You are on page 1of 16

April 27, 2005 14:20 WSPC/179-JIN 00071

Journal of Integrative Neuroscience, Vol. 4, No. 1 (2005) 145–159
c Imperial College Press

Research Report

INTEGRATIVE ASSESSMENT OF BRAIN AND COGNITIVE
FUNCTION IN POST-TRAUMATIC STRESS DISORDER

MELINDA D. VELTMEYER∗
Cognitive Neuroscience Laboratory and School of Psychology
Flinders University of South Australia
Adelaide, SA 5042, Australia
GPO Box 2100, Adelaide, South Australia. 5001
melinda.veltmeyer@flinders.edu.au

C. RICHARD CLARK
Cognitive Neuroscience Laboratory and School of Psychology
Flinders University, Adelaide, SA 5042, Australia

ALEXANDER C. McFARLANE
Department of Psychiatry, The University of Adelaide
Queen Elizabeth Hospital, Woodville, SA 5011, Australia

KIM L. FELMINGHAM
The Brain Dynamics Center, University of Sydney and
Westmead Hospital, NSW 2145, Australia

RICHARD A. BRYANT
School of Psychology, University of New South Wales
NSW 2052, Australia

EVIAN GORDON
The Brain Resource Company and the Brain Resource International Database
Ultimo, NSW 2007, Australia
Convenor, Integrative Neuroscience, Brain Dynamics Center, University of Sydney and
Westmead Hospital, NSW 2145, Australia
Discipline of Psychological Medicine, University of Sydney
NSW 2006, Australia

Received 7 February 2005
Revised 21 February 2005

The present study combined neuropsychological and electrophysiological measures to
obtain a comprehensive profile of the everyday attentional and memory dysfunction
reported in PTSD. The event-related potential (ERP) literature has consistently found

∗ Corresponding author.

145
April 27, 2005 14:20 WSPC/179-JIN 00071

146 Veltmeyer et al.

abnormalities in late components (N2, P3) reflecting working memory (WM) function.
However, the neuropsychological profile reported in the literature has considerable vari-
ation. The present study examined ERP activity in 33 PTSD participants and matched
controls during a standard two-tone auditory oddball task. Neuropsychological assessment
was carried out using a task battery assessing a wide range of cognitive functions. Consis-
tent with previous work, the PTSD group showed delayed N2 latency and reduced P3 target
amplitude, together with slower and less accurate target detection. Scalp topography pro-
vided evidence of widespread abnormality during WM function, but with strongest effects
broadly over the left hemisphere. Neuropsychological testing found concomitant difficul-
ties on factorial measures of verbal memory retention/access and sustained attention but
enhanced performance on measures of immediate recall. This integrative pattern of effects
reflects a specific impairment in the operation of working memory systems that guide ongo-
ing, planned behavior and that facilitate the acquisition and retention of new memories.

Keywords: ERPs; PTSD; neuropsychology; P3, N2; working memory.

1. Introduction
Current diagnostic criteria for PTSD emphasize symptoms such as hyperarousal,
intrusive recollections of the traumatic event and avoidance of trauma-related stim-
uli. However, it is also apparent that sufferers of PTSD have impaired everyday
memory and attention. Previous research using event-related potentials (ERPs),
which reflect the dynamics of brain activity associated with the controlled process-
ing of task-related information, has demonstrated disordered cognitive processing
of neutral (i.e., trauma-neutral) stimuli. For example, a diminished P3 component
to periodic target stimuli in a three-tone, auditory target detection task has been
found [21], which suggests a reduction in the resources allocated to attentional pro-
cesses [25], as well as a prolonged N2 latency, indicative of a slowing of cerebral
response to relevant environmental change.
Whilst subsequent studies have replicated and extended a number of these find-
ings [12, 13, 22, 23], there has been relative inconsistency in accounts of neuropsycho-
logical function in PTSD: some studies report global cognitive impairments, some
report relatively specific deficits, whilst others report no deficits at all [see 16 for a
review]. For example, dissociated impairments over groups have been found in ver-
bal abilities [6], inhibitory processes [32] and in attention [1, 14]. Thus, the key aim
of this study was to combine and examine the relationship between neuropsycholog-
ical and cognitive electrophysiological measures in the same group of experimental
participants.
The variability in results across the respective ERP and neuropsychological lit-
eratures may relate to the fact that no study has reported both sets of measures
on the same participant population. Specific hypotheses relating to ERP measures
predicted that N2 latency and P3 amplitude would be delayed and attenuated,
respectively, in the PTSD group relative to controls. It was also predicted that the
normality of performance on behavioral measures of working memory function and
operation would match that obtained for the ERP componentry.
April 27, 2005 14:20 WSPC/179-JIN 00071

Integrative Assessment of Brain and Cognitive Function in PTSD 147

2. Method
2.1. Participants
The participants consisted of 35 PTSD participants (17 Females, and 18 Males;
with an average age of 39.97 ± 10.85 years) and 140 (72 Males, 68 Females) controls
matched for gender and age, with an average age of 39.67± 11.27 years. Control par-
ticipants had normal vision, hearing and dexterity, spoke English as a first language,
and had no personal histories of mental and neurological disorders, including sub-
stance abuse and serious medical problems, such as cancer, heart disease, HIV and
Hepatitis. Control participants were also excluded on the basis of a family history
of psychiatric illness. PTSD participants met the criteria for a primary diagnosis of
PTSD according to DSM IV. The Clinician Administered PTSD Scale (CAPS) [2]
was used to confirm the diagnosis. Clinical participants were symptomatic for at least
6 months and had experienced the precipitating trauma within the last 10 years.
All PTSD participants were adults at onset and there were a range of precipitating
traumas, as shown in Table 1. Participants were also excluded if they had a Comor-
bid Axis 1 Anxiety Disorder, which developed prior to the onset of PTSD, a history
of Psychosis, Substance Dependence or were taking psychotropic medications other
than SSRIs.

2.2. Tasks and procedure
All PTSD and control participants completed the test protocol deployed for the
Brain Resource International Database (BRID). This database reflects a recent
international effort [17, 18] undertaken to develop a standardized cognitive and psy-
chophysiological battery of tests capable of tapping the primary cognitive domains.
The present study reports on a subset of the tasks completed as part of the battery,

Table 1. Breakdown of the type of trauma that precipitated PTSD
in study participants.

Trauma Type Male Female Total

Witnessed shooting 1 0 1
Physical assault 1 2 3
Police work 3 1 4
Motor vehicle accident 6 6 12
Witness to attempted suicide 0 1 1
Rock climbing accident 0 1 1
Workmate killed 1 0 1
Threatened with weapon 3 2 5
Rape 0 1 1
Observed dead bodies 1 0 1
Witness to someone being killed in accident 0 1 1
Fall 0 1 1
April 27, 2005 14:20 WSPC/179-JIN 00071

148 Veltmeyer et al.

including the auditory Oddball task and the “Cognitive test battery”. All partic-
ipants completed the tasks of the full battery in the same order over a period of
three hours.

2.2.1. ERP testing
Auditory Oddball task: Participants were presented with a series of high and low
tones via headphones, at 75 dB and lasting for 50 ms, with an ISI of 1 s. Rise and fall
times of the tones was 5 ms. Participants were instructed to press buttons with the
index finger of each hand in response to “target” tones (presented at 1000 Hz). They
were asked not to respond to ‘background’ tones (presented at 500 Hz). Speed and
accuracy of response were equally stressed in the task instructions. The background
and target tones were presented in a quasi-random order, with the only constraint
being that two targets cannot appear consecutively. During the task, participants
were required to sit still, relax and focus on a red dot on a computer monitor. The
duration of the task was six minutes.

2.2.2. Cognitive test battery
Participants were seated in a sound and light attenuated room, in front of a touch-
screen computer (NEC MultiSync LCD 1530 V). The cognitive tests were admin-
istered in a fixed order using pre-recorded task instructions (via headphones), and
the touch-screen computer was used for answers. The test battery consisted of the
following tests: motor tapping, choice reaction time, continuous performance test,
letter fluency, executive maze, span of visual memory, digit span, verbal list-learning
and recall, verbal interference and switching of attention. This computerized battery
has been shown to have good validity and reliability (see www.brainresource.com).
The measures are described in greater detail below.

Motor tapping task: Participants were required to tap a circle on the touch-screen
with their index finger, as fast as possible for 60 seconds. The dependant variables
were the number of taps, time between taps and the standard deviation of time
between taps.

Choice RT task: Participants were required to attend to the computer screen as one
of four circles was illuminated. Immediately following presentation, the participant
then had to touch the illuminated circle as quickly as possible. Twenty trials were
administered, with a random delay between trials of 2–4 seconds. The dependent
variable was the mean reaction time (RT) across trials.

Letter fluency: Participants were required to generate words that began with the
letters F, A and S. Sixty seconds were allowed for each letter and proper nouns were
not allowed. A second test involved the total number of animal names that could
be generated in sixty seconds. The total number of correct words generated across
April 27, 2005 14:20 WSPC/179-JIN 00071

Integrative Assessment of Brain and Cognitive Function in PTSD 149

the three letter trials and the number of animals generated were the two dependent
measures.

Sustained attention (vigilance) task: This is a continuous performance task. A series
of letters (B, C, D or G) were presented to the participant on the computer screen
(for 200 ms), separated by an interval of 2.5 seconds. If the same letter appeared twice
in a row, the participant was required to press buttons with the index finger of each
hand. Speed and accuracy of response were equally stressed in the task instructions.
There were 125 stimuli presented in total, 85 being non-target letters, 20 being target
letters (i.e., repetitions of the previous letter) and 20 other, unrelated stimuli, which
are not reported in this study. The dependent variables were RT, reaction time
variability (RTV), number of false positives and number of false negatives.

Maze: The participant was presented with a grid (8 × 8 matrix) of circles on the
computer screen. The object of the task was to identify the hidden path through
the grid, from the beginning point at the bottom of the grid to the end point at
the top. The participant was able to navigate around the grid by pressing arrow
keys. The participant was presented with one tone (and a red cross at the bottom
of the screen) if they made an incorrect move and a different tone (and a green
tick at the bottom of the screen) if they made a correct move. The purpose of the
task was to assess how quickly the participant learned the hidden route through
the maze and their ability to remember that route. The trial ended when the par-
ticipant completed the maze twice without error, or after 7 minutes had elapsed.
The dependent variables were the number of errors, the number of trials required
to complete the maze without error on two consecutive trials, and the trial by error
gradient.

Digit span: Participants were presented with a series of digits (e.g., 4, 2, 7), presented
individually for 500 ms and separated by a 1 s interval. The participant was then
immediately asked to enter the digits on a numeric keypad on the touch-screen,
either in forward order (forward digit span task) or backwards (reverse digit span
task). The number of digits in each sequence was gradually increased from 3 to 9.
The dependent measure was the maximum number of digits the participant recalled
without error in each task.

Span of visual memory: This test was a visual analogue of a digit span test. Partic-
ipants were presented with squares arranged in a random pattern on the computer
screen. The squares were highlighted in a sequential order on each trial. Participants
were required to repeat the order in which the squares were highlighted. The total
number of correct sequences was the dependent variable.

Verbal learning and memory task: The participants were read a list of 12 words,
which they were asked to memorize. The list contained 12 words from the English
language. Words were closely matched on concreteness, number of letters and
April 27, 2005 14:20 WSPC/179-JIN 00071

150 Veltmeyer et al.

frequency. The list was presented 4 times in total and the participant was required
to recall as many words as possible from the list in any order after each presenta-
tion. The participant was then presented with a list of distracter words and asked to
recall as many of these as possible in any order. The participant was then asked to
recall the 12 words from the original list, both immediately after the distracter list
(immediate free recall trial) and after a 20 minute delay (delayed free recall trial).
The dependent variables were the number of words correctly recalled across the four
initial learning trials, the number of words recalled on the long delay free recall trial,
the number of words correctly recognized, the number of words correctly rejected in
the recognition trial, the learning rate and the number of “out of the blue” words
(words found on neither list) in trials one to four.

Verbal interference task: This test was a computerized modification of the Stroop
test [15]. Participants were presented with color words printed in incongruent ink
and were required to nominate the name of the word in the first trial and the color
of the word in the second. Words were presented one at a time over a one-minute
interval. The dependent variables were the number of correct trials on each task.

Switching of attention task: This modified version of the Trail Making Test [26]
consisted of two parts. On the first part (Part 1), twenty-five numbers, in circles, were
placed on the touch-screen as a random array and the participant was instructed
to press them in the correct order (i.e., 1-2-3- etc). This tests the basic ability to
maintain attention during a simple task. The second part (Part 2) required the
connecting of numbers and letters in an ascending but alternating sequence (i.e.,
1-A-2-B etc). For this purpose, the numbers 1–13 and the letters A–L were presented
in circles as a random array on the touch-screen. The dependent variables were the
time to complete each task and the number of errors in each task.

2.2.3. EEG measurement
Participants were seated in a sound and light attenuated room, set at an ambient
temperature of 24◦ C. An electrode cap was used to acquire data from the Fp1, Fp2,
F7, F3, Fz, F4, F8, FC3, FCz, FC4, T3, C3, Cz, C4, T4, CP3, CPz, CP4, T5, P3, Pz,
P4, T6, O1, Oz and O2 electrode sites (32 channels; Nuamps; 10–20 International
system). Data was recorded relative to the virtual ground, but referenced offline
to linked mastoids. Horizontal eye movements were recorded with electrodes placed
1.5 cm lateral to the outer canthus of each eye. Vertical eye movements were recorded
with electrodes placed 3 mm above the middle of the left eyebrow and 1.5 cm below
the middle of the left bottom eye-lid. Skin resistance was <5 kOhms. A continuous
acquisition system was employed and data was EOG corrected offline. The sampling
rate was 500 Hz. A low pass filter with attenuation of 40 dB per decade above 100 Hz
was employed prior to digitization.
April 27, 2005 14:20 WSPC/179-JIN 00071

Integrative Assessment of Brain and Cognitive Function in PTSD 151

2.2.4. ERP data scoring and reduction
ERP data were extracted for target stimuli in the auditory oddball task using a
window from −300 ms to 700 ms (relative to the stimulus onset). Data were fil-
tered with a 25 Hz low-pass filter, and baselined relative to the 300 millisecond pre-
stimulus window. Peak and latency measures for ERP components were picked at
all scalp sites within specified window limits from each participant’s averaged wave-
forms using an automated algorithm. P3 was defined as the most positive point
between 300–600 ms, post-stimulus onset. N2 was defined as the most negative point
between 200–325 ms preceding the P3 component. Missing values were replaced with
group means. The total number of missing values replaced was less than 2% of the
overall data.

2.2.5. ERP statistical analyses
ERP data were analyzed as a series of analyses of variance with group (PTSD,
non-PTSD) as a between participants factor. A separate analysis of variance was
performed for each ERP component (N2 target latency, P3 target amplitude) with
each of the 26 electrodes included as a within-participants variable. Analyses were
performed on the full sample (N = 33) and a sub-sample of unmedicated participants
(N = 20).
Corrections for multiple comparisons of data from the 26 scalp electrodes were
based on the degree of correlation across the electrode topography. This was
determined using a principal components analysis over a 500 ms window of activ-
ity post-stimulus from the 26 scalp electrodes. It was shown that three spatial
component factors accounted for over 95% of the variance within each group and
condition (target and background). This value was used in the calculation of a
Bonferroni correction for multiple comparisons over the scalp [13], which produced
a maximum acceptable probability level of 0.01. A series of correlational analy-
ses were also performed to determine whether there was a relationship between
abnormalities of P3 amplitude and PTSD symptoms, adjusted significance levels
of p < 0.01.

2.2.6. Behavioral data scoring, reduction and analysis
Analyses of variance was performed on behavioral data from the Oddball task to
compare performance across the PTSD and non-PTSD groups.
The 33 measures from the 12 separate neuropsychological tasks were subjected
to principal components analysis with Varimax rotation over the data obtained from
control and PTSD participants, which yielded 10 factors. Analyses of variance were
conducted on refined factor scores derived via the standard regression method. The
meaning attributed to each factor was based on the mix of variables that contributed
individually to that factor with a loading greater than 0.50.
April 27, 2005 14:20 WSPC/179-JIN 00071

152 Veltmeyer et al.

3. Results
3.1. Clinical variables
Table 2 shows the clinical scores for the PTSD participants for whom clinical data
was available.

3.2. Behavioral measures from the oddball task
A MANOVA performed on behavioral measures from the oddball task indicated
significant increases in PTSD relative to controls for RT, RTV and number of false
negative errors, but no differences in false positive errors, as shown in Table 3.

3.3. Electrophysiological measures (ERPs)
Two of the 35 participants had insufficient ERP data to be included in the analysis.
The related matched controls (N = 8) were thus also removed from the analysis.

Table 2. Clinical variables for PTSD
participants (N = 29).

Mean SD

CAPSa
Re-experiencing 22.07 8.09
Avoidance 14.38 4.50
Numbing 21.43 6.88
Arousal 27.66 6.34
Total 85.89 20.39
b
CADSS
Amnesia 1.29 1.19
Depersonalization 2.43 2.96
Derealization 4.86 4.95
a
Clinician Administered PTSD Scale [2].
b
Clinician Administered Dissociative
States Scale [3].
SD = Standard Deviation.

Table 3. Behavioral performance on the Oddball task by the PTSD and
control groups.

PTSD (N = 33) Control (N = 129) F Sig.

Mean SD Mean SD

Average RT (ms) 379.38 71.47 333.94 39.67 23.07 <0.001
SD of RT (ms) 81.84 34.25 54.48 15.15 45.42 <0.001
False positives 0.63 1.50 0.52 0.84 0.28 0.599
False negatives 0.47 0.98 0.11 0.32 11.55 0.001
SD = Standard Deviation.
April 27, 2005 14:20 WSPC/179-JIN 00071

Integrative Assessment of Brain and Cognitive Function in PTSD 153

Control PTSD PTSD vs Controls
p-values
N2 latency
260

Ms
P<.001

P<.01
220
P3 amplitude P<.05
14
P<.10

NS
Uv

-14

Fig. 1. Topographic maps for N2 latency and P3 amplitude to target tones in the Oddball task.

P3 target amplitude: MANOVA showed there was a significant main effect of
group (F 1, 161 = 2.44, p < 0.001) that was evident broadly across the scalp (see
Fig. 1). The effect was strongest in the left hemisphere, particularly over prefrontal,
left centroparietal and temporal regions. There was also a clear topographic focus
over the right fronto-temporal region.
N2 target latency: There was a significant main effect of group (F1, 161 =
1.60, p = 0.04), with mean latency delayed in the PTSD group relative to con-
trols. The results for individual electrodes failed to identify particular regions of
significant difference (see Fig. 1).

3.4. Cognitive test battery
Ten factors explaining 67.67% of the variance in the sample were obtained. Analysis
of variance was conducted on those ten factors that each accounted for more than
3% of the variance in their own right. PTSD differed from controls in three of these
factors (see Table 4).

Table 4. Significant differences between PTSD and controls on the factors derived from the
cognitive test battery.

Factors %VE PTSD (N = 35) Controls (N = 140) Significance
Mean Factor Score Mean Factor Score
Mean SD Mean SD F p

Immediate verbal recall 2.37 0.45 0.70 −0.11 1.03 8.42 0.004
Verbal memory retention 9.09 −0.44 1.48 0.11 0.80 8.60 0.005
Vigilance 2.33 0.66 1.42 −0.17 0.79 19.92 < 0.001
%VE = percentage of variance explained by factor.
SD = Standard Deviation.
April 27, 2005 14:20 WSPC/179-JIN 00071

154 Veltmeyer et al.

The first of the differentiating factors reflected performance on the learning
and immediate recall trials (trials 1 to 6) of the verbal memory task (factor load-
ings: 0.72–0.92) and was taken to reflect verbal learning and short term memory.
The second factor was dominated by delayed recall, recognition memory and false
positive scores on the verbal recall task (factor loadings: 0.59, 0.93, −0.93 respec-
tively). This factor was interpreted to reflect verbal memory retention/access. The
third factor was dominated by scores on the sustained attention task (factor loadings
0.64–0.85), which suggests this factor reflects vigilance.
The PTSD group showed improved performance relative to controls on imme-
diate verbal memory, but impaired performance on the other two factors (verbal
memory retention/access and vigilance).

3.5. Correlational analyses
With respect to the hemispheric asymmetry observed in P3 amplitude, correlational
analyses were conducted between mean left hemisphere P3 peak amplitude, mean
right hemisphere peak P3 amplitude, P3 asymmetry [(left hemisphere P3 − right
hemisphere P3)/(left hemisphere P3 + right hemisphere P3)], and CAPS total and
sub-component scores (N = 35). As there was a significant main effect, but no
strong regional effects in N2 latency, correlations were also conducted between the
average N2 latency across all sites and clinical measures. There were no significant
correlations between clinical and ERP measures.
Correlational analyses were also conducted between behavioral (RT, RTV, num-
ber of missed targets) and ERP measures from the Oddball task. The only significant
correlation was between left hemisphere P3 and RTV in the PTSD group (r = −0.49,
p = 0.004). There were trends (r = −0.40, p < 0.05) between left hemisphere P3
amplitude and RT and also between right hemisphere P3 amplitude, RT and RTV in
the PTSD group. These relationships were not present in control participants. There
was also a significant relationship between N2 target latency and RTV (r = 0.47,
p = 0.006) in the PTSD group, but not in control participants.
There were no relationships between the electrophysiological measures and cog-
nitive factors found to be abnormal in PTSD.

4. Discussion
Combined neuropsychological and electrophysiological measures were obtained to
provide a more comprehensive profile of the everyday attentional and memory
dysfunction reported in PTSD. From a neuropsychological viewpoint, there was
impaired performance in PTSD on measures reflecting the long term retention of
(or access to) newly acquired auditory verbal material and the capacity for sus-
tained, as opposed to immediate, attention. This was matched by abnormalities in
the N2 and P3 components of the event-related potential, as reported previously
April 27, 2005 14:20 WSPC/179-JIN 00071

Integrative Assessment of Brain and Cognitive Function in PTSD 155

[12, 13, 21–23]. A delay in the N2 component has been argued to reflect delay in
the speed of detection of significant change in the environment, whilst reduction in
the amplitude of the P3 component reflects attenuation of the cerebral resources
allocated to the related working memory operations [13, 21]. Each of these outcomes
will now be discussed in more detail.

4.1. P3 target amplitude
The observed reduction in P3 was broadly evident across both hemispheres of the
scalp, with specific foci over left prefrontal, left centro-parietal left temporal and
right fronto-temporal regions. This can be related to recent PET work that iden-
tified localized cortical regions of abnormal working memory function in PTSD
[9, 27]. These studies found abnormal patterns of activation in the fronto-parietal
networks commonly associated with working memory operations [e.g., 7]. Corti-
cal network activation was relatively greater in the bilateral parietal lobes and
the left precentral gyrus whilst lower in the medial frontal lobe and right tem-
poral region. At the same time, cerebral activity was generally lower than normal
in both left dorsolateral prefrontal and left inferior parietal regions [9]. Thus, the
findings of the present study arguably reflect abnormal function involving the activ-
ity of the widely distributed cortical networks associated with working memory
function.
At the same time, the P3 effects in PTSD were clearly greater over the left
hemisphere. Thus, it may be that there is some lateralized aspect to the dysfunction
observed in PTSD. The P3 is thought to originate from interactions between frontal
and temporal-parietal regions, which include the hippocampus [25], the pathology
of which has been implicated in PTSD [see 20 for a review]. Several structural MRI
studies have found hippocampal pathology is lateralized to one side or the other.
For example, left lateralized hippocampal atrophy has been found in a group of
women [28] and a mixed-gender group [5], both with PTSD secondary to child-
hood sexual or physical abuse. On the other hand, right sided [4] and bilateral
[19] hippocampal atrophy have been found in men with combat related PTSD.
Alternatively, white matter damage has been found in some PTSD groups [8, 33],
and decreased P3 amplitudes have been associated with smaller callosal fibre tracts
[see 25 for a review]. A tendency towards greater lateralization of brain function
has been observed secondary to developmental trauma, and it has been suggested
that this may be related to the white matter atrophy found in this population
[29]. It has also been suggested that noradrenergic dysfunction may contribute to
P3 amplitude attenuation in PTSD [21]. Interestingly, primate studies have found
that noradrenergic antagonism attenuates P3 amplitude in the parietal, but not
frontal or temporal regions [24]. In this study, strong effects were found in pari-
etal regions, but equally strong effects were found in frontal and temporal regions,
suggesting that noradrenaline dysfunction is not responsible for P3 attenuation
in PTSD.
April 27, 2005 14:20 WSPC/179-JIN 00071

156 Veltmeyer et al.

4.2. N2 target latency
Previous research has been somewhat inconsistent with respect to the normality of
N2 latency in PTSD. For example, whilst a delayed N2 has been reported by some
groups [13, 21], others have found this component to be unaffected in PTSD [23].
This variation may be related to medication effects, since this factor often tends to
be uncontrolled in many studies, as in the present study. Nevertheless, the present
study replicated a delay for this ERP component in PTSD.

4.3. Behavioral measures
The PTSD group showed delayed and more variable RT. Previous research [e.g.,
13] has found that abnormalities in RT and N2 latency were related. The present
study found variable reaction times were related to both delayed N2 latency and
reduced P3 amplitude. This suggests that these psychophysiological indicators may
predict lability in behavioral measures. The PTSD group did not make more false
positive errors than controls, which is inconsistent with the theory that cognitive
dysfunction in PTSD is primarily related to impaired inhibitory processes [32]. The
PTSD group did make more errors of omission, indicative of deficient attention.

4.4. Cognitive test battery
The findings here are consistent with research that has found specific, rather than
global, cognitive deficits in PTSD. Enhanced immediate recall in the presence of
impaired delayed recall is consistent with previous research [34] and suggests that
memory problems in PTSD occur during the consolidation stage. In addition, the
finding of impaired vigilance is consistent with the common observation in PTSD
of a difficulty with the maintenance of attention to everyday matters over extended
periods of time. This difficulty with managing everyday demands over sustained
periods contrasts with the heightened attention of PTSD sufferers towards new
material that might constitute environmental threat [30]. This distinction might
also explain why immediate recall is above normal in PTSD, whilst the retention of
ongoing mental set is impaired during sustained mental activity.

4.5. Conclusions
This study combined neuropsychological and electrophysiological measurement in
the investigation of everyday cognition in PTSD. Previous findings of delayed N2
amplitude and attenuated P3 amplitude were replicated. In addition, topographic
analysis of brain event-related potential patterns provided evidence of widespread
abnormality across the scalp during working memory function, implicating multi-
ple focal sources of disturbed cortical network activity, but with strongest effects
broadly over the left hemisphere. This concurred with the profile of abnormality
provided by neuropsychological testing on factorial measures of verbal memory
April 27, 2005 14:20 WSPC/179-JIN 00071

Integrative Assessment of Brain and Cognitive Function in PTSD 157

retention/access and vigilance. In essence, this integrative pattern of effects reflects a
specific impairment in the operation of working memory systems that guide ongoing,
planned behavior and that facilitate the acquisition and retention of new episodic
memories.
Caution must still be exercised, however, in attributing this pattern of findings
solely to PTSD symptomatology. P3 amplitude, left lateralized and frontal dysfunc-
tion have been observed in other clinical populations, such as those with depression
[e.g., 10, 11]. ERP abnormalities have also been observed with the use of psychotropic
medications [e.g., 31]. Symptoms of depression and concurrent medication use were
not assessed in this study. It may be that concurrent depressive symptoms, which
are common in PTSD populations, were related to altered topography and impaired
performance. This study clearly indicates, however, that an integrative approach
to the assessment of brain and cognitive function adds considerable value to the
understanding of PTSD and has the potential to assess subtle brain abnormalities
in clinical populations.

Acknowledgments
The authors wish to acknowledge the Brain Resource International Database
(www.BrainResource.com) scientists and the study participants. Support was also
provided for this work by an NHMRC Program Grant (No. 300403) to R Bryant,
AC McFarlane, D Silove, M Creamer and CR Clark.

References
[1] Beers SR, De Bellis MD, Neuropsychological function in children with maltreatment-
related posttraumatic stress disorder, Am J Psychiat 159(3):483–486, 2002.
[2] Blake D et al., The development of a clinician-administered PTSD scale, J Trauma
Stress 8(1):75–90, 1995.
[3] Bremner J et al., Measurement of dissociative states with the Clinician-Administered
Dissociative States Scale (CADSS), J Trauma Stress 11(1):125–36, 1998.
[4] Bremner J et al., MRI-based measurement of hippocampal volume in patients with
combat-related posttraumatic stress disorder, Am J Psychiat 152(7):973–981, 1995.
[5] Bremner J et al., Magnetic resonance imaging-based measurement of hippocampal vol-
ume in posttraumatic stress disorder related to childhood physical and sexual abuse —
a preliminary report, Biol Psychiat 41(1):23–32, 1997.
[6] Bremner J et al., Deficits in verbal declarative memory function in women with child-
hood sexual abuse-related posttraumatic stress disorder, J Nerv Ment Dis 192(10):
643–9, 2004.
[7] Cabeza R, Nyberg L, Imaging cognition II: An empirical review of 275 PET and fMRI
studies, J Cogn Neurosci 12(1):1–47, 2000.
[8] Canive JM et al., MRI reveals gross structural abnormalities in PTSD, Ann NY Acad
Sci 821(1):512–515, 1997.
April 27, 2005 14:20 WSPC/179-JIN 00071

158 Veltmeyer et al.

[9] Clark C et al., Cerebral function in posttraumatic stress disorder during verbal working
memory updating: a positron emission tomography study, Biol Psychiat 53(6):474–
81, 2003.
[10] Debener S et al., Is resting anterior EEG alpha asymmetry a trait marker for depres-
sion? Findings for healthy adults and clinically depressed patients, Neuropsychobiology
41(1):31–7, 2000.
[11] Diego M, Field T, Hernandez-Reif M, CES-D depression scores are correlated with
frontal EEG alpha asymmetry, Depress Anxiety 13(1):32–7, 2001.
[12] Felmingham K et al., Event-related potential dysfunction in posttraumatic stress dis-
order: the role of numbing, Psychiat Res 109(2):171–9, 2002.
[13] Galletly C et al., Working memory in posttraumatic stress disorder-an event-related
potential study, J Trauma Stress 14(2):295–309, 2001.
[14] Gilbertson M et al., Multivariate assessment of explicit memory function in combat
veterans with posttraumatic stress disorder, J Trauma Stress 14(2):413–32, 2001.
[15] Golden C, Stroop Color and Word Task: A Manual for Clinical and Experimental Uses,
Stoeling Co., Wood Dale, IL, 1978.
[16] Golier J, Yehuda R, Neuropsychological processes in post-traumatic stress disorder,
Psychiat Clin N Am 25(2):295–315, vi, 2002.
[17] Gordon E, Integrative neuroscience and psychiatry, Neuropsychopharmacol 28(1):2–
8, 2003.
[18] Gordon E, Integrative neuroscience in psychiatry: the role of a standardized database,
Australas Psychiat 11(2):156–163, 2003.
[19] Gurvits T et al., Neurological status of Vietnam veterans with chronic posttraumatic
stress disorder, J Neuropsych Clin N 5(2):183–188, 1993.
[20] Hull AM, Neuroimaging findings in post-traumatic stress disorder: Systematic review,
Brit J Psychiat 181(2):102–110, 2002.
[21] McFarlane AD et al., Abnormal stimulus processing in posttraumatic stress disorder,
Biol Psychiat 34(5):311–20, 1993.
[22] Metzger L et al., Auditory event-related potentials to tone stimuli in combat-related
posttraumatic stress disorder, Biol Psychiat 42(11):1006–15, 1997.
[23] Metzger L et al., Evidence for diminished P3 amplitudes in PTSD, Ann NY Acad Sci
821(1):499–503, 1997.
[24] Pineda J, Westerfield M, Monkey P3 in an “oddball” paradigm: pharmacological sup-
port for multiple neural sources, Brain Res Bull 31(6):689–96, 1993.
[25] Polich J, Detection of change: event-related potential and fMRI findings, in Theoretical
overview of P3a and P3b, J. Polich, Editor. 2003. pp. 83–98.
[26] Reitan R, Validity of the Trail Making test as an indicator of organic brain damage,
Percept Motor Skill 8:271–276, 1958.
[27] Shaw M et al., Abnormal functional connectivity in posttraumatic stress disorder, Neu-
roimage 15(3):661–74, 2002.
[28] Stein M et al., Hippocampal volume in women victimized by childhood sexual abuse,
Psychol Med 27(4):951–9, 1997.
[29] Teicher M et al., Developmental neurobiology of childhood stress and trauma, Psychiat
Clin N Am 25(2):397–426, vii–viii, 2002.
[30] Van Der Kolk B, The psychobiology and psychopharmacology of PTSD, Hum Psy-
chopharmacol 16(S1):S49–S64, 2001.
April 27, 2005 14:20 WSPC/179-JIN 00071

Integrative Assessment of Brain and Cognitive Function in PTSD 159

[31] van Laar M et al., Differential effects of amitriptyline, nefazodone and paroxetine
on performance and brain indices of visual selective attention and working memory,
Psychopharmacology (Berl) 162(4):351–63, 2002.
[32] Vasterling J et al., Attention and memory dysfunction in posttraumatic stress disorder,
Neuropsychology 12(1):125–33, 1998.
[33] Villarreal G et al., Reduced hippocampal volume and total white matter volume in
posttraumatic stress disorder, Biol Psychiat 52(2):119–25, 2002.
[34] Yehuda R et al., Learning and memory in combat veterans with posttraumatic stress
disorder, Am J Psychiat 152(1):137–139, 1995.