the interaction of cognitive control and
emotional processing regions (Ochsner & Gross, 2005),
which can be readily studied with ERPs, e.g., using cogni-
tive control or interference tasks with emotional distracters
(Fales et al., 2008).
In future ERP studies of cognitive and affective
processing, it would also be important to examine subtypes
of depression, but this requires sufficiently large samples
because small subgroups (n < 10) are of limited value.
Comparison of ERPs in depressed versus control groups is
a useful first step, but fails to adequately account for the
clinical and biological heterogeneity of depressive dis-
orders. Studies comparing ERPs in subtypes with different
symptom features are particularly important when studying
P3 subcomponents (Bruder et al., 2002; Pierson et al.,
1996). The influence of comorbidity with anxiety also needs
further attention in these studies, as well as those of perfor-
mance monitoring (i.e., ERN) in depression. Another useful
approach is to subtype patients on the basis of their clinical
response to antidepressants with a specific mechanism of
action, which would require even larger samples. Some of
the most promising findings concern the relation between
ERN and intensity dependence of auditory ERPs (N1-P2) to
outcome of treatment with antidepressants. Studies with
larger samples comparing the value of these measures for
predicting response to different classes of antidepressants
(e.g., SSRI or NDRI antidepressants) are needed to confirm
the specificity of their relation to SSRI antidepressants.
Further study should also be directed toward determining
the neurotransmitter systems that may be related to ERP ab-
normalities in depressed patients. In addition to the relation
of the serotonin system to intensity dependence of N1-P2
and ERN, there is evidence that P3 subcomponents are
dependent on dopamine or norepinephrine systems (Polich,
2007; Turetsky & Fein, 2002). Pharmacological studies
measuring ERPs before and during acute treatment with
drugs that selectively act on specific neurotransmitter
systems would be of value here.
Findings linking the serotonin transporter gene (5-
HTTLPR) with both intensity dependence of N1-P2 (Stro bel
et al., 2003) and ERN (Fallgatter et al., 2004) also indicate
the importance of additional study of the genetic correlates
of ERP abnormalities in depressed patients. Lastly, both
intensity dependence of N1-P2 (Hegerl & Juckel, 1993) and
20 G.E. Bruder, J. Kayser, C.E. Tenke
ERN amplitude in depressed subjects (Pizzagalli et al.,
2006) have been hypothesized to be related to pre-test, tonic
EEG activity in specific frequency bands. Studies should
therefore examine how findings for antidepressant respon-
ders and nonresponders on these ERP measures depend on
differences in resting EEG oscillations.
Acknowledgements
This work was supported by National Institute of Mental
Health grants MH36295 (GEB) and MH58346 (JK). We
thank Drs. Jon Stewart and Patrick McGrath and the staff of
the Depression Evaluation Service at New York State
Psychiatric Institute, where the patients in our studies were
recruited, diagnosed, and treated with antidepressants.
References
American Psychiatric Association (1994). Diagnostic and Sta-
tistical Manual of Mental Disorders., 4th ed. APA, Washington, DC.
Ahveninen, J., Kahkonen, S., Pennanen, S., Liesivuori, J.,
Ilmoniemi, R. J., & Jaaskelainen, I. P. (2002). Tryptophan depletion
effects on EEG and MEG responses suggest serotonergic modulation
of auditory involuntary attention in humans. NeuroImage, 16, 1052-
1061.
Alexopoulos, G. S., Murphy, C. F., Gunning-Dixon, F. M.,
Kalayam, B., Katz, R., Kanellopoulos, D., Etwaroo, G. R., Klimstra,
S., & Foxe, J. J. (2007). Event-related potentials in an emotional go/
no-go task and remission of geriatric depression. NeuroReport, 18,
217-221.
Ancy, J., Gangadhar, B. N., & Janakiramaiah, N. (1996). “Normal”
P300 amplitude predicts rapid response to ECT in melancholia.
Journal of Affective Disorders, 41, 211-215.
Anderer, P., Saletu, B., Semlitsch, H. V., & Pascual-Marqui, R. D.
(2002). Structural and energetic processes related to P300: LORETA
findings in depression and effects of antidepressant drugs. Methods
and Findings in Experimental and Clinical Pharmacology, 24, 85-91.
Austin, M. P., Mitchell, P., & Goodwin, G. M. (2001). Cognitive
deficits in depression. British Journal of Psychiatry, 178, 200-206.
Bange, F., & Bathien, N. (1998). Visual cognitive dysfunction in
depression: An event- related potential study. Electroencephalography
and Clinical Neurophysiology, 108, 472-481.
Beauducel, A., Debener, S., Brocke, B., & Kayser, J. (2000). On
the reliability of augmenting/reducing: peak amplitudes and principal
components analysis of auditory evoked potentials. Journal of
Psychophysiology, 14, 226-240.
Blackburn, I. M., Roxborough, H. M., Muir, W. J., Glabus, M., &
Blackwood, D. H. R. (1990). Perceptual and physiological dysfunction
in depression. Psychological Medicine, 20, 95-103.
Blackwood, D. H. R., Whalley, L. J., Christie, J. E., Blackburn, I.
M., St. Clair D. M., & McInnes, A. (1987). Changes in auditory P3
event-related potential in schizophrenia and depression. British
Journal of Psychiatry, 150, 154-160.
Bremner, J. D., Vythilingam, M., Vermetten, E., Vaccarino, V., &
Charney, D. S. (2004). Deficits in hippocampal and anterior cingulate
functioning during verbal declarative memory encoding in midlife
major depression. American Journal of Psychiatry, 161, 637-645.
Bruder, G.E. (1992). P300 findings for deprssive and anxiety
disorders. Annuals of the New York Academy of Sciences, 658, 205-
222.
Bruder, G. E., Kayser, J., Tenke, C. E., Leite, P., Schneier, F. R.,
Stewart, J. W., & Quitkin, F. M. (2002). Cognitive ERPs in depressive
and anxiety disorders during tonal and phonetic oddball tasks. Clinical
Electroencephalography, 33, 119-124.
Bruder, G. E., Kroppmann, C. J., Kayser, J., Stewart, J. W.,
McGrath, P. J., & Tenke, C. E. (in press). Reduced brain responses to
novel sounds in depression: An electrophysiological study. Psychiatry
Research.
Bruder, G. E., Quitkin, F. M., Stewart, J. W., Martin, C.,
Voglmaier, M. M., & Harrison, W. M. (1989). Cerebral laterality and
depression: differences in perceptual asymmetry among diagnostic
subtypes. Journal of Abnormal Psychology, 98, 177-186.
Bruder, G. E., Tenke, C. E., Stewart, J. W., Towey, J. P., Leite, P.,
Voglmaier, M., & Quitkin, F. M. (1995). Brain event-related potentials
to complex tones in depressed patients: Relations to perceptual
asymmetry and clinical features. Psychophysiology, 32, 373-381.
Bruder, G. E., Tenke, C. E., Towey, J. P., Leite, P., Fong, R.,
Stewart, J. E., McGrath, P. J., & Quitkin, F. M. (1998). Brain ERPs of
depressed patients to complex tones in an oddball task: Relation of
reduced P3 asymmetry to physical anhedonia. Psychophysiology, 35,
54-63.
Bruder, G.E., Tenke, C.E., Warner, V., Weissman, M.M. (2007).
Grandchildren at high and low risk for depression differ in EEG
measures of regional brain asymmetry. Biological Psychiatry, 62,
1317-1323.
Bruder, G. E., Towey, J. P., Stewart, J. W., Friedman, D., Tenke,
C., & Quitkin, F. M. (1991). Event-related potentials in depression:
Influence of task, stimulus hemifield and clinical feautres on P3
latency. Biological Psychiatry, 30, 233-246.
Burkhart, M. A., & Thomas, D. G. (1993). Event-related potential
measures of attention in moderately depressed subjects.
Electroencephalography and Clinical Neurophysiology, 88, 42-50.
Burt, D. B., Zembar, M. J., & Niederehe, G. (1995). Depression
and memory impairment: a meta-analysis of the association, its
pattern, and specificity. Psychological Bulletin, 117, 285-305.
Cacioppo, J. T., Crites, S. L., Bernston, G. G., & Coles, M. G.
(1993). If attitudes affect how stimuli are processed, should they not
affect the event-related potential. Psychological Science, 4, 108-112.
Cacioppo, J. T., Crites, S. L., & Gardner, W. L. (1996). Attitudes
to the right: evaluative processing is associated with lateralized late
positive event-related potentials. Personality and Social Psychology
Bulletin, 22, 1205-1219.
Campbell, M. J., Lewis, D. A., Foote, S. L., & Morrison, J. H.
(1987). Distribution of choline acetyltransferase-, serotonin-,
dopamine-beta-hydroxylase-, tyrosine hydroxylase-immunoreactive
fibers in monkey primary auditory cortex. Journal of Comparative
Neurology, 261, 209-220.
Castaneda, A. E., Tuulio-Henriksson, A., Marttunen, M., Suvisaari,
J., & Lönnqvist, J. (2008). A review on cognitivie impairments in
depressive and anxiety disorders with a focus on young adults. Journal
of Affective Disorders, 106, 1-27.
Cavanagh, J., & Geisler, M. W. (2006). Mood effects on the ERP
processing of emotional intensity in faces: A P3 investigation with
depressed students. International Journal of Psychophysiology, 60, 27-
33.
Chiu, P. H., & Deldin, P. J. (2007). Neural evidence for enhanced
error detection in major depressive disorder. American Journal of
Psychiatry, 164, 608-616.
Debener, S., Strobel, A., Kurschner, K., Kranczioch, C.,
Hebenstreit, J., Maercker, A., Beauducel, A., & Brocke, B. (2002). Is
auditory evoked potential augmenting/reducing affected by acute
tryptophan depletion? Biological Psychology, 59, 121-133.
Dehaene S., Posner, M., & Tucker, D. (1994). Commentary:
localization of a neural system for error detection and compensation.
Psychological Medicine, 5, 303-305.
Deldin, P. J., Deveney, C. M., Kim, A. S., Casas, B. R., & Best, J.
L. (2001a). A slow wave investigation of working memory biases in
mood disorders. Journal of Abnormal Psychology, 110, 267-281.
Deldin, P. J., Keller, J., Gergen, J. A., & Miller, G. A. (2001b).
Cognitive bias and emotion in neuropsychological models of
ERPs in depression 21
depression. Cognition and Emotion, 15, 787-802.
Deldin, P. J., Keller, J., Gergen, J. A., & Miller, G. A. (2000).
Right-posterior face procesing anomaly in depression. Journal of
Abnormal Psychology, 109, 116-121.
Deveney, C. M., & Deldin, P. J. (2004). Memory of faces: A slow
wave ERP study of major depression. Emotion, 4, 295-304.
Dien, J., Spencer, K. M., & Donchin., E. (2003). Localization of
the event-related potential novelty response as defined by principal
components analysis. Cognitive Brain Research, 17, 637-650.
Dierks, T., Barta, S., Demisch, L., Schmeck, K., Englert, E.,
Kewitz, A., Maurer, K., & Poustka, F. (1999). Intensity dependence of
auditory evoked potentials (AEPs) as biological marker for cerebral
serotonin levels: Effects of tryptophan depletion in healthy subjects.
Psychopharmacology, 146, 101-107.
Dietrich, D. E., Emrich, H. M., Waller, C., Wieringa, B. M.,
Johannes, S., & Munte, T. F. (2000a). Emotion/cognition-coupling in
word recognition memory of depressive patients: an event-related
potential study. Psychiatry Research, 96, 15-29.
Dietrich, D. E., Kleinschmidt, A., Hauser, U., Schneider, U.,
Spannhuth, C. W., Kipp, K., Huber, T. J., Wieringa, B. M., Emrich, H.
M., & Johannes, S. (2000b). Word recognition memory before and
after successful treatment of depression. Pharmacopsychiatry, 33, 221-
228.
Diner, B. C., Holcomb, P. J., & Dykman, R. A. (1985). P300 in
major depressive disorder. Psychiatry Research, 15, 175-184.
Donkers, F. C. L., & van Boxtel, G. J. M. (2004). The N2 in go/no-
go tasks reflects conflict monitoring not response inhibition. Brain and
Cognition, 56, 165-176.
Drevets, W. (2000). Neuroimaging studies of mood disorders.
Biological Psychiatry, 48, 813-829.
Drevets, W. C., Price, J. L., Simpson, J. R. Jr., Todd, R. D., Reich,
T., Vannier, M., & Raichle, M. E. (1997). Subgenual prefrontal cortex
abnormalties in mood disorders. Nature, 386, 824-827.
Ehrlichman, H., & Bastone, L. (1992). Olfaction and emotion. In
J. Serby & K.L. Chobor (Eds.), Science of olfaction (pp.410-438).
New York: Springer.
El Massioui, F., Everett, J., Martin, M. T., Jouvent, R., &
Widlocher, D. (1996). Attention deficits in depression: an electrophy-
siological marker. Neuroreport, 7, 2483-2486.
El Massioui, F., & Lesevre, N. (1988). Attention impairment and
psychomotor retardation in depressed patients: an event-related
potential study. Electroencephalography and Clinical
Neurophysiology, 70, 46-55.
Elger, C. E., Grunwald, T., Lehnertz, K., Kutas, M., Helmstaedter,
C., Brockhaus A, Van Roost, D., & Heinze, H. J. (1997). Human
temporal lobe potentials in verbal learning and memory processes.
Neuropsychologia, 35, 657-667.
Fales, C. L., Barch, D. M., Rundle, M. M., Mintun, M. A., Snyder,
A. Z., Cohen, J. D., Mathews, J., & Sheline, Y. I. (2008). Altered
emotional interference processing in affective and cognitive-control
brain circuitry in major depression. Biological Psychiatry, 63, 377-
384.
Falkenstein, M., Hoormann, J., Christ, S., & Hohnsbein, J. (2000).
ERP components on reaction errors and their functional significance:
a tutorial. Biological Psychology, 51, 87-107.
Fallgatter, A. J., Herrmann, M. J., Roemmler, J., Ehlis, A.,
Wagener, A., Heidrich, A., Ortega, G., Zeng, Y., & Lesch, K. (2004).
Allelic variation of serotonin transporter function modulates the brain
electrical response for error processing. Neuropsychopharmacology,
29, 1506-1511.
Friedman, D. (1990). ERPs during continuous recognition memory
for words. Biological Psychology, 30, 61-87.
Friedman, D., Cycowicz, Y. M., & Gaeta, H. (2001). The novelty
P3: an event-related brain potential (ERP) sign of the brain’s
evaluation of novelty. Neuroscience and Biobehavioral Reviews, 25,
355-373.
Friedman, D., Simpson, G., & Hamberger, M. (1993). Age-related
changes in scalp topography to novel and target stimuli. Psychophy-
siology, 30, 383-396.
Gallinat, J., Bottlender, R., Juckel, G., Munke Puchner, A., Stotz,
G., Kuss, H. J., Mavrogiorgou, P., & Hegerl, U. (2000). The loudness
dependency of the auditory evoked N1/P2-component as a predictor
of the acute SSRI response in depression. Psychopharmacology, 148,
404-411.
Gangadhar, B. N., Ancy, J., Janakiramaiah, N., & Umapathy, C.
(1993). P300 amplitude in non-bipolar, melancholic depression.
Journal of Affective Disordors, 28, 57-60.
Gehring, W. J., Himle, J., Nisenson, L. G. (2000). Action-
monitoring dysfunction in obsessive-compulsive disorder.
Psychological Science, 11, 1-6.
Giese-Davis, J. E., Miller, G. A., & Knight, R. A. (1993). Memory
template comparison processes in anhedonia and dysthymia. Psycho-
physiology, 30, 646-656.
Hajcak, G., Franklin, M. E., Foa, E. B., & Simons, R. F. (2008).
Increased error-related brain activity in pediatric obsessive-compulsive
disorder before and after treatment. American Journal of Psychiatry,
165, 116-123.
Hajcak, G., McDonald, N., & Simons, R. F. (2004). Error-related
psychophysiology and negative affect. Brain and Cognition, 56, 189-
197.
Halgren, E., Baudena, P., Clarke, J. M., Heit, G., Marinkovic, K.,
Devaux, B., Vignal, J. P., & Biraben, A. (1995). Intracerebral
potentials to rare target and distractor auditory and visual stimuli. II.
Medial, lateral and posterior temporal lobe. Electroencephalography
and Clinical Neurophysiology, 94, 229-250.
Hansenne, M., Pitchot, W., Moreno, A. G., Zaldua, I. U., &
Ansseau, M. (1996). Suicidal behavior in depressive disorder: an
event-related potential study. Biological Psychiatry, 40, 116-122.
Hegerl, U., & Juckel, G. (1993). Intensity dependence of auditory
evoked potentials as an indicator of central serotonergic neuro-
transmission: A new hypothesis. Biological Psychiatry, 33, 173-187.
Hegerl, U., Gallinat, J. & Juckel, G. (2001). Event-related
potentials: Do they reflect central serotonergic neurotransmission and
do they predict clinical response to serotonin agonists? Journal of
Affective Disorders, 62, 93-100.
Heller, W. (1990). The neuropsychology of emotion:
developmental patterns and implications for psychopathology. In N.L.
Stein, B. Leventhal & T. Trabasso (Eds.), Psychological and
Biological Approaches to Emotion, (pp. 167-211). Hillside, NJ:
Lawrence Earlbaum Associates.
Heller, W. (1993). Neuropsychological mechanisms of individual
differences in emotion, personality and arousal. Neuropsychology, 7,
476-489.
Heller, W., Etienne, M. A., & Miller, G. A. (1995). Patterns of
perceptual asymmetry in depression and anxiety: implications for
neuropsychological models of emotion and psychopathology. Journal
of Abnormal Psychology, 104, 327-333.
Henson, R. N., Rugg, M. D., Shallice, T., & Dolan, R. J. (2000).
Confidence in recognition memory for words: dissociating right
prefrontal roles in episodic retrieval. Journal of Cognitive
Neuroscience, 12, 913-923.
Holmes, A. J., & Pizzagalli, D. A. (2008). Spatiotemporal
dynamics of error processing dysfunctions in major depressive
disorder. Archives of General Psychiatry, 65, 179-188.
Houston, R. J., Bauer, L. O. & Hesselbrock, V. M. (2003).
Depression and familial risk for substance dependence: a P300 study
of young women. Psychiatry Research: Neuroimaging, 124, 49-62.
Houston, R. J., Bauer, L. O., & Hesselbrock, V. M. (2004). P300
evidence of cognitive inflexibility in female adolescents at risk for
recurrent depression. Progress in Neuro-Psychopharmacology &
Biological Psychiatry 28, 529-536.
Jeon, Y.W., & Polich, J. (2003). Meta-analysis of P300 and
22 G.E. Bruder, J. Kayser, C.E. Tenke
schizophrenia: Patients, paradigms, and practical implications.
Psychophysiology, 40, 684-701.
Johnson, R., Jr. (1995). Event-related potential insights into the
neurobiology of memory systems. In F. Boller & J. Grafman (Eds.),
Handbook of Neuropsychology (Vol. 10, pp. 135-164). Amsterdam:
Elsevier.
Johnston, V. S., Miller, D. R., & Burleson, M. H. (1986). Multiple
P3s to eomtional stimuli and their theoretical significance.
Psychophysiology, 23, 684-694.
Juckel, G., Hegerl, U., Molnar, M., Csepe, V., & Karmos, G.
(1999). Auditory evoked potentials reflect serotonergic neuronal
activity--a study in behaving cats administered drugs acting on 5-
HT1A autoreceptors in the dorsal raphe nucleus.
Neuropsychopharmacology, 21, 710-716.
Kaiser, S., Unger, J., Kiefer, M., Markela, J., Mundt, C., &
Weisbrod, M. (2003). Executive control deficit in depression: Event-
related potentials in a Go/Nogo task. Psychiatry Research, 122, 169-
184.
Kalayam, B., & Alexopoulos, G. S. (2003). A preliminary study of
left frontal region error negativity and symptom improvement in
geriatric depression. American Journal of Psychiatry, 160 2054-2056.
Karaaslan, F., Gonul, A. S., Oguz, A., Erdinc, E., & Esel, E.
(2003). P300 changes in major depressive disorders with and without
psychotic features. Journal of Affective Disorders, 73, 283-287.
Kaustio, O., Partanen, J., Valkonen-Korhonen, M., Viinamäki, H.,
Lehtonen, J. (2002). Affective and psychotic symptoms relate to
different types of P300 alteration in depressive disorder. Journal of
Affective Disorders, 71, 43-50.
Kawasaki, T., Tanaka, S., Wang, J., Hokama, H., & Hiramatsu, K.
(2004). Abnormalities of P300 cortical current density in unmedicated
depressed patients revealed by LORETA analysis of event-related
potentials. Psychiatry and Clinical Neurosciences, 58, 68-75.
Kayser, J., Bruder, G. E., Tenke, C. E., Stewart, J. W., & Quitkin,
F. M. (2000). Event-related potentials (ERPs) to hemifield
presentations of emotional stimuli: Differences between depressed
patients and healthy adults in P3 amplitude and asymmetry.
International Journal of Psychophysiology, 36, 211-236.
Kayser, J., Fong, R., Tenke, C. E., & Bruder, G. E. (2003). Event-
related brain potentials during auditory and visual word recognition
memory tasks. Brain Research. Cognitive Brain Research, 16, 11-25.
Kayser, J., & Tenke, C.E. (2003). Optimizing PCA methodology
for ERP component identification and measurement: Theoretical
rationale and empirical evaluation. Clinical Neurophysiology, 114,
2307-2325.
Kayser, J., Tenke, C. (2005). Trusting in or breaking with
convention: towards a renaissance of principal components analysis in
electrophysiology. Clinical Neurophysiology, 116(8), 1747-1753.
Kayser, J., & Tenke, C. E. (2006a). Principal components analysis
of Laplacian waveforms as a generic method for identifying ERP
generator patterns: I. Evaluation with auditory oddball tasks. Clinical
Neurophysiology, 117, 348-368.
Kayser, J., & Tenke, C. E. (2006b). Principal components analysis
of Laplacian waveforms as a generic method for identifying ERP
generator patterns: II. Adequacy of low-density estimates. Clinical
Neurophysiology, 117, 369-380.
Kayser, J., & Tenke, C. E. (2006c). Consensus on PCA for ERP
data, and sensibility of unrestricted solutions. Clinical
Neurophysiology, 117, 703-707.
Kayser, J., Tenke, C. E., & Bruder, G. E. (1998). Dissociation of
brain ERP topographies for tonal and phonetic oddball tasks.
Psychophysiology, 35, 576-590.
Kayser, J., Tenke, C. E., Gates, N. A., & Bruder, G. E. (2007).
Reference-independent ERP old/new effects of auditory and visual
word recognition memory: joint extraction of stimulus- and response-
locked neuronal generator patterns. Psychophysiology, 44, 949-967.
Kayser, J., Tenke, C.E., Gates, N.A., Stewart, J.W., McGrath, P.J.,
Bruder, G.E. (in preparation). Event-related brain potentials (ERPs)
during auditory and visual word recognition memory tasks in affective
disorders.
Kayser, J., Tenke, C., Nordby, H., Hammerborg, D., Hugdahl, K.,
& Erdmann, G. (1997). Event-related potential (ERP) asymmetries to
emotional stimuli in a visual half-field paradigm. Psychophysiology,
34, 414-426.
Kiehl, K. A., Laurens, K. R., Duty, T. L., Forster, B. B., & Liddle,
P. F. (2001). Neural sources involved in auditory target detection and
novelty processing: an event-related fMRI study. Psychophysiology,
38, 133-142.
Kimble, M., Kaloupek, D., Kaufman, M., & Deldin, P. (2000).
Stimulus novelty differentially affects attentional allocation in PTSD.
Biological Psychiatry, 47, 880-890.
Knight, R. (1996). Contribution of human hippocampal region to
novelty detection. Nature, 383, 256-259.
Knight, R. T., & Scabiani, D. (1998). Anatomic bases of event-
related potentials and their rela tionship to novelty detection in humans.
Journal of Clinical Neurophysiology, 15, 3-13.
Knott, V., Lapierre, Y., Griffiths, L., de Lugt, D., & Bakish, D.
(1991). Event-related potentials and selective attention in major
depressive illness. Journal of Affective Disorders, 23, 43-48.
Knott, V. J., & Lapierre, Y. D. (1987). Electrophysiological and
behavioral correlates of pychomotor responsivity in depression.
Biological Psychiatry, 22, 313-324.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1999).
International Affective Picture System (IAPS): Technical manual and
affective ratings. Gainesville, FL: The Center for Research in
Psychophysiology, University of Florida.
Lewis, D. A., Campbell, M. J., Foote, S. L., & Morrison, J. H.
(1986). The monoaminergic innervation of primate neocortex. Human
Neurobiology, 5, 181-188.
Linka, T., Müller, B. W., Bender, S., & Sartory, G. (2004). The
STET intensity dependence of the auditory evoked N1 component as
a predictor of response to Citalopram treatment in patients with major
depression. Neuroscience Letters, 367, 375-378.
Linka, T., Müller, B. W., Bender, S., & Sartory, G., & Gastpar, M.
(2005). The intensity dependence of auditory evoked ERP components
predicts responsiveness to Reboxetine treatment in major depression.
Pharmacopsychiatry, 38, 139-143.
Luck, S.J. (2005). An introduction to the event-related potential
technique. Cambridge: The MIT Press.
Luu, P., Collins, P., & Tucker, D. M. (2000). Mood, personality,
and self-monitoring: Negative affect and emotionality in relation to
frontal lobe mechanisms of error monitoring. Journal of Experimental
Psychology: General, 129, 43-60.
Mao, W., Wang, Y., & Wang, D. (2005). Cognitive impairment in
major depressive disorder revealed by event-related potential N270.
Clinical EEG and Neuroscience, 36, 9-14.
Maris, E. (2004). Randomization tests for ERP topographies and
whole spatiotemporal data matrices. Psychophysiology, 41, 142-151.
Massioui, F., Everett, J., Martin, M. T., Jouvent, R., & Widlöcher,
D. (1996). Attention deficits in depression: an electrophysiological
marker. NeuroReport 7, 2483-2486.
Massioui, F., & Lesèvre, N. (1988). Attention impairment and
psychomotor retardiation in depressed patients: an event-related
potential study. Electroencephalography and Clinical
Neurophysiology, 70, 46-55.
Mayberg, H. S., Brannan, S. K., Mahurin, R. K., Jerabek, P. A.,
Brickman, J. S., Tekell, J. L., Silva, J. A., McGinnis, S., Glass, T. G.,
Martin, C. C., & Fox, P. T. (1997). Cingulate function in depression:
A potential predictor of treatment resonse. NeuroReport, 8, 1057-
1061.
Mecklinger, A., & Ullsperger, P. (1995). The P300 to novel and
target events: A spatiotemporal dipole model analysis. Neuroreport,
7, 241-245.
ERPs in depression 23
Mulert, C., Juckel, G., Brunnmeier, M., Karch, S., Leicht, G.,
Mergl, R., Moller, H. J., Hegerl, U., & Pogarell, O. (2007). Prediction
of treatment response in major depression: integration of concepts.
Journal of Affective Disorders, 98, 215-225.
Muir, W. J., St. Clair, D. M. & Blackwood, D. H. R. (1991). Long-
latency auditory event-related potentials in schziophrenia and in
bipolar and unipolar affective disorder. Psychological Medicine, 21,
867-879.
Nathan, P. J., Segrave, R., Phan, K. L., O'Neill, B., & Croft, R. J.
(2006). Direct evidence that acutely enhancing serotonin with the
selective serotonin reuptake inhibitor citalopram modulates the
loudness dependence of the auditory evoked potential (LDAEP)
marker of central serotonin function. Human Psychopharmacology,
21, 47-52.
Näätänen, R., & Gaillard, A. W. K. (1983). The orienting reflex
and the N2 deflection of the event-related potential (ERP). In A.W. K.
Gaillard & W. Ritter (Eds.), Tutorials in ERP research: Endogenous
components. Amsterdam: North-Holland.
Naumann, E., Bartussek, D., Diedrich, O., & Laufer, M. (1992).
Assessing cognitive and affective information processing functions of
the brain by the means of the late positive complex of the event-related
potential. Journal of Psychophysiology, 6, 285-298.
Neuhaus, A. H., Luborzewski, A., Rentzsch, J., Brakemeier, E. L.,
Opgen-Rhein, C., Gallinat, J., & Bajbouj, M. (2007). P300 is enhanced
in responders to vagus nerve stimulation for treatment of major
depressive disorder. Journal of Affective Disorders, 100, 123-128.
Nunez, P. (1981). Electric fields of the brain. New York: Oxford
University Press.
Nunez, P.L., & Srinivasan, R. (2006). Electric fields of the brain:
The neurophysics of EEG. New York: Oxford University Press.
Nurminen, M., Valkonen-Korhonen, M., Mervaala, E., Paakkonen,
A., Partanen, J., Viinamaki, H., & Lehtonen, J. (2005). Enhanced
attention-dependent auditory processing by electroconvulsive therapy
in psychotic depression. The Journal of ECT, 21, 19-24.
Ochsner, K., & Gross, J. J. (2005). The cognitive control of
emotion. Trends in Cognitive Science, 9, 242-249.
Ogura, C., Nageishi,Y., Omura, F., Fukao, K., Ohta, H.,
Kishimoto, A., & Matsubayashi, M. (1993). Biological Psychiatry, 33,
720-726.
Otto, M. W., Bruder, G. E., Fava, M., Delis, D. C., Quitkin, F. M.,
& Rosenbaum, J. F. (1994). Norms for depressed patients for the
California Verbal Learning Test: associations with depression severity
and self-report of cognitive difficulties. Archives of Clinical
Neuropsychology, 9, 81-88.
Paige, S. R., Fitzpatrick, D. F., Kline, J. P., Balogh, S. E., &
Hendricks, S. E. (1994). Event-related potential amplitude/intensity
slopes predict response to antidepressants. Neuropsychobiology, 30,
197-201.
Paige, S. R., Hendricks, S. E., Fitzpatrick, D. F., Balogh, S., &
Burke, W. J. (1995). Amplitude/intensity functions of auditory event-
related potentials predict responsiveness to bupropion in major
depressive disorder. Psychopharmacology Bulletin, 31, 243-248.
Palomba, D., Angrilli, A., & Mini, A. (1997). Visual evoked
potentials, heart rate responses and memory to emotional pictorial
stimuli. International Journal of Psychophysiology, 27, 55-67.
Pause, B. M., Miranda, A., Nysterud, M., & Ferstl, R. (2000).
Olfactory and emotional stimulus evaluation in patients with major
depression. Zeitschrift für Klinische Psycholologie und Psycho-
therapie, 29, 16-23.
Pause, B. M., Raack, N., Sojka, B., Goder, R., Aldenhoff, J. B., &
Ferstl, R. (2003). Convergent and divergent effects of odors and
emotions in depression. Psychophysiology, 40, 209-225.
Perrin, F., Pernier, J., Bertrand, O., & Echallier, J.F. (1989).
Spherical splines for scalp potential and current density mapping
[Corrigenda EEG 02274, EEG Neurophysiol, 1990, 76, 565].
Electroencephalography and Clinical Neurophysiology, 72, 184-187.
Pierson, A., Ragot, R., Van Hooff, J., Partiot, A., Renault, B., &
Jouvent, R. (1996). Heterogeneity of information-processing
alterations according to dimensions of depression: An event-related
potentials study. Biological Psychiatry, 40, 98-115.
Pizzagalli, D., Pascual-Marqui, R. D., Nitschke, J. B., Oakes, T. R.,
Larson, C. L., Abercrombie, H. C., Schaefer, S. M., Koger, J. V.,
Benca, R. M., & Davidson, R. J. (2001). Anterior cingulate activity as
a predictor of degree of treatment response in major depression:
Evidence from brain electrical tomography analysis. American Journal
of Psychiatry, 158, 405-415.
Pizzagalli, D. A., Peccoralo, L. A., Davidson, R. J., & Cohen, J. D.
(2006). Resting anterior cingulate activity and abnormal responses to
errors in subjects with elevated depressive symptoms: A 128-channel
EEG study. Human Brain Mapping, 27, 185-201.
Polich, J. (2007). Updating P300: An integrative theory of P3a and
P3b. Clinical Neurophysiology, 118, 2128-2148.
Polich, J., & Criado, J. R. (2006). Neuropsychology and
neuropharmacology of P3a and P3b. International Journal of
Psychophysiology, 60, 172-185.
Polich, J., & Herbst, K.L. (2000). P300 as a clinical assay:
rationale, evaluation, and findings. International Journal of
Psychophysiology, 38, 3-19.
Porter, R. J., Gallagher, P., Thompson, J. M., & Young, A. H.
(2003). Neurocognitive impairment in drug-free patients with major
depressive disorder. British Journal of Psychiatry, 1 82, 214-220.
Post, R. M., DeLisi, L. E., Holcomb, H. H., Uhde, T. W., Cohen,
R., & Buchsbaum, M. S. (1987). Glucose utilization in the temporal
cortex of affectively ill patients: Positron emission tomography.
Biological Psychiatry, 22, 545-553.
Purcell, R., Maruff, P., Kyrios, M., & Pantelis, C. (1997).
Neuropsychological function in young patients with unipolar major
depression. Psychological Medicine, 27, 1277-1285.
Röschke, J., & Wagner, P. (2003). A confirmatory study on the
mechanisms behind reduced P300 waves in depression.
Neuropsychopharmacology, 28, S9-S12.
Roth, W. T., Duncan, C. C., Pfefferbaum, A., & Timsit-Berthier,
M. (1986). Applications of cognitive ERPs in psychiatric patients. In
W.C. McCallum, R. Zappoli & F. Denoth (Eds.), Cerebral
Psychophysiology: Studies in Event-Related Potentials, (Suppl. 38, pp.
419-438). Amsterdam: Elsevier.
Ruchsow, M., Grothe, J., Spitzer, M., & Keifer, M. (2002). Human
anterior cingulate cortex is activated by negative feedback: evidence
from event-related potentials in a guessing task. Neuroscience Letters,
325, 203-206.
Ruchsow, M., Herrnberger, B., Beschoner, P., Grön, G., Spitzer,
M., & Kiefer, M. (2006). Error processing in major depressive
disorder: evidence from event-related potentials. Journal of
Psychiatric Research, 40, 37-46.
Ruchsow, M., Herrnberger, B., Beschoner, P., Gron, G., Spitzer,
M., & Kiefer, M.Drevets, W. (2000). Neuroimaging studies of mood
disorders. Biological Psychiatry, 48, 813-829.
Ruchsow, M., Herrnberger, B., Wiesend, C., Gron, G., Spitzer, M.,
& Kiefer, M. (2004). The effect of erroneous responses on response
monitoring in patients with major depressive disorder: A study with
event-related potentials. Psychophysiology, 41, 833-840.
Rugg, M. D., Roberts, R. C., Potter, D. D., Pickles, C. D., & Nagy,
M. E. (1991). Event-related potentials related to recognition memory.
Effects of unilateral temporal lobectomy and temporal lobe epilepsy.
Brain, 114, (5), 2313-2332.
Sams, M., Alho, K., & Näätänen, R. (1983). Sequential effects on
the ERP in discriminating two stimuli. Biological Psychology, 17, 41-
58.
Sandman, C., Gerner, R., O’Halloran, J., & Isenhart, R. (1987).
Event-related potentials and item recognition in depressed,
schizophrenic and alcoholic patients. International Journal of
Pyschophysiology, 5, 215-225.
24 G.E. Bruder, J. Kayser, C.E. Tenke
Sandman, C. A., Vigor-Zierk, C. S., Isenhart, R., Wu, J., & Zetin,
M. (1992). Cardiovascular phase relationships to the cortical event-
related potential of schizophrenic, depresed and normal subjects.
Biological Psychiatry, 32-778-789.
Santosh, P. J., Malhotra, S., Raghunathan, M., & Mehra, Y. N.
(1994). A study of P300 in melancholic depression--correlation with
psychotic features. Biological Psychiatry, 35, 474-479.
Sapolsky, R. M. (2000). Glucocorticoids and hippocampal atrophy
in neuropsychiatric disorders.- Archives of General Psychiatry, 57,
925-935.
Sara, G., Gordon, E., Kraiuhin, C., Coyle, S., Howson, A., &
Meares, R. (1994). The P300 ERP component: an index of cognitive
dysfunction in depression? Journal of Affective Disorders, 31, 29-38.
Schlegel, S., Nieber, D., Herrmann, C., & Bakauski, E. (1991).
Latencies of the P300 component of the auditory event-related
potential in depression are related to the Bech-Rafaelsen Melancholia
Scale but not to the Hamilton Rating Scale for Depression. Acta
Psychiatrica Scandinavica, 83, 438-440.
Shah, P. J., Ebmeier, K. P., Glabus, M. F., & Goodwin, G. M.
(1998). Cortical grey matter reductions associated with treatment-
resistant chronic unipolar depression. Controlled magnetic resonance
imaging study. British Journal of Psychiatry, 172, 527-532.
Shestyuk, A. Y., Deldin, P. J., Brand, J. E., & Deveney, C. M.
(2005). Reduced sustained brain activity during processing of positive
emotional stimuli in major depression. Biological Psychiatry, 57,
1089-1096.
Sidtis, J. J. (1981). The complex tone test: Implication for the
assesment of auditory laterality effects. Neuropsychologia, 19, 103-
112.
Siegle, G. J., Steinhauer, S. R., & Thase, M. E., (2004). Pupillary
assessment and computational modeling of the Stroop task in
depression. International Journal of Psychophysiology, 52, 63-76.
Simons, R., Graham, F., Miles, M., & Chen, X. (2001). On the
relationship of P3a and the novelty-P3. Biological Psychology, 56,
207-218.
Smith, M. E., & Halgren, E. (1989). Dissociation of recognition
memory components following temporal lobe lesions. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 15, 50-
60.
Smith, M. E., Stapleton, J. M., & Halgren, E. (1986). Human
medial temporal lobe potentials evoked in memory and language tasks.
Electroencephalography and Clinical Neurophysiology, 63, 145-159.
Spencer, K., Dien, J., & Donchin, E. (1999). A componential
analysis of the ERP elicited by novel events using a dense electrode
array. Psychophysiology, 36, 409-414.
Steiner, J.E., Lidar-Lifschitz, D., & Perl, E. (1993). Taste and odor:
Reactivity in depressive disorders, a multidisciplinary approach.
Perceptual and Motor Skills, 77, 1331-1346.
Strobel, A., Debener, S., Schmidt, D., Hünnerkopf, R., Lesch, K.
P., & Brocke, B. (2003). Allelic variation in serotonin transporter
function associated with the intensity dependence of the auditory
evoked potential. American Journal of Medical Genetics, 118B, 41-47.
Sumich, A. L., Kumari, V., Heasman, B. C., Gordon, E., &
Brammer, M. (2006). Abnormal asymmetry of N200 and P300 event-
related potentials in subclinical depression. Journal of Affective
Disorders, 92, 171-183.
Tenke, C. E., Bruder, G. E., Towey, J. P., Leite, P., & Sidtis, J. J.
(1993). Correspondence between brain ERP and behavioral asymme-
tries in a dichotic complex tone test. Psychophysiology, 30, 62-70.
Tenke, C. E., & Kayser, J. (2005). Reference-free quantification
of EEG spectra: combining current source density (CSD) and
frequency principal components analysis (fPCA). Clinical Neuro-
physiology, 116, 2826-2846.
Tenke, C.E., & Kayser, J. (2008). ERP generators within the longi-
tudinal fissure: Are putative inverses flawed? Program No. 872.28.
2008 Neuroscience Meeting Planner. Washington, DC: Society for
Neuroscience, 2008. [Available at: http://psychophysiology.cpmc.
columbia.edu/mmedia/sfn2008/sfn2008.pdf]
Tenke, C. E., Kayser, J., Fong, R., Leite, P., Towey, J. P., &
Bruder, G. E. (1998). Response- and stimulus-related ERP
asymmetries in a tonal oddball task: A Laplacian analysis. Brain
Topography, 10, 201-210.
Tenke, C. E., Kayser, J., Shankman, S. A., Griggs, C. B., Leite, P.,
Stewart, J. W., & Bruder, G. E. (2008). Hemispatial PCA dissociates
temporal from parietal ERP generator patterns: CSD components in
healthy adults and depressed patients during a dichotic oddball task.
International Journal of Psychophysiology, 67, 1-16.
Tenke, C.E., Kayser, J., Stewart, J.W., Bruder, G.E. (in press).
Novelty P3 reductions in depression: characterization using principal
components analysis (PCA) of current source density (CSD)
waveforms. Psychophysiology.
Tucker, D., Luu, P., Frishkoff, F., Quiring, J., & Poulsen, C.
(2003). Frontolimbic response to negative feedback in clinical
depression. Journal of Abnormal Psychology, 112, 667-678.
Turetsky, B. I., & Fein, G. (2002). Alpha2-noradrenergic effects
on ERP and behavioral indices of auditory information processing.
Psychophysiology, 39, 147-157.
Tzur, G., & Berger, A. (2007). When things look wrong: theta
activity in rule violation. Neuropsychologia, 45, 3122-3126.
Umbricht, D., Koller, R., Schmid, L., Skrabo, A., Grübel, C.,
Huber, T., & Stassen, H. (2003). How specific are deficits in mismatch
negativity generation to schizophrenia? BiologicalPsychiatry, 53,
1120-1131.
Urretavizcaya, M., Moreno, I., Benlloch, L., Cardoner, N.,
Serrallonga, J., Menchon, J. M., & Vallejo, J. (2003). Auditory event-
related potentials in 50 melancholic patients: Increased N100, N200
and P300 latencies and diminished P300 amplitude. Journal of
Affective Disorders, 74, 293-297.
Wagner, P., Röschke, J., Fell, J., & Frank, C. (1997). Differential
pathophysiological mechanisms of reduced P300 amplitude in
schizophrenia and depression: a single trial analysis. Schizophrenia
Research, 25(3), 221-229.
Van Veen, V., & Carter, C. (2002). The timing of action-
monitoring processes in the anterior cingulate cortex. Journal of
Cognitive Neuroscience, 14, 593-602.
Vandoolaeghe, E. van, Hunsel, F., Nuyten, D., & Maes, M. (1998).
Auditory event related potentials in major depression: prolonged P300
latency and increased P200 amplitude. Journal of Affective Disorders,
48, 105-113.
Wang, C., Ulbert, I., Schomer, D.L., Marinkovic, K., & Halgren,
E. (2005). Responses of human anterior cingulate cortex micro-
domains to error detection, conflict monitoring, stimulus-response
mapping, familiarity, and orienting. Journal of Neuroscience, 25, 604-
613.
Yanai, I., Fujikawa, T., Osada, M., Yamawaki, S., & Touhouda, Y.
(1997). Changes in auditory P300 in patients with major depression
and silent cerebral infarction. Journal of Affective Disorders, 46, 263-
271.
Zatorre, R. J., Evans, A. C., Meyer, E., & Gjedde, A. (1992).
Lateralization of phonetic and pitch discrimination in speech
processing. Science, 256, 846-849.
Zhang, Y., Hauser, U., Conty, C., Emrich, H. M., & Dietrich, D.
E. (2007). Familial risk for depression and P3b component as a
possible neuro cognitive vulnerability marker. Neuropschobiology, 55,
14-20.
