Psychiatry Research: Neuroimaging 214 (2013) 48–55

Contents lists available at ScienceDirect

Psychiatry Research: Neuroimaging

journal homepage: www.elsevier.com/locate/psychresns

Neural responses during emotional processing before and after

cognitive trauma therapy for battered women$
Robin L. Aupperle a,c,d,n, Carolyn B. Allard d, Alan N. Simmons a,b,d, Taru Flagan d,f,
Steven R. Thorp a,b,d, Sonya B. Norman a,b,d, Martin P. Paulus a,b,d, Murray B. Stein a,d,e
a
VA San Diego Healthcare System, San Diego, CA, USA
b
Center of Excellence for Stress and Mental Health, VA San Diego Healthcare System, San Diego, CA, USA
c
Department of Psychology, University of Missouri—Kansas City, Kansas City, MO, USA
d
Department of Psychiatry, University of California, San Diego, La Jolla, CA, USA
e
Family and Preventive Medicine, University of California, San Diego, La Jolla, CA, USA
f
Department of Psychology, University of Texas, Austin, TX, USA

art ic l e i nf o a b s t r a c t

Article history: Therapy for combat and accident-related posttraumatic stress disorder (PTSD) has been reported to
Received 19 October 2012 influence amygdala and anterior cingulate cortex (ACC) response during emotional processing. It is not
Received in revised form yet understood how therapy influences different phases of emotional processing, and whether previous
27 February 2013
findings generalize to other PTSD populations. We hypothesized that cognitive trauma therapy for
Accepted 16 May 2013
battered women (CTT-BW) would alter insula, amygdala, and cingulate responses during anticipation
and presentation of emotional images. Fourteen female patients with PTSD related to domestic violence
Keywords: completed the Clinician Administered PTSD Scale (CAPS) and functional magnetic resonance imaging
Prefrontal (fMRI) before and after CTT-BW. The fMRI task involved cued anticipation followed by presentation of
Insula
positive versus negative affective images. CTT-BW was associated with decreases in CAPS score,
Amygdala
enhanced ACC and decreased anterior insula activation during anticipation, and decreased dorsolateral
Anticipation
Functional magnetic resonance imaging prefrontal cortex and amygdala response during image presentation (negative–positive). Pre-treatment
Posttraumatic stress disorder ACC activation during anticipation and image presentation exhibited positive and negative relationships
to treatment response, respectively. Results suggest that CTT-BW enhanced efficiency of neural responses
during preparation for upcoming emotional events in a way that reduced the need to recruit prefrontal-
amygdala responses during the occurrence of the event. Results also suggest that enhancing ACC function
during anticipation may be beneficial for PTSD treatment.
& 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Treatment for posttraumatic stress disorder (PTSD) usually
includes psychotherapy, medication, or a combination of both
(Foa, 2006; Stein et al., 2006). There is strong empirical support
for cognitive-behavioral therapies (CBT) for PTSD, including pro-
longed exposure therapy (Foa, 2006) and cognitive processing
therapy (Resick and Schnicke, 1992), among others. Cognitive
trauma therapy for battered women (CTT-BW) is a CBT treatment
developed by Kubany et al. (2004) to specifically target issues
faced by women who developed PTSD after intimate partner
☆
This work was completed at the VA San Diego Healthcare System (VASDHS) and
the University of California, San Diego (UCSD; Department of Psychiatry), La Jolla,
CA.
n
Corresponding author at: University of Missouri—Kansas City, 5030 Cherry
Street, Cherry Hall Room 301, Kansas City, MO 64114, USA. Tel.: +1 816 235 5865;
fax: +1 816 235 1062.
E-mail address: aupperler@umkc.edu (R.L. Aupperle).
violence (IPV). While response rates for CTT-BW (Kubany et al.,
2004) and other CBTs are relatively high (50–80% of completers),
there is still room for improvement, with 20–60% of patients either
dropping out or not responding optimally (Schottenbauer et al.,
2008). Neuroimaging techniques may be useful in determining
mechanisms of healing and in offering insights into potential novel
treatment targets.
Neuroimaging studies with PTSD patients have focused primar-
ily on processing of trauma-related and emotional stimuli. In
general, results suggest PTSD subjects have greater activation in
amygdala and anterior insula regions and less activation of anterior
cingulate cortex (ACC) and/or medial and lateral prefrontal cortex
(PFC) (Etkin and Wager, 2007; Liberzon and Sripada, 2008; Shin and
Liberzon, 2010) compared to non-PTSD subjects, though there have
been some inconsistencies regarding directionality of findings
(Etkin and Wager, 2007). While the amygdala is implicated in
processing affective salience (Davis and Whalen, 2001; Morrison
and Salzman, 2010), the insula is implicated in processing inter-
oceptive signals (i.e., bodily responses) (Critchley et al., 2004; Craig,

0925-4927/$ - see front matter & 2013 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.pscychresns.2013.05.001
 R.L. Aupperle et al. / Psychiatry Research: Neuroimaging 214 (2013) 48–55
2011) and anticipating future events (Simmons et al., 2004; Knutson
and Greer, 2008). The overall pattern of dysfunction in PTSD is
thought to relate to difficulties appropriately engaging prefrontal
regions to monitor and regulate amygdala and insula activation and
the cognitive and behavioral responses to emotional stimuli
(Liberzon and Sripada, 2008; Shin and Liberzon, 2010).
A few studies have used fMRI to investigate changes in neural
response with PTSD treatment. Felmingham et al. (2007) and Roy
et al. (2010) observed that exposure-based CBT for PTSD was
associated with decreased amygdala and increased rostral and/or
dorsal ACC response during emotional face processing and the
affective Stroop task, respectively. These findings suggest therapy
may (a) enhance function within cingulate regions important for
conflict monitoring, inhibition, and cognitive-emotion regulation
(Ochsner and Gross, 2005; Botvinick, 2007; Levine, 2009) and
(b) reduce responsivity within the amygdala, perhaps reflecting a
reduction in the salience attributed to presented stimuli (Davis
and Whalen, 2001; Morrison and Salzman, 2010). A more recent
fMRI study by Thomaes et al. (2012) reported decreased dorsal
ACC and anterior insula activation during the affective Stroop task
after group CBT for complex PTSD. This suggests that CBT treat-
ment for PTSD may decrease insula as well as amygdala activation
during certain tasks and that the directional relationship between
treatment and ACC responses may be complex and influenced by
either subject or paradigm factors that differ between studies.
Neuroimaging methodologies may also enhance our under-
standing of why certain individuals experience more or less
benefit from treatment. Bryant et al. (2008) reported that greater
pre-treatment activation within both ventral ACC and amygdala to
backward-masked negative emotional faces was predictive of
worse response to exposure-based CBT for PTSD. While the
negative relationship between amygdala activation and treatment
response was in the expected direction, the authors had theorized
a positive relationship between ACC activation and treatment
response. The authors suggested the directionality of the finding
may have been due to the subconscious presentation of stimuli
and hypothesized that greater baseline ACC response during
conscious emotional processing may relate to better treatment
response.
The current fMRI study investigated neural outcomes and
predictors of CTT-BW using a task involving conscious anticipation
(ANI, API) followed by presentation of negative (NI) and positive
(PI) affective images. Enhanced emotional or physiological
responses to cues or “triggers” of emotional events are an impor-
tant aspect of PTSD diagnostic criteria. Anticipatory anxiety can
also lead to avoidance behavior, which is another important aspect
of PTSD that CBT often targets specifically (Foa, 2006). The task
used in the current study was designed to enable investigation of
neural processing during cued anticipatory processing. In previous
work, women with IPV-related PTSD have exhibited enhanced
activation within bilateral anterior insula and reduced activation
activation within lateral regions of the prefrontal cortex (PFC)
during emotional anticipation compared to women without PTSD
(Simmons et al., 2008; Aupperle et al., 2012). However, the
employed anticipation task also allows for investigation of neural
responses during the experience of the emotional stimulus itself
(Simmons et al., 2006). This paradigm therefore has potential for
elucidating the relationship between treatment and both antici-
patory and stimulus presentation phases of emotional processing.
In the current study, we sought to determine (a) effects of CTT-
BW on prefrontal-amygdala-insula responses during anticipation
and presentation of affective stimuli and (b) what pattern of
baseline prefrontal-amygdala-insula responses are predictive of
CTT-BW treatment response. We hypothesized that CTT-BW would
reduce activation in anterior insula during anticipatory processing
and both amygdala and anterior insula during stimulus processing,
49
but enhance ACC response during both anticipatory and stimulus
processing. Further, we hypothesized that greater ACC activation
and lower amygdala and/or anterior insula activation pre-
treatment would relate to better treatment response.
2. Methods
2.1. Participants and measures
Women who had experienced IPV and were seeking treatment for PTSD
symptoms were recruited for the study. Exclusionary criteria included substance
abuse in the past year; history of 45 years of alcohol abuse; use of psychotropic
medications within 4 weeks prior to the study, bipolar disorder or schizophrenia,
irremovable ferromagnetic bodily material, pregnancy, or claustrophobia. Forty-one
women with IPV-PTSD completed the anticipation task during fMRI and were
offered CTT-BW (Kubany et al., 2004). Several of these participants were excluded
due to not initiating or completing treatment or excessive movement during and
lower amygdala description of exclusions). A total of 14 women with IPV-PTSD
(mean [S.D.] age, 40.07 [7.44] years; mean [S.D.] education¼ 14.43 [1.99] years)
were included in final analyses. These subjects represent a subset of a cohort for
which pre-treatment fMRI results were reported previously (Aupperle et al., 2012).
All participants were seeking treatment for PTSD symptoms and met full (N¼11)
or partial (N¼ 3) DSM-IV PTSD criteria, verified using the Clinician-Administered
PTSD Scale (CAPS) (Blake et al., 1995). Treatment involved 90-minute weekly sessions
of CTT-BW (mean number of sessions¼ 11.57, S.D.¼1.60). CTT-BW was developed by
Kubany et al. (2004; Kubany and Ralston, 2008) and is a manualized, modular
intervention designed to treat PTSD and functional impairment specifically for
IPV victims. CTT-BW is based on cognitive behavioral principles and includes
psychoeducation, skills training, exposure to reminders of trauma, and assessment
and correction of irrational beliefs. CTT-BW focuses these intervention strategies on
areas of distress most relevant to IPV survivors. See Supplementary Material for
further of CTT-BW.
Subjects completed, among other measures, the CAPS (Blake et al., 1995), PTSD
Checklist (PCL) (Blanchard et al., 1996), and Beck Depression Inventory version 2
(BDI-II) (Beck et al., 1996) pre- and post-treatment. The CAPS was additionally
completed after treatment at 3-month follow-up. See Supplementary Material for
IPV and full/partial PTSD definitions and exclusionary criteria.
After complete description of the study to the subjects, written informed
consent was obtained at the initial study session. The study protocol was approved
by the University of California—San Diego Human Research Protections Program
and the Veterans Affairs San Diego Healthcare System Research and Development
Office; all procedures were completed at these institutions.
2.2. fMRI data acquisition
The anticipation task was conducted pre- and post-treatment during fMRI
scans sensitive to blood oxygenation level-dependent (BOLD) contrast using a Signa
Excite (GE Healthcare, USA) 3.0 T scanner (T2n-weighted echoplanar [EPI] imaging,
repetition time (TR) ¼ 2000 ms, echo time (TE) ¼ 32 ms, 64  64 matrix, 30 2.6-mm
axial slices with a 1.4-mm gap, 290 scans). During each scan session, a high-
resolution T1-weighted image (spoiled gradient recalled (SPGR), TR ¼ 8 ms,
TE¼ 3 ms, 172 sagittal slices with approximately 1 mm3 voxels) was obtained for
anatomical reference.
The anticipation task, conducted as previously described (Simmons et al., 2008)
and as shown in Fig. 1, combines a continuous performance task (CPT) with
interspersed presentation of positive (PI) and negative (NI) affective images. During
the CPT, subjects were asked to press a ‘LEFT’ or ‘RIGHT’ button corresponding to
the direction of an arrow on the screen. Simultaneously, a 250-ms long 500-Hz
tone was presented every 2 s. During baseline conditions, the background screen
was gray. Subjects were instructed prior to the task that when the background
screen turned blue, accompanied by a 250-Hz tone, a positive image would appear
on the screen, whereas when the background turned yellow, accompanied by a
1000-Hz tone, a negative image would appear. The trials in which the background
was either blue or yellow represented the anticipation periods. The picture stimuli
comprised 17 positive and 17 negative images taken from the International
Affective Picture System (IAPS Lang et al., 2008). The positively and negatively
valenced images were matched on level of arousal (as reported in the IAPS manual)
to allow examination of valence effects specifically. The anticipation periods during
the task lasted 6 s, and the image presentation lasted 2 s. The baseline CPT task was
interspersed for variable duration averaging about 8 s in between trials. There was
no inter-stimulus interval between anticipation and image presentation phase. The
total duration of the task was 580 s. Response accuracy and reaction time were
obtained for the CPT during baseline, anticipation of a positive image (API), and
anticipation of a negative image (ANI) conditions. See Supplementary Material for
further description of this task.
50 R.L. Aupperle et al. / Psychiatry Research: Neuroimaging 214 (2013) 48–55
Fig. 1. fMRI anticipation task. This task combines a continuous performance task
with the interspersed presentation of affective stimuli. Subjects are asked to press a
left or right button based on the direction of the arrow. Subjects are instructed
prior to the task that a blue square accompanied by a low tone indicates a positive
image is going to appear. In contrast, a switch to a yellow square accompanied by a
high tone signals an impending negative image. Images used in this paradigm were
taken from the International Affective Picture System. (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of
this article.)
2.3. Data analysis
2.3.1. Clinical and behavioral data
Clinical measures (CAPS, PCL, BDI-II) were compared between pre-treatment,
post-treatment, and follow-up using repeated-measures analysis of variance
(ANOVA) in the Statistical Package for the Social Sciences (SPSS). Repeated
measures ANOVA was used to examine main and interaction effects of condition
(ANI vs. API) and treatment (pre- vs. post-treatment) on CPT reaction time and
accuracy (results in Supplementary Material). Clinical and behavioral results were
considered significant at p o 0.05.
2.3.2. fMRI BOLD data
Data were preprocessed and analyzed using Analysis of Functional NeuroI-
mages (AFNI) (Cox, 1996) and the R statistical package (r-project.org). EPI images
were aligned to high-resolution anatomical images. Voxel data points representing
outliers relative to surrounding data points were eliminated and interpolated.
Voxel time series were interpolated to correct for non-simultaneous slice acquisi-
tion within each volume and corrected for three-dimensional motion. Individual
time-series data were analyzed using a multiple regression model and BOLD
hemodynamic response function (  4–6 s peak). Four orthogonal regressors of
interest were used to quantify neural activation for PI, NI, API and ANI periods.
Six regressors of no interest were also entered into the model: (1) baseline
regressor, (2–4) motion-related regressors (roll, pitch, and yaw), (5) white-matter
mask to control for physiological noise, and (6) linear trend used to eliminate slow
signal drifts.
Percent signal change (PSC) was calculated by dividing the regressor of interest
by the baseline regressor. Data were spatially blurred with a 4-mm full width at
half-maximum (FWHM) spatial filter and normalized to Talairach space. Paired-
samples t-tests were used to examine task effects (ANI vs. API; NI vs. PI) on PSC at
pre- and post-treatment separately (Supplementary Tables 2–3).
Linear mixed effect (LME) models were implemented using the R statistical
package (using Restricted Maximum Likelihood Estimation procedures) to examine
condition (ANI vs. API; NI vs. PI) by treatment (pre- vs. post-treatment) effects on
PSC (subject as random effects). LME analyses were also conducted to investigate
the relationship between change in symptom severity and change in brain
activation by examining time (pre- vs. post-treatment) by treatment response
(change in CAPS score) interaction effects on BOLD response during anticipation
(ANI-API; NI-PI). Huber robust regression analyses (Huber, 1973) with bootstrap-
ping (50 maximum iterations) were conducted to examine the relationship
between pre-treatment BOLD response (ANI-API; NI-PI) and post-treatment CAPS
score, while covarying for pre-treatment CAPS score.
For LME results, average F and p values are reported for each cluster, while
unstandardized coefficients (B) and t values are reported for regression results. PSC
was extracted from each significant cluster of activation and statistics provided
concerning post-hoc analyses conducted with each condition separately (ANI and
API; NI and PI). These statistics aid in understanding the relative contribution of
each to analyses originally conducted with ANI-API or NI-PI differential PSC.
Analyses were conducted voxel-wise for whole-brain and regions of interest
(ROIs) related to emotional processing in PTSD, including bilateral insula, amygdala
and cingulate. See Supplementary Material and Supplementary Fig. 1 for a full
description of ROI masks. fMRI results were considered significant at p o0.05,
Monte Carlo corrected for multiple comparisons, resulting in a minimum cluster
extent (and adjusted p threshold) of 832 mm3 (p o 0.00001) for whole-brain,
448 mm3 (p o 0.0002) for cingulate, 192 mm3 (p o0.0008) for amygdala, and
320 mm3 (po 0.0007) for insula regions.
PSC was extracted for clusters significant from the condition by treatment LME
analyses. Spearman's rho correlations (p o 0.05) between anticipation (ANI-API)
and image (NI-PI) phase activations are reported in Supplementary Material. This
exploratory analysis aimed to understand whether activations during anticipation
may influence activation during subsequent image presentation.
3. Results
3.1. Treatment effects
3.1.1. Clinical measures
CAPS total severity (F(2,24) ¼67.06, p o0.001), cluster B (F(2,
24)¼46.65, p o0.001), cluster C (F(2,24)¼56.23, p o0.001), and
cluster D (F(2,24)¼17.34, po 0.001) subscale scores significantly
changed between pre-treatment, post-treatment, and follow-up
periods. Post-hoc t-tests indicated that post-treatment and follow-
up CAPS scores were substantially lower than pre-treatment, but
there was no significant difference between post-treatment and
follow-up. PCL scores (missing for one subject at pre-treatment
and another at follow-up) also significantly changed between pre-
treatment, post-treatment, and follow-up (F(2,22) ¼38.87,
po 0.001). PCL scores showed the same pattern as the CAPS, with
post-treatment and follow-up scores being lower than pre-treat-
ment, but with no significant difference between post-treatment
and follow-up. BDI scores were available for pre- (M¼20.50; S.
D.¼9.28) and post-treatment (M ¼7.07, S.D.¼ 7.26) and also
showed a significant decline (t(13) ¼ 4.71, p o0.001). Thus, results
support that completion of CTT-BW was associated with decreases
in PTSD and depressive symptoms. See Supplementary Table 1 and
Fig. 2.
There was a greater than 50% symptom reduction (CAPS
severity score) in 12 of the 14 subjects. However, there was
variability in level of treatment response with percent decrease
in CAPS score ranging from 23.53% to 100% (M ¼26.43, S.
D.¼22.09), thus allowing for investigations concerning neuroima-
ging predictors of level of treatment response.
3.1.2. Image anticipation phase fMRI
Whole-brain analyses revealed differential activation during
anticipation (ANI-API) increased from pre- to post-treatment within
clusters at the rostral/dorsal intersect of the left ACC (Brodmann
area (BA) 9 and 32) and in the left posterior cingulate (PCC; BA 29)
and decreased from pre- to post-treatment within right anterior
insula (BA 13) (other regions listed in Table 1). No additional regions
were identified through ROI analysis. See Fig. 2.
Better treatment response (greater decrease in CAPS score)
significantly related to decreased BOLD response from pre- to
post-treatment within left anterior insula (BA 13; 1216 mm3; x,y,
z¼−37,−2,2; F(1,12) ¼6.94), posterior cingulate (BA 29; 1472 mm3;
x,y,z ¼−4,−34,9; F(1,12) ¼8.89; ROI analysis: BA 31; 448 mm3; x,y,
z¼−1,−57,26; F(1,12) ¼7.38), and precuneus (BA 7; 896 mm3; x,y,
z¼8,−60,32; F(1,12) ¼7.38). ROI analyses additionally identified a
right posterior insula cluster (BA 13; 384 mm3; x,y,z ¼45,−33,20; F
(1,12) ¼6.39) in which better treatment response related to greater
increases in BOLD activation.
3.1.3. Image presentation phase fMRI
Whole-brain analyses revealed enhanced differential activation
during presentation of affective images (NI-PI) within bilateral
 R.L. Aupperle et al. / Psychiatry Research: Neuroimaging 214 (2013) 48–55 51

Table 1
Treatment by condition interaction effect on BOLD response during anticipation of affective images.

Side Region BA Cluster size (mm3) Center of mass coordinates (x,y,z) F value t value for ANI only t value for API only

Increased activation (ANI-API) from pre- to post-treatment

Left Posterior cingulate cortex 29 2176 −5 −38 17 7.10 3.13 −2.86
Right Cerebellum 2048 26 −58 −30 6.70 2.78 −2.88
Right Superior/middle temporal gyrus 21,22 1664 61 −16 −1 6.91 3.33 −2.58
Left ACC, medial frontal gyrus 32,9 1472 −16 37 24 7.12 3.06 −2.97
Left Cerebellum 1088 −8 −61 −29 6.45 2.59 −3.57
Decreased activation (ANI-API) from pre- to post-treatment
Right Middle frontal gyrus 10 960 29 55 14 5.46 −1.67 2.13
Right Mid Insula 13 832 34 4 −2 6.58 −1.65 3.16

Notes: Results shown from linear mixed model analysis for treatment (pre- vs. post-treatment) by condition (ANI vs. API) interaction effect on percent signal change. Analyses
included N ¼ 14 women with domestic violence related PTSD who completed cognitive trauma therapy for battered women (CTT-BW). Clusters listed met significance cutoff
of p o0.05, Monte Carlo adjusted for whole-brain volume (832 mm3) multiple comparisons. All coordinates are Talairach coordinates (x,y,z) based on Talairach Daemon
software (Lancaster et al., 2000). Abbreviations: BA¼Brodmann area; ACC ¼ anterior cingulate cortex; ANI¼ anticipation of negative images; API ¼ anticipation of positive
images; mm3 ¼millimeters cubed.

posterior cingulate (BA 31) and right inferior parietal cortex (BA
40) and decreased activation in dorsolateral prrefrontal cortex
(dlPFC; BA 9) from pre- to post-treatment. ROI analyses addition-
ally revealed a cluster within the right amygdala in which
differential activation decreased from pre- to post-treatment. See
Table 2 and Fig. 2.
Better treatment response related to greater increases in BOLD
response from pre- to post-treatment within bilateral precuneus
(BA 18,19; left: 9344 mm3; x,y,z¼ −7,−76,12; F(1,12) ¼8.19; right:
1664 mm3; x,y,z ¼7,−75,35; F(1,12) ¼7.02) and right posterior cin-
gulate (BA 31; 832 mm3; x,y,z¼ 9,−29,37; F(1,12) ¼ 11.70), medial
frontal gyrus (BA 6; 1920 mm3; F(1,12) ¼8.04), precentral gyrus
(BA 4; 1216 mm3; x,y,z¼40,−14,46; F(1,12) ¼9.09), lingual gyrus
(BA 18; 1216 mm3; x,y,z ¼9,−77,3; F(1,12) ¼9.06), cerebellum
(2048 mm3; x,y,z ¼19,−53,−5; F(1,12) ¼8.68), and inferior temporal
gyrus (BA 20; 448 mm3; x,y,z¼ 56,−11,−24; F(1,12) ¼ 9.76). Better
treatment response related to decreases in BOLD response within
right superior/transverse temporal gyrus (BA 41; 1344 mm3; x,y,
z ¼39,−33,12; F(1,12) ¼ 7.14), cerebellum (3840 mm3; x,y,z¼31,−58,
−32; F(1,12) ¼7.26) and caudate body (896 mm3; x,y,z ¼0,13,10;
F(1,12) ¼7.38). ROI analyses identified a left posterior insula cluster
(BA 13; 448 mm3; x,y,z¼ 39,−5,16; F(1,12) ¼8.46) in which better
treatment response significantly related to greater increases in
BOLD activation.
A primary pattern revealed by results from anticipation and
image phase presentation was that CTT-BW was associated with
enhanced prefrontal (ACC) and decreased insula response during
anticipation and decreased prefrontal (dlPFC) and amygdala
response during image presentation.
3.2. Baseline fMRI predictors of treatment response
3.2.1. Image anticipation phase fMRI
Huber robust regression analysis (with bootstrapping) was
conducted using baseline differential activation during anticipa-
tion (ANI-API) as the predictor, post-treatment CAPS score as the
dependent variable, and pre-treatment CAPS as covariate. Whole-
brain analyses revealed that pre-treatment right dorsal/rostral ACC
(BA 32) activation related negatively to post-treatment CAPS score
(Fig. 3). ROI analysis additionally revealed a right posterior
cingulate cortex (BA 31) cluster that related negatively to post-
treatment CAPS score. See Table 3 for full list of results.
3.2.2. Image presentation phase fMRI
Differential activation during presentation of affective images
(NI-PI) within right middle temporal gyrus (BA 39) and precuneus
(BA 7) related negatively to post-treatment CAPS score. ROI
analysis revealed that right dorsal/rostral ACC (BA 32; shown in
Fig. 3) and left posterior cingulate (BA 31) related positively to
post-treatment CAPS score, while right posterior insula (BA 13)
related negatively to post-treatment CAPS score (Table 3).
Analysis of baseline predictors of treatment response therefore
supported a relationship between ACC activation and treatment
response, though the directionality was positive for the anticipa-
tion phase but negative for the image presentation phase.
4. Discussion
This study examined the relationship between a cognitive-
behavioral therapy for IPV-PTSD (CTT-BW) and neural activations
during anticipation and image presentation phases of emotional
processing. Exposure-based CBT for PTSD has been associated with
increased ACC and other PFC activations in previous studies
(Felmingham et al., 2007; Roy et al., 2010; Thomaes et al., 2012).
Individuals with IPV-PTSD showed an increase in ACC activation
during anticipation (ANI-API) after CTT-BW. As we have not pre-
viously identified reduced ACC activation for IPV-related PTSD
subjects compared to controls (Simmons et al., 2008; Aupperle
et al., 2012) on the employed anticipation task, this finding may not
reflect a simple normalization in activations from pre- to post-
treatment. It is possible that CTT-BW may enhance recruitment of
ACC regions in a way that helps to compensate and normalize
activation in other regions (i.e., insula, amygdala). The ACC clusters
identified in the current study lie in dorsal aspects of pregenual
ACC, at the intersect of dorsal and rostral ACC subregions. The ACC
in general has been associated with conflict monitoring, inhibition,
and emotion regulation (Ochsner and Gross, 2005; Botvinick, 2007;
Etkin et al., 2011). Dorsal ACC has been proposed to play a primary
role in cognitive processing or appraisal of emotion while ventral/
rostral aspects play a more primary role in emotional processing or
automatic emotional regulation (Mohanty et al., 2007; Etkin et al.,
2011). Given the focus of CBT on learning to regulate emotions and
restructure cognitions, we propose that enhanced ACC activation in
this study likely reflects increased use of conscious reappraisal or
cognitive-emotion regulation strategies during anticipation.
Our results suggest that CTT-BW may influence neural activa-
tions differently depending on the phase of emotional processing.
Treatment was associated with enhanced ACC activation during
anticipation, but with decreased dlPFC response during image
presentation. We suggest that CBTs may enhance ACC involvement
specifically during individuals' initial opportunity to regulate
52 R.L. Aupperle et al. / Psychiatry Research: Neuroimaging 214 (2013) 48–55

Fig. 2. Treatment effect on fMRI BOLD activation. Cognitive trauma therapy for battered women (CTT-BW; Kubany and Ralston, 2008) was associated with increased
differential activation during anticipation of negative [API] versus positive [API] images within left anterior cingulate cortex (A; ACC; BA 32/9; shown at x ¼ −12) and reduced
differential activation within right anterior insula (B; BA 13; shown at X ¼ 40). CTT-BW was associated with decreased differential activation during presentation of negative
[NI] versus positive [PI] images within left dorsolateral prefrontal cortex (C; dlPFC; BA 9; shown at x ¼−44) and right amygdala (D; shown at x¼ 24).

emotional responses to an event. Perhaps, more efficient prepara- by the identified negative correlation between ACC anticipatory
tion during anticipation reduces the need to recruit prefrontal and activation and amygdala activation during image presentation
amygdala regions during the event. This idea is partially supported pre-treatment (Supplementary Material).
 R.L. Aupperle et al. / Psychiatry Research: Neuroimaging 214 (2013) 48–55 53

Table 2
Treatment by condition interaction effect on BOLD response during presentation of affective images.

Side Region BA Cluster size (mm3) Center of mass coordinates (x,y,z) F value t value for NI only t value for PI only

Increased activation (NI-PI) from pre- to post-treatment

Right Precuneus, inferior parietal 40 1344 26 −46 53 5.57 2.60 −2.58
Left Precuneus, posterior cingulate 31 896 −18 −53 26 5.30 2.80 −2.71
Right Precuneus 31 896 18 −54 29 6.81 4.92 −2.46
Decreased activation (NI-PI) from pre- to post-treatment
Left Dorsolateral PFC 9 1280 −44 6 27 5.75 −2.84 3.11
Right Amygdala* 448 25 −5 −15 5.64 −2.80 2.41

Notes: Results shown from linear mixed model analysis for treatment (pre- vs. post-treatment) by condition (NI vs. PI) interaction effect on percent signal change. Analyses
included N ¼ 14 women with domestic violence related PTSD who completed cognitive trauma therapy for battered women (CTT-BW). Clusters listed met significance cutoff
of po 0.05, Monte Carlo adjusted for whole-brain volume (832 mm3) multiple comparisons or region of interest analysis (n). All coordinates are Talairach coordinates (x,y,z)
based on Talairach Daemon software (Lancaster et al., 2000). Abbreviations: BA¼ Brodmann area; PFC ¼ prefrontal cortex; NI ¼negative image; PI ¼positive image;
mm3 ¼ millimeters cubed.

Fig. 3. Brain regions for which baseline activation was predictive of post-treatment CAPS score. Greater differential activation during anticipation (anticipation of negative
images [ANI]—anticipation of positive images [API]) in the anterior cingulate cortex (ACC; A; BA 32; shown a x ¼ 9) at pre-treatment related to lower post-treatment
symptom severity. However, greater differential activation during presentation of images (negative images [NI]—positive images [PI]) in the ACC (B; BA 32; shown at x¼ 9) at
pre-treatment related to greater post-treatment symptom severity.

Greater pre-treatment ACC activation during anticipation
(ANI-API) and less ACC activation during image presentation (NI-
PI) related to lower CAPS score post-treatment (i.e., better treat-
ment response). The latter finding is consistent with the results of
Bryant et al. (Bryant et al., 2008) for ventral ACC during subcon-
scious emotional face processing. Greater propensity to engage
ACC regions when preparing for emotional events may reflect
greater potential for regulating emotional responses, therefore
helping a patient respond optimally to CBT. In contrast, less
efficiency during preparation phases of emotional processing
may indicate a greater need to recruit ACC regions during the
image phase—thus relating to less treatment response.
Our findings suggest that enhancing ACC function prior to
therapy may have potential for boosting treatment response. It is
uncertain whether techniques aimed at enhancing ACC function in
general may be helpful in this regard, or if only condition- or
trauma-specific techniques (such as CBT) can be effective. How-
ever, preliminary findings with transcranial magnetic stimulation
(TMS) (Boggio et al., 2010) suggest that targeting PFC regions in
general can have beneficial effects for PTSD. Other potential
strategies may include cognitive training (Haut et al., 2010) or
neurofeedback (Hamilton et al., 2011).
Differential posterior cingulate cortex (PCC) activation during
anticipation (ANI-API) increased while PCC activation during
54 R.L. Aupperle et al. / Psychiatry Research: Neuroimaging 214 (2013) 48–55

Table 3
Regions for which pre-treatment BOLD response related to post-treatment symptom severity (CAPS score).

Side Region BA Cluster size (mm3) Center of mass coordinates (x,y,z) B coefficient t value t value for ANI only t value for API only

Anticipation: Less activation (ANI-API) pre-treatment ¼ better treatment response

N/A
Anticipation: Greater activation (ANI-API) pre-treatment¼ better treatment response
Right Dorsal ACC 32 1408 9 32 28 −0.70 −3.22 −4.14 3.02
Right Posterior cingulate cortex* 31 512 3 −43 27 −0.56 −3.09 −2.36 2.22
Image processing: Less activation (NI-PI) pre-treatment ¼ better treatment response
Right Dorsal ACC* 32 576 10 26 24 0.74 3.11 3.89 −3.42
Left Posterior cingulate* 31 512 −5 −34 39 −0.70 3.25 1.22 −3.18
Image processing: Greater activation (NI-PI) pre-treatment ¼ better treatment response
Right Middle temporal gyrus 39 1216 52 −23 28 −0.68 −3.20 −6.21 1.80
Right Precuneus 7 1024 28 −65 33 −0.69 −2.81 −9.40 −0.70

Notes: Results shown Huber robust regression analyses investigating relationship between pre-treatment differential percent signal change between either anticipation of
negative (ANI)—positive (API) images or processing of negative (NI)—positive images (PI) and post-treatment symptom severity (CAPS) while covarying for pre-treatment
CAPS. Analyses included N ¼ 14 women with domestic violence related PTSD who completed cognitive trauma therapy for battered women (CTT-BW). Clusters listed
met significance cutoff of p o 0.05, Monte Carlo adjusted for whole-brain volume (832 mm3) multiple comparisons or region of interest analyses (n). All coordinates
are Talairach coordinates (x,y,z) based on Talairach Daemon software (Lancaster et al., 2000). Abbreviations: BA¼Brodmann area; ACC ¼anterior cingulate cortex;
mm3 ¼millimeters cubed.

image presentation decreased after CTT-BW. The relationship of
these changes to treatment response was inconsistent, however—
with greater increases during anticipation and greater decreases
during image presentation relating to worse treatment response.
Farrow et al. (2005) reported increased PCC activation during
empathy and forgiveability judgments after CBT for PTSD. Our
findings suggest that while CBT influences PCC activation during
emotional processing, such changes may not be contributing to
successful treatment response.
CTT-BW was also associated with reduced anterior insula
response during anticipation (ANI-API) and reduced amygdala
response during image presentation (NI-PI). Additionally, the
amount of decreased anticipatory activation in left anterior insula
related to better treatment response. The anterior insula has been
implicated in monitoring internal bodily state and predicting state
changes (Critchley et al., 2004; Paulus and Stein, 2006; Craig,
2011), thus having an important role in anticipation (Simmons
et al., 2004; Simmons et al., 2008; Aupperle et al., 2012). The insula
has also been consistently implicated in the pathophysiology of
PTSD (Etkin and Wager, 2007; Shin and Liberzon, 2010) and
previous studies with the current anticipation task have identified
enhanced insula activation for IPV-related PTSD compared to
healthy controls (Simmons et al., 2008; Aupperle et al., 2012).
Thus, decreased recruitment of insula regions during anticipation
may represent a normalization of response from pre- to
post-treatment. Perhaps, with successful CBT, an individual's
homeostasis is less disrupted by anticipation of emotional changes
in the environment. This could result from enhanced top-down
modulation, i.e. cognitive regulation, or attenuated bottom-up
modulation, i.e. decreases in neural signals representing change
in emotional homeostasis. Future research could be aimed at
clarifying how treatment influences top-down versus bottom-up
processes related to insula response.
The amygdala has been implicated in processing emotional
salience, particularly negative valence and fear learning (Davis and
Whalen, 2001; Morrison and Salzman, 2010), and in the patho-
physiology of PTSD and anxiety disorders (Etkin and Wager, 2007;
Liberzon and Sripada, 2008; Shin and Liberzon, 2010). Findings
of decreased amygdala activation during image presentation (NI-
PI) after CTT-BW are consistent with the findings of Felmingham
et al. (2007) and Roy et al. (2010) exposure-based CBT and most
likely represent a normalization of activation from pre- to post-
treatment. It is also consistent with findings reported for anxioly-
tic medications (Paulus et al., 2005; Aupperle et al., 2011). Thus,
decreased amygdala activation may be a common mechanism of
action across various anxiety-related treatments. Whereas anxio-
lytics may reduce prefrontal-insula-amygdala responsivity overall,
CBT treatments may influence amygdala and insula responses via
enhanced top-down inhibitory ACC networks.
Due to this study not being a randomized, controlled trial, pre-
to post-treatment changes could have been influenced by natural
recovery processes or repeat-testing effects rather than CTT-BW.
The relationship between amount of change in some implicated
regions (e.g., anterior insula) and change in symptoms provides
some support for the changes being related to treatment. How-
ever, it cannot be determined from the current study whether
effects were directly due to CTT-BW. Future research using
randomized, placebo-controlled designs in conjunction with neu-
roimaging is needed to further delineate mechanisms of treat-
ment. Although this study represents one of the largest fMRI/
treatment outcome studies with PTSD, statistical power was
still limited due to sample size. With the majority of patients
responding well to CTT-BW, variability in treatment outcome was
constricted and may have further limited statistical power.
As previous fMRI/treatment outcome studies have involved PTSD
related to combat or accidents, one of the current study's strengths
is that it included female patients with IPV-PTSD. However,
generalizability to other PTSD populations cannot be assumed.
The task used in the current study was designed to specifically
examine valence effects on emotional anticipation. We did not
include a jittered inter-stimulus interval between anticipation and
image presentation phases, thus reducing the independence of
findings from the image versus the anticipation phases. Notably,
there were findings unique to the image vs. anticipation phases QJ;
(i.e., amygdala), supporting the unique contribution of each to our
understanding of CTT-BW treatment effects. In addition, as
the negatively and positively valenced images inclu-
ded in the task were matched on arousal, we were not able to
identify how CTT-BW may influence activations related to arousal.
Future research could include a neutral condition in addition to
negatively and positively valenced conditions in order to investi-
gate effects on both valence- and arousal-specific activation
patterns.
Despite these limitations, the current study suggests that
CTT-BW is associated with changes in prefrontal-insula-amygdala
responding, though directionality of change may depend upon the
phase of emotional processing. CTT-BW may increase activation in
regions involved in cognitive-emotion regulation (ACC) and
decrease activation in regions associated with emotional anticipa-
tion and interoception (anterior insula) during preparation for
 R.L. Aupperle et al. / Psychiatry Research: Neuroimaging 214 (2013) 48–55
upcoming emotional events. CTT-BW may also decrease the need
to recruit dlPFC and amygdala regions during subsequent presen-
tation of emotional stimuli. The more PTSD patients initially
engage ACC regions when anticipating negative events, the more
they may benefit from CTT-BW. Novel interventions that enhance
ACC function (e.g., prior to initiating therapy) could be beneficial
for PTSD treatment and warrant further investigation.
Acknowledgment
This study was funded by the Department of Veteran Affairs
(Merit Award to MBS). Dr. Thorp is funded by a VA Career
Development Award.
We would also like to acknowledge the contributions of
Michelle Behrooznia, M.A., Shadha Cissell, M.S.W., Erin Grimes,
Psy.D., Kelly Hughes-Berardi, Ph.D., Lindsay Reinhardt, B.A., and
Sarah Sullivan, B.A., towards coordinating the study, recruiting
subjects, and/or collecting behavioral and fMRI data. We also
thank Greg Fonzo, M.S., for creating the anatomical masks used
for functional magnetic resonance imaging (fMRI) region of inter-
est analyses.
Appendix A. Supplementary material
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.pscychresns.2013.
05.001.
References
Aupperle, R.L., Allard, C.B., Grimes, E.M., Simmons, A., Flagan, T., Behrooznia, M.,
Cissel, S.H., Twamley, E.W., Thorp, S.R., Norman, S.B., Paulus, M.P., Stein, M.B.,
2012. Dorsolateral prefrontal cortex activation during emotional anticipation
and neuropsychological performance in PTSD. Archives of General Psychiatry
69, 360–371.
Aupperle, R.L., Ravindran, L., Tankersley, D., Flagan, T., Stein, N.R., Simmons, A.N.,
Stein, M.B., Paulus, M.P., 2011. Pregabalin influences insula and amygdala
activation during anticipation of emotional images. Neuropsychopharmacology
36, 1466–1477.
Beck, A.T., Steer, R.A., Brown, G.K., 1996. Manual for Beck Depression Inventory II
(BDI-II). Psychological Corporation, San Antonio, TX.
Blake, D.D., Weathers, F.W., Nagy, L.M., Kaloupek, D.G., Gusman, F.D., Charney, D.S.,
Keane, T.M., 1995. The development of a Clinician-Administered PTSD Scale.
Journal of Traumatic Stress 8, 75–90.
Blanchard, E.B., Jones-Alexander, J., Buckley, T.C., Forneris, C.A., 1996. Psychometric
properties of the PTSD Checklist (PCL). Behaviour Research and Therapy 34,
669–673.
Boggio, P.S., Rocha, M., Oliveira, M.O., Fecteau, S., Cohen, R.B., Campanha, C.,
Ferreira-Santos, E., Meleiro, A., Corchs, F., Zaghi, S., Pascual-Leone, A., Fregni,
F., 2010. Noninvasive brain stimulation with high-frequency and low-intensity
repetitive transcranial magnetic stimulation treatment for posttraumatic stress
disorder. Journal of Clinical Psychiatry 71, 992–999.
Botvinick, M.M., 2007. Conflict monitoring and decision making: reconciling two
perspectives on anterior cingulate function. Cognitive Affective and Behavioral
Neuroscience 7, 356–366.
Bryant, R.A., Felmingham, K., Kemp, A., Das, P., Hughes, G., Peduto, A., Williams, L.,
2008. Amygdala and ventral anterior cingulate activation predicts treatment
response to cognitive behaviour therapy for post-traumatic stress disorder.
Psychological Medicine 38, 555–561.
Cox, R.W., 1996. AFNI: software for analysis and visualization of functional
magnetic resonance neuroimages. Computers and Biomedical Research 29,
162–173.
Craig, A.D., 2011. Significance of the insula for the evolution of human awareness of
feelings from the body. Annals of the New York Academy of Sciences 1225,
72–82.
Critchley, H.D., Wiens, S., Rotshtein, P., Ohman, A., Dolan, R.J., 2004. Neural systems
supporting interoceptive awareness. Nature Neuroscience 7, 189–195.
Davis, M., Whalen, P.J., 2001. The amygdala: vigilance and emotion. Molecular
Psychiatry 6, 13–34.
Etkin, A., Egner, T., Kalisch, R., 2011. Emotional processing in anterior cingulate and
medial prefrontal cortex. Trends in Cognitive Sciences 15, 85–93.
55
Etkin, A., Wager, T.D., 2007. Functional neuroimaging of anxiety: a meta-analysis of
emotional processing in PTSD, social anxiety disorder, and specific phobia.
American Journal of Psychiatry 164, 1476–1488.
Farrow, T.F., Hunter, M.D., Wilkinson, I.D., Gouneea, C., Fawbert, D., Smith, R., Lee,
K.H., Mason, S., Spence, S.A., Woodruff, P.W., 2005. Quantifiable change in
functional brain response to empathic and forgivability judgments with
resolution of posttraumatic stress disorder. Psychiatry Research: Neuroimaging
140, 45–53.
Felmingham, K., Kemp, A., Williams, L., Das, P., Hughes, G., Peduto, A., Bryant, R.,
2007. Changes in anterior cingulate and amygdala after cognitive behavior
therapy of posttraumatic stress disorder. Psychological Science 18, 127–129.
Foa, E.B., 2006. Psychosocial therapy for posttraumatic stress disorder. Journal of
Clinical Psychiatry 67 (Suppl 2), 40–45.
Hamilton, J.P., Glover, G.H., Hsu, J.J., Johnson, R.F., Gotlib, I.H., 2011. Modulation of
subgenual anterior cingulate cortex activity with real-time neurofeedback.
Human Brain Mapping 32, 22–31.
Haut, K.M., Lim, K.O., MacDonald 3rd, A., 2010. Prefrontal cortical changes following
cognitive training in patients with chronic schizophrenia: effects of practice,
generalization, and specificity. Neuropsychopharmacology 35, 1850–1859.
Huber, P.J., 1973. Robust regression: asymptomatics, conjectures, and Monte Carlo.
Annals of Statistics 1, 799–821.
Knutson, B., Greer, S.M., 2008. Anticipatory affect: neural correlates and conse-
quences for choice. Philosophical Transactions of the Royal Society of London.
Series B: Biological Sciences 363, 3771–3786.
Kubany, E.S., Hill, E.E., Owens, J.A., Iannce-Spencer, C., McCaig, M.A., Tremayne, K.J.,
Williams, P.L., 2004. Cognitive trauma therapy for battered women with PTSD
(CTT-BW). Journal of Consulting and Clinical Psychology 72, 3–18.
Kubany, E.S., Ralston, T.C., 2008. Treating PTSD in Battered Women: A Step-by-Step
Manual for Therapists and Counselors. New Harbinger, Oakland.
Lancaster, J.L., Woldorff, M.G., Parsons, L.M., Liotti, M., Freitas, C.S., Rainey, L.,
Kochunov, P.V., Nickerson, D., Mikiten, S.A., Fox, P.T., 2000. Automated Talairach
atlas labels for functional brain mapping. Human Brain Mapping 10, 120–131.
Lang, P.J., Bradley, M.M., Cuthbert, B.N., 2008. International Affective Picture System
(IAPS): Affective Ratings of Pictures and Instruction Manual, Technical Report
A-8. University of Florida, Gainesville, FL.
Levine, D.S., 2009. Brain pathways for cognitive-emotional decision making in the
human animal. Neural Networks 22, 286–293.
Liberzon, I., Sripada, C.S., 2008. The functional neuroanatomy of PTSD: a critical
review. Progress in Brain Research 167, 151–169.
Mohanty, A., Engels, A.S., Herrington, J.D., Heller, W., Ho, M.H., Banich, M.T., Webb,
A.G., Warren, S.L., Miller, G.A., 2007. Differential engagement of anterior
cingulate cortex subdivisions for cognitive and emotional function. Psychophy-
siology 44, 343–351.
Morrison, S.E., Salzman, C.D., 2010. Re-valuing the amygdala. Current Opinion in
Neurobiology 20, 221–230.
Ochsner, K.N., Gross, J.J., 2005. The cognitive control of emotion. Trends in Cognitive
Sciences 9, 242–249.
Paulus, M.P., Feinstein, J.S., Castillo, G., Simmons, A.N., Stein, M.B., 2005. Dose-
dependent decrease of activation in bilateral amygdala and insula by lorazepam
during emotion processing. Archives of General Psychiatry 62, 282–288.
Paulus, M.P., Stein, M.B., 2006. An insular view of anxiety. Biological Psychiatry 60,
383–387.
Resick, P.A., Schnicke, M.K., 1992. Cognitive processing therapy for sexual assault
victims. Journal of Consulting and Clinical Psychology 60, 748–756.
Roy, M.J., Francis, J., Friedlander, J., Banks-Williams, L., Lande, R.G., Taylor, P., Blair, J.,
McLellan, J., Law, W., Tarpley, V., Patt, I., Yu, H., Mallinger, A., Difede, J., Rizzo, A.,
Rothbaum, B., 2010. Improvement in cerebral function with treatment of
posttraumatic stress disorder. Annals of the New York Academy of Sciences
1208, 142–149.
Schottenbauer, M.A., Glass, C.R., Arnkoff, D.B., Tendick, V., Gray, S.H., 2008.
Nonresponse and dropout rates in outcome studies on PTSD: review and
methodological considerations. Psychiatry 71, 134–168.
Shin, L.M., Liberzon, I., 2010. The neurocircuitry of fear, stress, and anxiety
disorders. Neuropsychopharmacology 35, 169–191.
Simmons, A., Matthews, S.C., Stein, M.B., Paulus, M.P., 2004. Anticipation of
emotionally aversive visual stimuli activates right insula. Neuroreport 15,
2261–2265.
Simmons, A., Strigo, I., Matthews, S.C., Paulus, M.P., Stein, M.B., 2006. Anticipation of
aversive visual stimuli is associated with increased insula activation in anxiety-
prone subjects. Biological Psychiatry 60, 402–409.
Simmons, A.N., Paulus, M.P., Thorp, S.R., Matthews, S.C., Norman, S.B., Stein, M.B.,
2008. Functional activation and neural networks in women with posttraumatic
stress disorder related to intimate partner violence. Biological Psychiatry 64,
681–690.
Stein, D.J., Ipser, J.C., Seedat, S., 2006. Pharmacotherapy for post traumatic stress
disorder (PTSD). Cochrane database of systematic reviews 1, CD002795.
Thomaes, K., Dorrepaal, E., Draijer, N., de Ruiter, M.B., Elzinga, B.M., van Balkom, A.J.,
Smit, J.H., Veltman, D.J., 2012. Treatment effects on insular and anterior
cingulate cortex activation during classic and emotional Stroop interference
in child abuse-related complex post-traumatic stress disorder. Psychological
Medicine, 1–13.