Perceived threat predicts the neural sequelae of combat stress.

0.5 MB·English

by  van Wingen, G A

MolecularPsychiatry(2011)16,664–671
&2011MacmillanPublishersLimited Allrightsreserved1359-4184/11
www.nature.com/mp
ORIGINAL ARTICLE
Perceived threat predicts the neural sequelae of combat
stress
GA van Wingen1,2,3, E Geuze4,5, E Vermetten4,5 and G Ferna´ndez1,6
1DondersInstituteforBrain, Cognitionand Behaviour, RadboudUniversityNijmegen, Nijmegen, The Netherlands;
2DepartmentofNeurology, RadboudUniversity NijmegenMedical Centre, Nijmegen, TheNetherlands; 3Department of
Psychiatry, AcademicMedical Center, Universityof Amsterdam,Amsterdam,The Netherlands;4ResearchCentre, Military
Mental Health, MinistryofDefense, Utrecht,The Netherlands;5Department ofPsychiatry,Rudolf Magnus Instituteof
Neuroscience,Utrecht University MedicalCenter, Utrecht,The Netherlands and6DepartmentforCognitive Neuroscience,
Radboud UniversityNijmegen MedicalCentre, Nijmegen, The Netherlands
Exposure to severe stressors increases the risk for psychiatric disorders in vulnerable
individuals, but can lead to positive outcomes for others. However, it remains unknown how
severe stress affects neural functioning in humans and what factors mediate individual
differences in the neural sequelae of stress. The amygdala is a key brain region involved in
threat detection and fear regulation, and previous animal studies have suggested that stress
sensitizes amygdala responsivity and reduces its regulation by the prefrontal cortex. In this
study, we used a prospective design to investigate the consequences of severe stress in
soldiersbeforeandafterdeploymenttoacombatzone.Wefoundthatcombatstressincreased
amygdala and insula reactivity to biologically salient stimuli across the group of combat-
exposedindividuals.Incontrast,itsinfluenceonamygdalacouplingwiththeinsulaanddorsal
anterior cingulate cortex was dependent on perceived threat, rather than actual exposure,
suggestingthatthreatappraisalaffectsinteroceptiveawarenessandamygdalaregulation.Our
results demonstrate that combat stress has sustained consequences on neural responsivity,
andsuggestakeyrolefortheappraisalofthreatonanamygdala-centeredneuralnetworkin
the aftermath ofseverestress.
MolecularPsychiatry(2011)16,664–671;doi:10.1038/mp.2010.132;publishedonline18January2011
Keywords: ACC; amygdala; combat; fMRI; PTSD; stress
Introduction tion.9–11TheseneuralalterationsinPTSDpresumably
result from a combination of stress exposure and
Highlystressfulexperiences,suchasnaturaldisaster,
stress vulnerability.12–14 Nevertheless, animal studies
terrorism, assault or military combat, are significant
suggest that severe stress has lasting influences on
eventsinthelivesofpeople.Suchstressfullifeevents
neural functioning.15,16 Specifically, recent animal
have large impact on the exposed individual, but its
studies have shown that prolonged severe stress
influence on subsequent psychological well-being is
sensitizes amygdala responsivity and reduces its
highly variable between individuals. Whereas trau-
regulation by the prefrontal cortex.17,18 Furthermore,
matic experiences lead to positive changes for
acutestressexposureincreasesamygdalareactivityin
some,1,2 they increase the risk for post-traumatic
humans.19 To investigate whether prolonged severe
stress disorder (PTSD) and depression for others,3,4
stress also affects amygdala functioning in humans,
which suggests that the influence of stress is highly
even in the absence of acute stress, we used a
variable between individuals. However, the neural
prospective study design. We assessed amygdala
consequenceofseverestressonneuralfunctioningin
functioning in soldiers before and after deployment
humansremainsunknown.PatientswithPTSDshow
to a combat zone, which is typically associated with
exaggerated amygdala responses and deficient pre-
severe stress exposure. In addition, we included a
frontal cortex function, in particular in the anterior
groupofsoldierswhowereneverdeployedtocontrol
cingulatecortexandventromedialprefrontalcortex.5–8
for repeated testing and unspecific time effects.
The amygdala and prefrontal cortex are key brain
regions involved in threat detection and fear regula-
Materials and methods
Correspondence: Dr GAvan Wingen, Department of Psychiatry, Participants
Academic Medical Center, University of Amsterdam, PO Box The combat stress group consisted of 33 healthy
22660,Amsterdam1100DD,TheNetherlands.
soldierswhoweredeployedfor4monthstoAfghanistan
E-mail:[email protected]
as part of the North Atlantic Treaty Organization
Received 17 July 2010; revised 4 October 2010; accepted 13
November2010;publishedonline18January2011 InternationalSecurityAssistanceForcepeacekeeping
Neuralsequelaeofcombatstress
GAvanWingenetal
665
operation. We investigated them before and on validated Dutch questionnaire, the self-rating inven-
average1.5(s.d.=0.8)monthsaftertheirfirstmilitary tory for PTSD, was used to assess self-reported PTSD
deployment. They were recruited from a larger symptoms. We excluded individuals with possible
prospective study on the development of stress- PTSDfromallanalyses,asdefinedbythecutoffscore
related disorders following military deployment in of 52.20 The Dutch version of the Positive and
theDutcharmedforces.Theirdutiesincludedcombat Negative Affect Schedule was used to assess positive
patrols, clearing or searching homes and buildings, and negative mood,21 and Dutch state version of the
participation in demining operations and transporta- State Trait Anxiety Inventory was administered to
tion across enemy territory. They were exposed to assess state anxiety.22 To quantify individual differ-
typicalwar-zonestressors, suchasenemyfire, armed ences in stress exposure, we measured combat
combat and seeing seriously injured fellow soldiers exposure and perceived threat during deployment
and civilians (including women and children). The using the Combat experiences and Perceived threat
control group consisted of 26 soldiers who were scales of the Deployment Risk and Resilience Inven-
recruited from training bases and army divisions tory.23 The scores on these scales were not signifi-
currently not involved in combat missions. One cantly correlated (r=0.10, P=0.58), suggesting that
soldier of the control group scored above a clinical theamountofperceivedthreatduringthetimeofthe
threshold for PTSD symptoms on a self-report ques- military deployment was not directly related to the
tionnaire (see below) at baseline, and one soldier of number of actual combat experiences.
the combat group scored above the threshold after
deployment. Both participants were therefore ex- Behavioral task
cluded from the analyses. The groups did not differ The experimental paradigm that was performed
significantly in sex ratio, age and intelligence quo- during functional magnetic resonance imaging scan-
tient(seeTable1).Furthermore,weinvestigatedboth ning consisted of a blocked design, including an
groups at the same test-retest interval (mean±s.d.; emotionconditionwithangryandfearfulfacestimuli
6.9±1.6 months; t(55)=0.5, P=0.65). Because the (http://www.macbrain.org) and a visuomotor control
number of females in the study was small, we also condition with scrambled ellipse stimuli.24–26 Two
conducted all analyses with males only, which did emotion blocks were interleaved with three control
not alter the pattern of results (data not shown). The blocks, and each 30s block consisted of six 5s trials.
study was in accordance with the declaration of Eachtrialconsistedofthreesimultaneouslypresented
Helsinki and institutional guidelines of the local stimuli, with the cue stimulus presented above the
ethicscommittee(CMORegioArnhem-Nijmegen,The target and distractor. In the emotion condition, an
Netherlands), and all participants provided written angryorfearfulfacewaspresentedontopascue,and
informedconsentafterwrittenandoraldescriptionof participants had to indicate by an appropriate button
the study. press which of the bottom two faces (one angry and
onefearful)matchedthecueinemotionalexpression.
Questionnaires In the control condition, a horizontally or vertically
Toevaluatetheinfluenceofseverestressexposureon oriented ellipse was presented as cue above two
PTSD symptoms, mood and anxiety, we used three ellipses (one vertical and one horizontal), and parti-
questionnaires. The short version of a previously cipants had to select the identically oriented ellipse.
Table1 Demographicandquestionnairedata(mean±s.d.)
Combatstressgroup(n=32) Controlgroup(n=25) Statistics
Baseline Follow-up Baseline Follow-up Baselinea Group(cid:2)timeb
Mean s.d. Mean s.d. Mean s.d. Mean s.d. P P
Age 24.3 8.0 23.9 6.7 0.82
IQ 89.4 10.6 92.1 13.3 0.40
Sex(M/F) 31/1 23/2 0.41c
Questionnaires
PTSDsymptoms 27.6 4.6 27.6 5.9 26.5 3.8 26.8 4.6 0.35 0.77
Positiveaffect 32.0 5.5 33.1 5.6 32.6 5.0 31.4 6.7 0.64 0.10
Negativeaffect 12.9 4.8 11.6 2.3 11.5 2.6 11.2 2.0 0.21 0.26
Stateanxiety 31.6 7.6 30.7 6.7 30.2 6.8 28.2 5.3 0.54 0.50
Abbreviations:F,female;IQ,intelligencequotient;M,male;PTSD,post-traumaticstressdisorder.
aTwo-samplet-test
bGroup(cid:2)timeanalysisofvariance.
cPearson’sw2.
MolecularPsychiatry
Neuralsequelaeofcombatstress
GAvanWingenetal
666
Functional magnetic resonance image acquisition Voxel-based morphometry
Magnetic resonance data were acquired with a 1.5T To assess the influence of severe stress on regional
Siemens (Erlangen, Germany) Avanto magnetic reso- gray matter volume, we performed voxel-based
nance scanner, equipped with a standard head coil. morphometry using SPM5 with standard proce-
T2*-weighted bold images were acquired using echo dures29 and default parameters of the VBM5 toolbox
plannerimagingwithanechotimeof35mstoreduce (http://dbm.neuro.uni-jena.de/vbm.html).30 Analyses
signal dropout in the medial temporal lobes. Each were performed on gray matter segmentations, which
image volume consisted of 32 axial slices (3.5mm, weremultipliedbythenonlinearcomponentsderived
0.35mm slice gap, repetition time (TR)=2.340s, from the normalization matrix in order to preserve
64(cid:2)64 matrix, field of view=212mm, flip angle local gray matter volumes (that is, the modulated
(FA)=901). In addition, a high-resolution T1- images) and spatially smoothed (12mm full width at
weighted structural magnetic resonance image with half maximum).
optimized gray/white matter contrast was acquired
(three-dimensional magnetization-prepared rapid ac- Statistical testing
quisition with gradient echo, 1(cid:2)1(cid:2)1mm3 voxels, Mixed model analysis of variances with the factors
TR=2.730s, echo time (TE)=2.95ms, inversion time group and time were used to test whether changes in
(TI)=1000ms, field of view=256mm, FA=7). brain structure and function over time were different
betweenthecombatstressandcontrolgroups.Totest
whether stress-induced changes were related to
Functional magnetic resonance data analysis individual differences in combat exposure and per-
Image analysis was performed with SPM5 (http:// ceived threat, additional correlation analyses were
www.fil.ion.ucl.ac.uk/spm). The first five echo plan- performed.Statisticaltestswerefamily-wiseerrorrate
nerimagingvolumeswerediscardedandtheremain- corrected (P<0.05) for multiple comparisons across
ing imageswere realignedto the firstvolume. Images the entire brain, or for the search volume for regions
were then co-registered to the anatomical scan, of interest using a small volume correction.31 The
corrected for differences in slice acquisition time, search volumes for the amygdala, insula and anterior
spatially normalized to the Montreal Neurological cingulate cortex (Brodmann areas 24 and 32) were
Institute (MNI) T1 template, resampled into anatomically defined using the WFU Pickatlas tool-
2(cid:2)2(cid:2)2mm3 voxels, and spatially smoothed (8mm box implemented in SPM5.32
full width at half maximum). Statistical analysis was
performedwithintheframeworkofthegenerallinear
Results
model.27 To assess the influence of combat stress on
neural responsivity, the two experimental conditions Questionnaires
were modeled as box-car regressors convolved with We observed no significant differences in self-
the canonical hemodynamic response function of reported PTSD symptoms, state anxiety, positive
SPM5. In addition, the realignment parameters were affect and negative affect scores at baseline, nor
included to model potential movement artefacts, as different changes in these variables between the
well as a constant. Furthermore, a high-pass filter combat stress group and control group (see Table 1),
(cutoff 1/128Hz) was included, and temporal auto- indicating that changes in neural activity did not
correlation was modelled with an AR(1) process. reflect increases in symptomatology. In addition, we
Contrast images subtracting the visuomotor control measured combat exposure and perceived threat
condition fromthe emotion condition were obtained, during deployment to quantify the individual differ-
and analyzed in subsequent random effects models. encesinstressexposure.Theaveragescoreforcombat
exposure (mean±s.d.; 5.0±2.5) was higher than that
of a previously reported reference population of Gulf
Functional connectivity War veterans (4.0±3.2),23 confirming that the combat
To assess the influence of severe stress on the group was exposed to typical combat zone stressors
amygdala network, we performed an additional such as armed combat, combat patrols, and exposure
functionalconnectivity(psychophysiologicalinterac- toenemyfire,aswellasmodernwarfarewithariskof
tion) analysis using SPM5. The time course of exposuretoimprovisedexplosivedevises.Toexplore
amygdala activity was obtained for each scanning whether there was a delayed onset of PTSD symp-
session. The first eigenvariate of a sphere with 5mm toms, we contacted the deployed individuals again 6
radius around the peak group(cid:2)time interaction in months after their deployment, but observed no
therightamygdala(24,(cid:3)4and(cid:3)18)wasextractedand significant changes in PTSD symptoms over time in
entered as an additional regressor to the original those individuals that completed all three question-
functional magnetic resonance imaging model, as naires (n=16; F(2,14)=1.3, P=0.31).
well as the interaction between the experimental
conditions and amygdala time course.28 Psychophy- Behavioral performance
siological interaction (time course(cid:2)condition) Acrossbothgroupsandmeasurements,task accuracy
images were obtained, and analyzed in subsequent was higher in the emotion than control condition
random effects models. (Z=(cid:3)2.0, P=0.04; median±interquartile range;
MolecularPsychiatry
Neuralsequelaeofcombatstress
GAvanWingenetal
667
emotion condition: 100.0±4.2%, control condition: control group (P<0.001), but these additional tests
97.2±5.6%)butreactiontimeswereslower(F(1,55)= did not remain significant after correction for multi-
171.1, P<0.001; mean±s.d.; emotion condition: ple comparisons. These results demonstrate that
1863±346ms, control condition: 1141±391ms). severe and long-term stress exposure sensitized
However,nosignificantchangesinaccuracy(emotion amygdala and insula reactivity to biologically salient
condition: U=362.0, Z=(cid:3)0.73, P=0.47; control con- stimuli in humans. Next, to evaluate whether these
dition:U=387.0,Z=(cid:3)0.22,P=0.83)orreactiontimes activity changes over time were related to individual
(F<1) over time between the combat stress and differences in stress exposure, we performed addi-
control groups were observed, which suggests that tional correlationanalyses. However, weobserved no
the imaging results are unlikely explained by differ- significant correlations between activity changes and
ences in behavioral performance. combat experiences or perceived threat (P >0.05).
cor
Neural responsivity Amygdala connectivity
Toverifythatthetaskindeedactivatedtheamygdala, To investigate whether the influence of severe stress
we compared the emotion condition with the control on amygdala reactivity also affected the neural
condition across both groups and measurements. As network that is centered around the amygdala,11,35
expected,thetaskincreasedactivityintheamygdala, we performed a functional connectivity analysis.
as well as in the insula, dorsal anterior cingulate Across groups and investigations, the amygdala was
cortex (dACC), occipitotemporal cortex, hippocam- negatively coupled to cingulate cortex areas (see
pus, orbitofrontal cortex, inferior and middle frontal Figure 2a) including the dorsal anterior cingulate
gyri, thalamus, precuneus, angular gyrus and cere- cortex (dACC; (6, 8 and 44), Z=4.1, P =0.017),
cor
bellum(allP <0.05).Theseactivationpatternswere pregenual ACC (6, 48 and 2; Z=3.8, P =0.040,
cor cor
not significantly different between groups at baseline middle cingulate cortex; 4, (cid:3)20 and 42; Z=4.3,
(P >0.05), suggesting that neural functioning was P =0.006) and posterior cingulate cortex ((cid:3)2,(cid:3)36
cor cor
comparable before stress exposure. and 26), Z=5.9, P <0.001). These connectivity
cor
Next, we compared the change in amygdala patterns were not significantly different between
reactivity over time between groups. The increase in groups at baseline (P >0.05), suggesting that amyg-
cor
amygdala reactivity was significantly larger in the dalacouplingwascomparablebeforestressexposure.
combat stress group than the control group (peak Changes in functional connectivity over time were
MontrealNeurologicalInstitutecoordinate(x,yandz); not significantly different between the groups
(24, (cid:3)4 and (cid:3)18); Z=3.1, P =0.037). Subsequent (P >0.05), but were related to the amount of
cor cor
tests showed that combat stress exposure increased perceived threat in stress-exposed individuals. Im-
amygdala reactivity (30, 0 and(cid:3)18; Z=3.1, portantly,thestrengthofamygdalacouplingwiththe
P =0.044; see Figure 1a), whereas no significant dACCwaspositivelyrelatedtoperceivedthreat((cid:3)10,
cor
change in activity was observed in the control group 0 and 42; Z=4.5, P =0.003; see Figures 2b and c).
cor
(P >0.05). A similar group(cid:2)time interaction was Thus, negative coupling of the amygdala with the
cor
also observed in the anterior insula (32, 24 and 2; dACC was enhanced in individuals that perceived
Z=3.8, P =0.037; see Figure 1b), which is a brain littlethreat,butreducedinindividualsthatperceived
cor
region involved in interoceptive awareness and much threat. In addition, perceived threat also
thoughttosignalinternalbodystates.33,34Subsequent enhanced amygdala coupling with the insula (42, 16
tests showed that combat stress exposure also and(cid:3)2;Z=3.8,P =0.039;seeFigure3).Incontrast,
cor
increasedinsulareactivity(P=0.002),whereasinsula no significant correlations between actual combat
reactivity decreased with repeated testing in the exposure and functional connectivity changes were
Figure 1 Severe stress exposure increases amygdala and insula reactivity to biologically salient stimuli. (a) Combat
exposure increased amygdala reactivity in military soldiers, whereas no significant change in amygdala reactivity was
observed in soldiers that were never deployed. (b) Combat exposure also increased insula reactivity in soldiers relative to
responsehabituationovertimeinthecontrolgroup.Thefiguresshowstatisticalparametricmapsillustratingthesignificant
effects(P<0.05,corrected)atP<0.005,uncorrected.
MolecularPsychiatry
Neuralsequelaeofcombatstress
GAvanWingenetal
668
Figure2 Individualdifferencesinperceivedthreatduringmilitarydeploymentaffectsfunctionalcouplingoftheamygdala
with the dorsal anterior cingulate cortex (dACC). (a) The dACC and other midline structures were in general negatively
coupled to the amygdala when analyzed across stress exposure groups and testing sessions. (b) The change in amygdala
coupling with the dACC after stress exposure was positively correlated to perceived threat during military deployment.
(c)Thescatterplotillustratesthecorrelationinpanelbatthepeakvoxel.Panelsaandbshowstatisticalparametricmaps
illustratingthesignificanteffects(P<0.05,corrected)atP<0.005,uncorrected.
Figure3 Individualdifferencesinperceivedthreatduringmilitarydeploymentaffectsfunctionalcouplingoftheamygdala
with the insula. (a) The change in amygdala coupling with the insula after stress exposure was positively correlated to
perceived threat during military deployment. The statistical parametric map illustrates the significant effect (P<0.05,
corrected)atP<0.005,uncorrected.(b)Thescatterplotillustratesthecorrelationinpanelaatthepeakvoxel.
observed(P >0.05).Moreover,theassociationswith neural changes were not observed in the control
cor
perceived threat remained significant after correcting group and occurred in the absence of self-reported
for combatexposure,suggestingthat perceivedthreat changes in post-traumatic stress symptoms. This
rather than actual exposure influenced amygdala suggests that sustained enhanced reactivity to biolo-
coupling. gicallysalientstimuliisacommonadaptiveresponse
toprolongedenvironmentalthreat.Theamygdalaisa
Brain structure crucialbrainregionforthedetectionofthreatandthe
To investigate whether the stress-induced changes in enhancementofvigilance,andboththeamygdalaand
neural activity and connectivity were related to insula arepart of a larger salience network.9,36There-
changes in brain structure, we compared regional fore these adaptations may be beneficial to maintain
brain volumes using voxel-based morphometry. We sustained vigilance in continuously dangerous situa-
observed no significant volume differences at base- tions, such as the combat zone the soldiers were
line, no significant changes over time between the deployed to. But as heightened amygdala and insula
combat stress group and control group, and no reactivityisthoughttoincreasetheriskformoodand
significant correlations between combat experiences anxiety disorders,37,38 these alterations may not be
or perceived threat and regional volume changes adaptive for safe environments. Although our second
(P >0.05). follow-up assessment suggests that the combat group
cor
did not develop PTSD symptoms within half a year
after deployment, it remains possible that stress
Discussion
symptoms do evolve later on. Furthermore, the
To study the neural consequences of severe stress amygdala and insula sensitization may also increase
exposure, we investigated soldiers before and after the vulnerability to future stressors.
deployment to a combat zone. Our results demon- Toinvestigatewhethertheinfluenceofseverestress
strate that combat stress increases amygdala and on amygdala reactivity also affected the neural
insula reactivitytobiologically salientstimuli. These network that is centered around the amygdala, we
MolecularPsychiatry
Neuralsequelaeofcombatstress
GAvanWingenetal
669
performed an additional functional connectivity prevent the development of symptoms. Interestingly,
analysis.Ourresultsshowthattheinfluenceofsevere patients with PTSD have abnormalities in the
stress on amygdala coupling with the dACC and hypothalamic–pituitary–adrenal axis,43 and soldiers
insula was dependent on individual differences in that develop PTSD symptoms after combat exposure
perceived threat. Individuals that perceived little have a high preexisting number of glucocorticoid
threat duringdeployment showed enhanced negative receptors.44ThissuggeststhatPTSDsymptomscould
amygdala–dACC coupling, whereas individuals that resultfromunsuccessfulnormalizationafterexposure
perceived much threat showed reduced coupling. to consecutive stressors.45,46 Therefore, we propose
Given that the dACC is thought to regulate amygdala that whereas the stress-induced changes in neural
activity,10,11 this finding suggests that the perception reactivity appear a common adaptive response to
ofthreatduringseverestressexposurealtersamygda- prolongedenvironmentalthreat,maladaptivenormal-
la regulation thereafter. In addition, perceived threat ization of neural hyperactivity may lead to the
also enhanced amygdala–insula coupling. The insula development of stress symptoms, which could be
isinvolvedininteroceptiveawarenessandisthought addressed in future studies.
to signalinternalbody states,33,34which suggests that The neural sequelae of severe stress exposure were
this may reflect increased bodily awareness in those remarkably similar to the neural alterations in PTSD,
individuals that perceive the most threat. whichincludeamygdalaandinsulahyperactivityand
Theinfluenceofperceivedthreatratherthanactual dACC hypoactivity.8,47 This is in line with the idea
combat exposure on the amygdala network is in line that PTSD reflects the upper end of a stress response
with the appraisal theory, which posits that the continuum.48 However, the neural basis of PTSD
cognitive appraisal of threat rather than the actual presumably results from a combination of stress
environmental stressor determines the impact of exposure and stress vulnerability.12,49 Previous stu-
stress exposure.39,40 Thus, our results suggest that dies have shown that high amygdala reactivity to
cognitive appraisal also has a critical role in deter- masked stimuli (which prevents conscious aware-
mining the impact of severe stress on the amygdala ness) and dACC metabolism before stress exposure
network. Interestingly, perceived threat also seems a increase the risk for subsequent stress symptoms.13,14
better predictor for PTSD symptoms than actual Our results now suggest that sustained increases in
combat exposure.23 Even though previous studies amygdala and insula reactivity to biologically salient
have demonstrated a close relation between combat stimuli reflect a consequence of stress exposure,
experiences and the prevalence of PTSD,4 path which may be additive to preexisting vulnerability
analyses suggest that the influences of combat factors. In contrast, the individual differences in
exposure on PTSD symptoms is mediated by its altered amygdala coupling suggest that the impact of
influence on perceived threat.41 stress on the amygdala network is a consequence of
The divergent influences of severe stress on the theinteractionbetweenstressvulnerabilityandstress
amygdala network may help to explain why some exposure.
individuals are vulnerable to stress, whereas others The similarity between the neural consequences of
are stress resilient. Our results show opposite effects stress exposure and PTSD pathophysiology has
on amygdala–dACC and amygdala–insula coupling important implications for studies investigating the
depending on how the stressful experience is per- neural basis of PTSD. Previous neuroimaging studies
ceived. In turn, this may lead to opposite effects on with PTSD patients have included a control group of
amygdala regulation and interoceptive awareness, trauma-exposed individuals that did not develop
with better emotion regulation in those individuals PTSD, to control for trauma exposure (for example,
that perceive little threat but worse emotion regula- see Rauch et al.5). Our results underscore the
tion in those that experience much threat. These importance for this procedure by demonstrating that
individual differences may explain in part why we trauma exposure by itself also influences brain
did not observe consistent changes in symptomatol- functioning. Comparing PTSD patients to non-trau-
ogy across the group of combat-exposed individuals, ma-exposed individuals may result in misattribution
eventhoughtheiramygdalaandinsulareactivityhad of group differences to PTSD pathophysiology.
increased, and suggestaneuralmechanismbywhich One of the 33 soldiers (3%) scored above the
traumatic experiences could lead to highly variable clinicalthresholdforPTSDsymptomsonaself-report
outcomes.1–4 questionnaire after their 4-month deployment to
The stress-induced changes in the amygdala net- Afghanistan. This is comparable with PTSD preva-
work didnot leadto consistentchanges inmood and lence rates for Dutch and United Kingdom soldiers
anxiety as assessed here. We have measured the that have been deployed to Iraq or Afghanistan,50,51
changes in neural reactivity during affective stimula- but lower than the estimate for United States
tion. But in addition to increased reactivity to soldiers.4 Differences in these prevalence rates may
biologically salient stimuli, combat exposure may have various causes, including differences in study
also have altered the manner in which the brain populations,level ofcombatexposure,and screening
restoresafterstressexposure.Forexample,therelease methods.52Althoughoursampleseemsrepresentative
of cortisol after stress exposure normalizes amygdala for the military population, it may not be representa-
reactivity,42andadaptivechangesinthisprocessmay tive for the general population. For example, the
MolecularPsychiatry
Neuralsequelaeofcombatstress
GAvanWingenetal
670
process of self-selection to serve in the military may 11 Stein JL, Wiedholz LM, Bassett DS, Weinberger DR, Zink CF,
lead to a relatively resilient population. To address Mattay VS et al. A validated network of effective amygdala
this issue, future studies may investigate the con- connectivity.Neuroimage2007;36:736–745.
12 TrueWR,RiceJ,EisenSA,HeathAC,GoldbergJ,LyonsMJetal.A
sequences of severe and traumatic stress in non-
twinstudyofgeneticandenvironmentalcontributionstoliability
military samples.
forposttraumaticstresssymptoms.ArchGenPsychiatry1993;50:
In conclusion, our results demonstrate that severe 257–264.
stress exposure sensitizes amygdala and insula 13 AdmonR,LubinG,SternO,RosenbergK,SelaL,Ben-AmiHetal.
reactivity.Inaddition,individualdifferencesinthreat Humanvulnerabilitytostressdependsonamygdala’spredisposi-
tion and hippocampal plasticity. Proc Natl Acad Sci USA 2009;
perception predicted divergent influences on the
106:14120–14125.
amygdala network, which may explain why some 14 Shin LM, Lasko NB, Macklin ML, Karpf RD, Milad MR, Orr SP
individuals are vulnerable to stress, whereas others et al. Resting metabolic activity in the cingulate cortex and
are stress resilient. Long-term follow-up studies are vulnerabilitytoposttraumaticstressdisorder.ArchGenPsychiatry
2009;66:1099–1107.
required to determine whether these stress-induced
15 Pavlides C, Nivon LG, McEwen BS. Effects of chronic stress on
neural changes indeed represent resilience orvulner-
hippocampal long-term potentiation. Hippocampus 2002; 12:
ability factors. 245–257.
16 Alfarez DN, Joe¨ls M, Krugers HJ. Chronic unpredictable stress
impairslong-termpotentiationinrathippocampalCA1areaand
Conflict of interest dentategyrusinvitro.EurJNeurosci2003;17:1928–1934.
17 CorrellCM,RosenkranzJA,GraceAA.Chroniccoldstressalters
The authors declare no conflict of interest. prefrontal cortical modulation of amygdala neuronal activity in
rats.BiolPsychiatry2005;58:382–391.
18 Rosenkranz JA, Venheim ER, Padival M. Chronic stress causes
amygdala hyperexcitability in rodents. Biol Psychiatry 2010; 67:
Acknowledgments 1128–1136.
19 vanMarleHJ,HermansEJ,QinS,Ferna´ndezG.Fromspecificityto
We thank A Beekman and M Groenewald for their sensitivity: how acute stress affects amygdala processing of
assistancewithdatacollectionandPGaalmanforhis biologicallysalientstimuli.BiolPsychiatry2009;66:649–655.
technical support. This work was supported by a 20 Hovens JE, van der Ploeg HM, Bramsen I, Klaarenbeek MT,
Schreuder JN, Rivero VV. The development of the self-rating
grant(918.66.613)fromtheNetherlandsOrganization
inventoryforposttraumaticstressdisorder.ActaPsychiatrScand
forScientificResearch(NWO)andtheDutchMinistry 1994;90:172–183.
ofDefense.Thesponsorshadnoroleinstudydesign; 21 WatsonD,ClarkLA,TellegenA.Developmentandvalidation of
in the collection, analysis and interpretation of data; briefmeasuresofpositiveandnegativeaffect:thePANASscales.
JPersSocPsychol1988;54:1063–1070.
in the writing of the report; and in the decision to
22 Spielberger CD, Gorsuch RL, Lushene RE. STAI Manual for the
submit the paper for publication.
StateTraitAnxietyInventory.ConsultingPsychologistsPress:Palo
Alto,1970.
23 KingLA,KingDW,VogtDS,KnightJ,SamperRE.Deploymentrisk
References and resilience inventory: a collection of measures for studying
deployment-related experiences of military personnel and veter-
1 Helgeson VS, Reynolds KA, Tomich PL. A meta-analytic review ans.MilPsychol2006;18:89–120.
of benefit finding and growth. J Consult Clin Psychol 2006; 74: 24 HaririAR,MattayVS,TessitoreA,KolachanaB,FeraF,Goldman
797–816. Detal.Serotonintransportergeneticvariationandtheresponseof
2 Linley PA, Joseph S. Positive change following trauma and thehumanamygdala.Science2002;297:400–403.
adversity:areview.JTraumaStress2004;17:11–21. 25 van Wingen GA, van Broekhoven F, Verkes RJ, Petersson KM,
3 Galea S, Ahern J, Resnick H, Kilpatrick D, Bucuvalas M, Gold J Ba¨ckstro¨mT,BuitelaarJKetal.Progesteroneselectivelyincreases
etal.PsychologicalsequelaeoftheSeptember11terroristattacks amygdalareactivityinwomen.MolPsychiatry2008;13:325–333.
inNewYorkCity.NEnglJMed2002;346:982–987. 26 van Wingen GA, Zylicz SA, Pieters S, Mattern C, Verkes RJ,
4 HogeCW,CastroCA,MesserSC,McGurkD,CottingDI,Koffman Buitelaar JK et al. Testosterone increases amygdala reactivity in
RL.CombatdutyinIraqandAfghanistan,mentalhealthproblems, middle-aged women to a young adulthood level. Neuropsycho-
andbarrierstocare.NEnglJMed2004;351:13–22. pharmacology2009;34:539–547.
5 RauchSL,WhalenPJ,ShinLM,McInerneySC,MacklinML,Lasko 27 Friston KJ, Holmes AP, Worsley KJ, Poline JB, Frith CD,
NBetal.Exaggeratedamygdalaresponsetomaskedfacialstimuli Frackowiak RSJ. Statistical parametric maps in functional
in posttraumatic stress disorder: a functional MRI study. Biol imaging: a general linear approach. Hum Brain Mapp 1995; 2:
Psychiatry2000;47:769–776. 189–210.
6 ShinLM,WhalenPJ,PitmanRK,BushG,MacklinML,LaskoNB 28 Friston KJ, Buechel C, Fink GR, Morris J, Rolls E, Dolan RJ.
et al. An fMRI study of anterior cingulate function in posttrau- Psychophysiological and modulatory interactions in neuroima-
maticstressdisorder.BiolPsychiatry2001;50:932–942. ging.Neuroimage1997;6:218–229.
7 Shin LM, Wright CI, Cannistraro PA, Wedig MM, McMullin K, 29 AshburnerJ,FristonKJ.Unifiedsegmentation.Neuroimage2005;
MartisBetal.Afunctionalmagneticresonanceimagingstudyof 26:839–851.
amygdala and medial prefrontal cortex responses to overtly 30 LudersE,TogaAW,LeporeN,GaserC.Theunderlyinganatomical
presented fearful faces in posttraumatic stress disorder. Arch correlatesoflong-termmeditation:largerhippocampalandfrontal
GenPsychiatry2005;62:273–281. volumesofgraymatter.Neuroimage2009;45:672–678.
8 EtkinA,WagerTD.Functionalneuroimagingofanxiety:ameta- 31 Worsley KJ, Marrett S, Neelin P, Vandal AC, Friston KJ,
analysisofemotionalprocessinginPTSD,socialanxietydisorder, Evans AC. A unified statistical approach for determining sig-
andspecificphobia.AmJPsychiatry2007;164:1476–1488. nificantsignalsinimagesofcerebralactivation.HumBrainMapp
9 Davis M, Whalen PJ. The amygdala: vigilance and emotion. Mol 1996;4:58–73.
Psychiatry2001;6:13–34. 32 Maldjian JA, Laurienti PJ, Kraft RA, Burdette JH. An automated
10 Quirk GJ, Beer JS. Prefrontal involvement in the regulation of method for neuroanatomic and cytoarchitectonic atlas-
emotion: convergence of rat and human studies. Curr Opin based interrogation of fMRI data sets. Neuroimage 2003; 19:
Neurobiol2006;16:723–727. 1233–1239.
MolecularPsychiatry
Neuralsequelaeofcombatstress
GAvanWingenetal
671
33 Craig AD. How do you feel? Interoception: the sense of the predicts development of posttraumatic stress symptoms after
physiological condition of the body. Nat Rev Neurosci 2002; 3: militarydeployment.AmJPsychiatry2011;168:89–96.
655–666. 45 de Kloet ER, Joe¨ls M, Holsboer F. Stress and the brain: from
34 PaulusMP,SteinMB.Aninsularviewofanxiety.BiolPsychiatry adaptationtodisease.NatRevNeurosci2005;6:463–475.
2006;60:383–387. 46 KarstH,BergerS,ErdmannG,SchutzG,JoelsM.Metaplasticityof
35 Ghashghaei HT,HilgetagCC,Barbas H.Sequenceofinformation amygdalar responses to the stress hormone corticosterone. Proc
processing for emotions based on the anatomic dialogue NatlAcadSciUSA2010;107:14449–14454.
between prefrontal cortex and amygdala. Neuroimage 2007; 34: 47 Rauch SL, Shin LM, Phelps EA. Neurocircuitry models of
905–923. posttraumatic stress disorder and extinction: human neuroima-
36 SeeleyWW,MenonV,SchatzbergAF,KellerJ,GloverGH,Kenna gingresearch-past,present,andfuture.BiolPsychiatry2006;60:
H et al. Dissociable intrinsic connectivity networks for salience 376–382.
processingandexecutivecontrol.JNeurosci2007;27:2349–2356. 48 Ruscio AM, Ruscio J, Keane TM. The latent structure of
37 Stein MB, Simmons AN, Feinstein JS, Paulus MP. Increased posttraumatic stress disorder: a taxometric investigation
amygdala and insula activation during emotion processing in of reactions to extreme stress. J Abnorm Psychol 2002; 111:
anxiety-pronesubjects.AmJPsychiatry2007;164:318–327. 290–301.
38 Wolfensberger SP, Veltman DJ, Hoogendijk WJ, Boomsma DI, de 49 Gilbertson MW, Shenton ME, Ciszewski A, Kasai K, Lasko NB,
Geus EJ. Amygdala responses to emotional faces in twins Orr SP et al. Smaller hippocampal volume predicts pathologic
discordantorconcordantfortheriskforanxietyanddepression. vulnerability to psychological trauma. Nat Neurosci 2002; 5:
Neuroimage2008;41:544–552. 1242–1247.
39 LazarusRS, Folkman S.Stress, Appraisal,andCoping.Springer 50 HotopfM,HullL,FearNT,BrowneT,HornO,IversenAetal.The
PublishingCompany:NewYork,1984. health of UK military personnel who deployed to the 2003 Iraq
40 Solomon Z, Mikulincer M, Benbenishty R. Locus of control and war:acohortstudy.Lancet2006;367:1731–1741.
combat-relatedpost-traumaticstressdisorder:theinterveningrole 51 Engelhard IM, van den Hout MA, Weerts J, Arntz A, Hox JJ,
ofbattleintensity,threatappraisalandcoping.BrJClinPsychol McNally RJ. Deployment-related stress and trauma in Dutch
1989;28:131–144. soldiers returning from Iraq. Prospective study. Br J Psychiatry
41 King DW, King LA, Gudanowski DM, Vreven DL. Alternative 2007;191:140–145.
representations of war zone stressors: relationships to posttrau- 52 Sundin J, Fear NT, Iversen A, Rona RJ, Wessely S. PTSD after
matic stress disorder in male and female Vietnam veterans. deploymenttoIraq:conflictingrates,conflictingclaims. Psychol
JAbnormPsychol1995;104:184–195. Med2010;40:367–382.
42 Henckens MJAG, van Wingen GA, Joe¨ls M, Ferna´ndez G. Time-
dependenteffectsofcorticosteroids onhumanamygdalaproces-
sing.JNeurosci2010;30:12725–12732. This work is licensed under the Creative
43 Yehuda R, LeDoux J. Response variation following trauma: a Commons Attribution-NonCommercial-
translational neuroscience approach to understanding PTSD. NoDerivativeWorks3.0UnportedLicense.Toview
Neuron2007;56:19–32.
a copy of this license, visit http://creativecommons.
44 vanZuidenM,GeuzeE,WillemenHLDM,VermettenE,MaasM,
HeijnenCJetal.Pre-existinghighglucocorticoidreceptornumber org/licenses/by-nc-nd/3.0/
MolecularPsychiatry