Early-life stress is associated with impairment in cognitive control in adolescence: An fMRI study

Page 1

Ea

SDa

Nb

c

d

a

ARRAA

KSDITMC

1

rt(WRSba

tS

0d

Neuropsychologia 48 (2010) 3037–3044

Contents lists available at ScienceDirect

Neuropsychologia

journa l homepage: www.e lsev ier .com/ locate /neuropsychologia

arly-life stress is associated with impairment in cognitive control indolescence: An fMRI study�

ven C. Muellera,∗, Francoise S. Maheua,b, Mary Dozierc, Elizabeth Pelosoc,arcy Mandell a, Ellen Leibenluftd, Daniel S. Pinea, Monique Ernsta

Section of Developmental and Affective Neuroscience, Mood and Anxiety Disorders Program, National Institute of Mental Health,ational Institutes of Health, 15K North Drive, Bethesda, MD 20892, USAResearch Centre of the CHU Ste-Justine, and Psychiatry Department, University of Montreal, CanadaDepartment of Psychology, University of Delaware, Newark, DE, USASection of Bipolar Spectrum Disorders, Mood and Anxiety Disorders Program, National Institute of Mental Health, National Institutes of Health, Bethesda, MD, USA

r t i c l e i n f o

rticle history:eceived 30 November 2009eceived in revised form 18 May 2010ccepted 10 June 2010vailable online 16 June 2010

eywords:tressevelopment

nhibitionask switching

a b s t r a c t

Early-life stress (ES) has been associated with diverse forms of psychopathology. Some investigatorssuggest that these associations reflect the effects of stress on the neural circuits that support cognitivecontrol. However, very few prior studies have examined the associations between ES, cognitive control,and underlying neural architecture. The present study compares adolescents with a documented historyof ES to typical adolescents on a cognitive control task using functional magnetic resonance imaging(fMRI). Twelve ES adolescents who were adopted because of early caregiver deprivation (9 females,age = 13 years ± 2.58) and 21 healthy control adolescents without a history of ES (10 females, age = 13years ± 1.96) who resided with their biological parents performed the change task (Nelson, Vinton et al.,2007) – a variant of the stop task – during fMRI. Behaviourally, ES adolescents took longer to switch from aprepotent response (“go”) to an alternative response (“change”) than control adolescents. During correct

altreatmenthange task

“change” responses vs. correct “go” responses, this behavioural group difference was accompanied byhigher activation in ES subjects than controls. These differences were noted in regions involved in primarysensorimotor processes (pre- and postcentral gyri), conflict monitoring (dorsal anterior cingulate gyrus),inhibitory and response control (inferior prefrontal cortex and striatum), and somatic representations(posterior insula). Furthermore, correct “change” responses vs. incorrect “change” responses recruitedthe inferior prefrontal cortex (BA 44/46) more strongly in ES subjects than controls. These data suggest

l in y

impaired cognitive contro

. Introduction

Early-life stress (ES) influences brain development and confersisk for later psychopathology, including anxiety, depression, post-raumatic stress disorder (PTSD), substance abuse, and psychosisBremner, Southwick, Johnson, Yehuda, & Charney, 1993; Carrion,

eems, et al., 2009; Fisher et al., 2009; Kilpatrick et al., 2003;

itchie et al., 2009; Schenkel, Spaulding, DiLillo, & Silverstein, 2005;tein et al., 1996). These psychopathological consequences mighte mediated by the disruption of cognitive processes and theirssociated neural underpinnings (Bremner & Vermetten, 2001). In

� The views expressed in this article do not necessarily represent the views ofhe National Institute of Mental Health, National Institutes of Health, or the Unitedtates Government.∗ Corresponding author. Tel.: +1 301 402 6955; fax: +1 301 402 2010.

E-mail address: [email protected] (S.C. Mueller).

028-3932/$ – see front matter. Published by Elsevier Ltd.oi:10.1016/j.neuropsychologia.2010.06.013

outh who experienced ES.Published by Elsevier Ltd.

humans, the impact of ES on cognitive functions such as mem-ory, cognitive control, visuospatial processing, language, attentionprocessing, and manual dexterity has been documented (Chuganiet al., 2001; McEwen, 1998; Nelson, Zeanah et al., 2007; Pears &Fisher, 2005). Concomitantly, some initial studies found an asso-ciation between ES and neural perturbations in affective (Maheuet al., 2010), reward (Dillon et al., 2009), memory (Carrion, Haas,Garrett, Song, & Reiss, 2010), and executive (Carrion, Garrett,Menon, Weems, & Reiss, 2008) processing.

Notably, substantial improvement in cognitive control occursacross development (Davidson, Amso, Anderson, & Diamond, 2006;Rubia et al., 2006), with mature skills finally emerging duringearly adulthood (Bunge & Wright, 2007). These final cognitive

refinements are accompanied by changes in neural activation,particularly in structures implicated in executive control (Bunge,Dudukovic, Thomason, Vaidya, & Gabrieli, 2002; Luna et al., 2001)and goal-directed behaviour (Ernst & Mueller, 2008; Ernst, Pine,& Hardin, 2006). Specifically, these structures include the infe-

Page 2

3 sychol

rPc&c2Thc(flapda&(wasf&rhta

minetaecff

bhtrdhftrpEewacSp

2

2

awtctts

038 S.C. Mueller et al. / Neurop

ior prefrontal cortex for inhibitory processes (Aron, Robbins, &oldrack, 2004), the dorsal anterior cingulate cortex (dACC) foronflict detection/resolution (Botvinick, Nystrom, Fissell, Carter,

Cohen, 1999), and the striatal network involved in responseoding and switching (Atallah, Lopez-Paniagua, Rudy, & O’Reilly,007; Casey et al., 2004; Hikosaka & Isoda, 2010; Loose, Kaufmann,ucha, Auer, & Lange, 2006). Developmental neuroimaging studiesave shown that these same regions are recruited in adoles-ents during tasks that involve inhibition of a prepotent responseDurston, Mulder, Casey, Ziermans, & van Engeland, 2006), con-ict between two responses (e.g., stop-signal task) (Leibenluft etl., 2007; Pliszka et al., 2006), task switching (Casey et al., 2004), orlanning/execution of action (Bunge & Wright, 2007). Unlike tra-itional studies of task switching that require a switch betweenbstract categories such as letters and numbers (Mueller, Swainson,Jackson, 2007; Rogers & Monsell, 1995), the current change task

Nelson, Vinton et al., 2007) is similar to the antisaccade task, inhich a prepotent response must be inhibited and replaced with

n alternative one (Hallett & Adams, 1980). Electrophysiologicaltudies identified distinct inhibitory components when switchingrom prepotent to non-prepotent responses (Mueller, Swainson,

Jackson, 2009). Consequently, these types of tasks inherentlyequire a strong motoric response. Not surprisingly, activationsave also been reported in primary motor and sensorimotor cor-ices in adults (Menon, Adleman, White, Glover, & Reiss, 2001) anddolescents (Schulz et al., 2004) completing go/no-go tasks.

Separation from the primary caregiver, as well as experiencingaltreatment or neglect, are all associated with severe physiolog-

cal stress responses. Such experiences have been shown to altereuroendocrine function (McEwen, 1998; Sanchez et al., 2010). Forxample, one study found that children in foster care exhibit dis-urbed daytime patterns of cortisol production (Dozier, Manni, etl., 2006). Moreover, work with monkeys has found that excessivexposure to cortisol can impair inhibitory control and prefrontalortical function (Lyons, Lopez, Yang, & Schatzberg, 2000). There-ore, ES may have important influences on prefrontal corticalunction via the neuroendocrine system.

Systems in flux, like the adolescent brain, are typically unsta-le and vulnerable to disruption. The current study addressed theypothesis that adolescents with a history of ES would display per-urbed performance on a cognitive control task. We predicted that,elative to a group of healthy peers, ES adolescents would showeficits in cognitive control on the change task. Specifically, weypothesized that reaction time [RT] on trials requiring a switch

rom a prepotent to a non-prepotent response would be slower inhe ES group than the control group. With respect to neural cor-elates, we expected altered regional activation within the inferiorrefrontal cortex, dorsal anterior cingulate cortex and striatum inS adolescents compared to controls (Bunge et al., 2002; Carriont al., 2008; Casey et al., 1997; Durston et al., 2006). Consistentith the motoric requirements of this task (Menon et al., 2001)

nd the differences in sensorimotor processing reported betweenontrols and youth displaying psychopathology (Schulz et al., 2004;uskauer et al., 2008), we also predicted functional alterations inrimary motor and sensorimotor cortices.

. Materials and methods

.1. Subjects

Twelve ES adolescents (13.16 years SD 2.58, 9 female) who resided with theirdoptive parents and 21 non-ES adolescents (13.86 years SD 1.96, 10 female)

ho resided with their biological parents participated in the study. We chose

o recruit ∼50% more controls than ES subjects in order to increase statisti-al power, reduce inter-subject variance, and obtain a truer representation ofhe mean for typical adolescents, given that individual developmental trajec-ories can differ greatly in adolescent samples. Groups did not differ on IQ,ocioeconomic status (SES) (Hollingshead, 1975), sex distribution, parental edu-

ogia 48 (2010) 3037–3044

cation, or State/Trait anxiety scores (Spielberger, Gorsuch, & Lushene, 1970) (allp > .05) (Table 1). All subjects were carefully assessed by physical examinationand structured psychiatric interviews (Kiddie-Schedule-for-Affective-Disorders –Present-and-Lifetime-version (KSADS); Kaufman et al., 1997). Parents (adoptiveparents for the ES group, biological parents for the control group) were also inter-viewed about the adolescent participant’s behaviours. The psychiatric interviewswere conducted by experienced clinicians and achieved excellent inter-rater relia-bility (k > .75). IQ was assessed using the Wechsler Abbreviated Scale of Intelligencefor Adolescents (Wechsler, 1999).

Nine of the twelve ES participants had a history of neglect and maltreatmentprior to their adoption. The other three ES participants had experienced unstableearly environments as reflected by two to three foster placements before adop-tion. ES adolescents had resided in U.S. foster care or international orphanages foran average of 28.36 months (SD = 30.22) prior to their adoptive placement. All ESyouths experienced their first placement between the ages of 1 and 74 months(mean = 16.25 months, SD 25.87 months); they had been placed in foster carebetween one and five times (mean 2.42, SD 1.16). ES adolescents were recruitedas part of a larger, ongoing study and have been well characterized in terms of theirearly-life experiences (Dozier, Peloso, et al., 2006; Dozier, Peloso, Lewis, Laurenceau,& Levine, 2008). In the ES group, one adolescent suffered from enuresis, one fromOppositional Defiant Disorder, one from Generalised Anxiety Disorder, and two fromSpecific Phobia. None of these adolescents met criteria for a mood or anxiety disor-der, including PTSD. Only one subject from the ES group was receiving medication(sertraline and methylphenidate). This subject discontinued methylphenidate for30 h prior to the fMRI study. With regards to prenatal risk in the ES group, one par-ticipant was born with a low birth weight and the mothers of four participants hadsuspected use of alcohol and/or drugs/cigarettes during pregnancy.

Adolescents from the control group were screened for an absence of medicalor psychiatric problems, IQ greater than 70, and no history of maltreatment (asevaluated by the trauma section of the KSADS). The parents of participants pro-vided written informed consent, and the adolescents provided written assent toparticipate in protocols approved by the Institutional Review Boards of the NationalInstitute of Mental Health (NIMH) and the University of Delaware.

2.2. Experimental paradigm

Cognitive control was probed using a variant of the stop-signal task, the changetask (McClure et al., 2005; Nelson, Vinton et al., 2007). The change task measures theability to inhibit a prepotent response (go response) and to switch to an alternativenon-prepotent response (change response). To establish prepotency, go trials consti-tuted 66% of the total experimental trials (excluding blank trials), and change trials33%. Each trial began with a 500 ms fixation cross at the center of the screen. Fixationwas then replaced with an “X” or an “O”, which was displayed for 1000 ms. Subjectswere instructed to press button 1 when they saw an “X”, and button 2 when theysaw an “O”. However, for 33% of the trials, a blue square appeared in the backgroundafter the appearance of the “X” or the “O”. On these trials, subjects were requiredto press button 3. Trials without the appearance of the blue square constituted thego trials, and those with the appearance of the blue square constituted the changetrials.

On the first change trial, the change signal was displayed 250 ms after the go sig-nal. Consistent with the Stop/Signal paradigm (Logan, Schachar, & Tannock, 1997),the delay between the onset of the go signal (“X” or “O”) and the onset of the changesignal (blue square) varied from trial to trial and was adjusted to the subject’s per-formance in order to ensure an approximately 50% correct response rate on changetrials for all participants. If the previous change trial was performed correctly, theonset of the subsequent change signal was delayed by 50 ms, thereby making exe-cution of a correct change response more difficult. If, on the other hand, the previouschange trial was performed incorrectly, the onset of the change signal occurred 50 mscloser to the onset of the X/O signal, making execution of a correct change responseeasier. In addition, 88 blank fixation trials were randomly distributed across thetask and served as the implicit fMRI baseline. The task was conducted over four3.5-min runs inside the scanner. Subjects were trained to proficiency prior to enter-ing the scanner, i.e., mean reaction time of less than 1000 ms on go trials. Subjectsalso received feedback after each block, prompting them to speed up if the mean goreaction time exceeded 1000 ms.

2.3. fMRI data acquisition

All fMRI data were collected at the NIMH. Scanning took place in a General Elec-tric 3 Tesla magnet scanner (Waukesha, WI, USA). Images were acquired with echo-planar single shot gradient echo T2* weighting. A total of 23 5-mm axial slices wereacquired with the following parameters: repetition time (TR) 2000 ms; echo time(TE) 40; field of view (FOV) 240 mm; matrix 64 × 64. A single high-resolution struc-tural image was also acquired for each subject to assist with spatial normalization.

This image consisted of a spoiled gradient (SPGR) sequence: 180 1-mm axial slices;TR 11.4; TE 4.4; FOV 240 mm; number of acquisitions (NEX) 1; matrix 256 × 256;inversion time (TI) 300; bandwidth 130 Hz, pixel, 33 kHz/256 pixels. The paradigmwas presented to the subjects inside the magnet via Avotec Silent Vision goggles(Stuart, FL, USA) mounted on the head coil directly above subjects’ eyes. Each of thefour runs consisted of 86 trials: 44 go trials, 20 change trials, and 22 blank fixation tri-

Page 3

S.C. Mueller et al. / Neuropsychologia 48 (2010) 3037–3044 3039

Table 1Demographic information: demographic information for ES and control groups including mean age, IQ, social economic status, use of medication, and any co-morbid disorders.

Demographic information Adopted youth (n = 12) Unaffected controls (n = 21) P-Value

Age 13.16 (2.58) 13.85 (1.96) .39IQ 105.67 (10.47) 109.62 (12.05) .35SES income 7.25 (1.96) 6.82 (1.67) .53SES education 5.83 (1.19) 5.72 (.89) .77Female 9 10 .13BMI (Mean, SD) 21.16 (4.21) 20.91 (2.35)a .85State anxiety (STAI) 28.08 (1.24) 26.41 (1.05)a .32Trait anxiety (STAI) 29.58 (1.47) 28.67 (1.20)a .62Medication 1 –Anxiety disorder 3 –ODD 1 –Enuresis 1

Behavioural performance dataMean Go RT (ms) 920.78 (39.58) 914.37 (29.62) .90Accuracy go correct (%) 91.28 (2.53) 89.20 (1.89) .52CSRT (ms) 289.03 (17.84) 231.09 (13.35) .02Mean inhibit delay (ms) 434.31 (40.18) 490.70 (30.07) .28

The bold value indicates that significance (at p < .05) was met.y volu

s se cono d on ti

arddZ1

2

mf

FwER

a Unfortunately, due to some missing scores, the comparison of BMI in the healthubset of 17 controls for whom these data were available. The remaining data for then the change task with regards to reaction time to go stimuli (ms), errors committenhibit delay time (MID) of the change command.

ls. Go trials, change trials, and fixation trials were distributed randomly across eachun. In this event-related design, randomly interspersed fixation trials (2250 ms inuration each) allowed the experimental and statistical techniques to successfullyeconvolve unique events occurring relatively close in time (Friston et al., 1998;arahn & Slifstein, 2001). In this statistical approach (Zarahn, Aguirre, & D’Esposito,997), fixation trials were not modelled and provided an implicit baseline.

.4. fMRI analysis

Pre-processing steps of imaging data included slice timing correction, realign-ent, spatial normalization, and reslicing. To facilitate comparison with studies

rom our laboratory that used the same task in different populations (Leibenluft et

ig. 1. (A) Axial slices through the brain showing increased activation in the ES group (relatas set at p < .001, uncorrected. IFG = inferior frontal gyrus, PreCG = precingulate gyrus, P

xtracted peak voxel BOLD signal changes for correct inhibition vs. fixation and correct gOI, given that all clusters showed the same pattern of activation. Data for the ES group ar

nteers is based on a subset of 14 controls and the STAI state/trait comparison on atrols were missing. Behavioural performance data show the performance of subjectshese trials, mean reaction time to execute the change command (CSRT), and mean

al., 2007; Nelson, Vinton et al., 2007), we analysed the data with SPM99 (WellcomeDepartment of Imaging Neuroscience, UCL, London, UK) using the same parametersas in earlier studies. A statistical model was generated for each subject for correctchange trials, incorrect change trials, correct go trials, and incorrect go trials. Withregards to the number of trials used for modelling data, and in line with the groupaccuracy rates provided in Table 1, the statistical analyses included 160 (±3) go trials

and 88 (±3) fixation trials on average for the ES group, and 156 (±3) go trials and 88(±3) fixation trials on average for the control group. With regards to successful vs.unsuccessful change trials, 40 (±5) correct and 40 (±5) incorrect change trials wereused on average for the ES group, and 38 (±4) correct and 42 (±4) incorrect changetrials on average for controls. The wave form used to model each event consisted of arectangular pulse convolved with the hemodynamic response function provided by

ive to controls) during the correct change vs. correct go trials contrast. The thresholdostCG = post-cingulate gyrus, CingG = cingulate gyrus, Cl/ins = Claustrum/insula. (B)o trials vs. fixation as a comparison contrast are shown for the most representativee displayed with dashed bars and the control group in grey. Error bars denote S.E.M.

Page 4

3040 S.C. Mueller et al. / Neuropsychologia 48 (2010) 3037–3044

F (relatB correR was sC

Sea1jl

uteai

srdcr

TSc

ig. 2. (A) Axial and coronal slices (left) showing increased activation in the ES groupOLD signal changes from baseline for the peak activations for correct changes and inOI, given that all clusters showed the same pattern of activation. The thresholdl/insula = Claustrum/insula.

PM99. A high-pass filter of .024 Hz was applied to each subject’s data prior to modelstimation to improve the signal-to-noise ratio. Contrast images were proportion-lly scaled to session baseline and smoothed with an isotropic Gaussian kernel of1.4-mm full width-half maximum (FWHM) (Leibenluft et al., 2007). Individual sub-

ects’ contrasts were then entered into a second order (i.e., random effects) groupevel analysis.

All group level analyses employed t-tests of the contrast values that each individ-al subject’s data generated in the pre-determined event contrasts. The statisticalhreshold was set at p < .001 uncorrected for multiple comparisons with a spatialxtent of at least 10 contiguous voxels (Nelson, Vinton et al., 2007). Sex was used ascovariate of nuisance in all analyses because of a trend towards group differences

n the distribution of this variable (x2(1) = 2.34, p = .13).

To examine the neural correlates involved in inhibiting a planned response and

witching to an alternative response, we compared correct change responses to cor-ect go responses. Moreover, to examine the neural correlates of brain responsesuring incorrect changes, we examined correct vs. incorrect change responses. Toonfirm that significant activations on the correct change vs. correct go contrastesulted from the change responses and not from mere responding to go trials,

able 2ignificantly (p < .001, uncorrected) activated regions on correct change trials (correct chahanges) along with the direction of the effect (ES > control or control > ES), coordinates, B

Region Hemisphere Co

Correct change vs. correct goES > Control

Precentral gyrus Right 46Postcentral gyrus Right 22Cingulate gyrus Right 22Cingulate gyrus Left −2Inferior frontal gyrus/precentral gyrus Left −5Caudate Left −8Putamen/lentiform nucleus Left −2Claustrum/insula Left −3

Correct change vs. incorrect changeES > Control

Precentral gyrus Left −5Inferior frontal gyrus Left −4Claustrum/insula Left −2

ive to controls) during the correct vs. incorrect changes contrast. (B) Bar graphs showct changes relative to fixation for comparison are shown for the most representativeet at p < .001, uncorrected. IFG = inferior frontal gyrus, PreCG = precingulate gyrus,

contrasts were decomposed into correct change trials vs. baseline and correct gotrials vs. baseline (Fig. 1B). Similarly, to tease out whether the activity in the cor-rect vs. incorrect change trials stemmed from either correct or incorrect responding,each component was compared against baseline (Fig. 2B). All activations are in MNIcoordinates (mm).

Finally, to control for a potential confounding effect of group differences inresponse time in the [correct change vs. correct go] contrast, a multivariate ANCOVAwas conducted on extracted individual peak voxel values with Group as thebetween-subjects factor and individual change signal reaction time (CSRT) as thecovariate of nuisance. Statistical threshold was set at p < .05, corrected using thestep-down Bonferroni–Holm procedure.

2.5. Behavioural analysis

Behavioural data were analysed using Multivariate Analysis of Variance(MANOVA). In addition to go and change accuracy, we examined the mean inhibitdelay (MID) and the CSRT. The MID is an index of task difficulty and is calculated bysubtracting the mean onset time of the change signal from the mean onset time of

nge vs. correct go trials) and erroneous change trials (correct changes vs. incorrectroadmann area, T-value, and cluster size.

ordinates BA t-Score Cluster size

−8 42 BA4 4.08 154−28 46 BA3 4.06 1496 40 BA24 3.59 260 18 32 BA32 3.66 650 18 8 BA44/45 3.53 6310 16 3.60 230 18 6 3.75 1440 −22 10 BA13 3.63 46

8 4 8 BA44 4.38 5960 34 6 BA46 3.44 106 26 −2 3.68 64

Page 5

sychol

tWgctsf

3

3

tcr(pctppcvtfso

3

3

sEaitraimrxuafxtw

tmsTvenF

3

arlc

S.C. Mueller et al. / Neurop

he go signal. The CSRT is an estimation of the speed of the subject’s change response.hen a subject changed successfully on 50% of change trials, the CSRT was the mean

o reaction time minus the MID (Logan et al., 1997). Because a subject’s accuracyould deviate from 50%, an interpolation algorithm was used to calculate CSRT, suchhat CSRT = go reaction time at the Xth percentile minus mean MID, where X is theubject’s percent accuracy on change trials (Nelson, Vinton et al., 2007). As with theMRI data, sex was used as a covariate in all behavioural analyses.

. Results

.1. Behavioural performance (Table 1)

ES youths switched to an alternative response on changerials (CSRT: 289.03 ms ± 17.84) significantly more slowly thanontrols (CSRT: 231.09 ms ± 13.35) (F(1,30) = 6.54, p = .02). Accu-acy for go trials did not differ significantly between the ES91.282 ± 2.53%) and control groups (89.20 ± 1.89%) (F(1,30) = .42,= .52). Similarly, the mean RT of go trials (ES group: 920 ± 39 ms;ontrols: 914 ± 29 ms) and the MID time (ES: 434 ± 40 ms; con-rol: 490 ± 30 ms) did not differ between groups (F(1,30) = .16,= .89 and F(1,30) = 1.22, p = .28, respectively). Finally, because thearadigm was designed to keep error rates around 50%, ES andontrol groups did not differ on errors on change trials (49.95%s. 46.88%) (F(1,30) = .42, p = .52) (Table 1). With respect to sex,he error rate for change trials was lower for males (41%) thanemales (53%) (F(1,30) = 8.26, p < .05). CSRT did not correlate withtate (r2(29) = .16, p = .38) or trait (r2(29) = −.05, p = .83) measuresf anxiety.

.2. fMRI

.2.1. Correct change trials vs. correct go trialsThe SPM analysis of correct change (vs. correct go) trials revealed

ignificantly higher activation (all at p ≤ .001, uncorrected) for theS group relative to controls in a set of regions involved in motornd cognitive control (see Table 2). Specifically, these regionsncluded: the left inferior prefrontal cortex (BA44, xyz = −50 18 8),ypically involved in response inhibition and task switching; theight dorsal anterior cingulate cortex (dACC, BA24, xyz = 22 6 40)nd left anterior cingulate cortex (BA32, xyz = −20 18 32), bothnvolved in conflict monitoring; striatal structures (left puta-

en, xyz = −20 18 6 and left caudate, xyz = −8 10 16), involved inesponse planning and execution; and the primary motor (BA4,yz = 46 −8 42) and sensorimotor cortices (BA3, xyz = 22 −28 46)nderlying the motor and sensorimotor aspects of the task. Inddition to expected changes seen in these regions, a group dif-erence emerged in the posterior left claustrum/insula (BA13,yz = −30 −22 10) (Table 2), also reflecting enhanced activation inhe ES group relative to controls (Fig. 1A). No regional activationsere found to be greater in the controls than in the ES group.

To examine whether the significantly increased activations forhe ES group reflected the switch process (change trials) or were a

ere result of motor response (go trials), the contrasts were furthereparated into change trials vs. fixation, and go trials vs. fixation.he results revealed that the contributions to the significant acti-ation stemmed from the change trials (all F(1,29) > 6.00, p < .02,xcept for the cingulate activation BA32 which approached sig-ificance (F(1,29) = 3.93, p = .057)), and not from the go trials (all(1,29) < 1.26, all p > .21) (Fig. 1B).

.3. Correct change vs. incorrect change trials

In contrast to correct change trials vs. incorrect change tri-ls, the ES group showed significantly greater activation in threeegions (all at p < .001, uncorrected) compared to controls: theeft inferior prefrontal gyrus (BA46, xyz = -40 34 6), the left pre-entral/inferior frontal gyrus (BA44, xyz = -58 4 8), and the left

ogia 48 (2010) 3037–3044 3041

claustrum/insula (xyz = −26 26 −2) (Table 2/Fig. 2A). Further anal-ysis of this contrast revealed that the significant trial by groupinteraction was primarily related to group differences in correctchange trials (all F(1,29) > 5.50, all p < .03), and not incorrect changetrials (all F(1,29) < .82, ns). However, the deactivation of the inferiorfrontal gyrus seen in the ES group during these incorrect trials mayalso have contributed to the interaction (Fig. 2B).

3.4. Additional analyses

A multivariate ANCOVA controlling for individual CSRT (covari-ate of nuisance) in the go vs. change contrast revealed that groupdifferences in activations in the ACC (BA24), postcentral gyrus andIFG remained significant (at p < .05, corrected), while the other areaswere trending (p < .10, corrected). These results suggest that theobserved group differences were not accounted for by group dif-ferences in motoric speed.

4. Discussion

The present study addressed the extent to which ES is associ-ated with perturbations in cognitive control, which, in turn, couldconfer risk for psychopathology in this population. Our findingssupport this hypothesis at both the behavioural and neural level. Inthis study, and consistent with previous reports of perturbations oncognitive control tasks in adopted youth (Lewis, Dozier, Ackerman,& Sepulveda-Kozakowski, 2007), the ES group showed prolongedreaction times compared to controls when inhibiting a prepotentresponse and executing an alternative one. Furthermore, group dif-ferences in neural activity showed greater activation in the ES groupthan in controls in several regions involved in cognitive control.These included the inferior frontal cortex (BA junction 44/45 and46), a region previously identified as playing a critical role in cogni-tive inhibitory control (Aron et al., 2004; Bunge et al., 2002; Durstonet al., 2006; Leung & Cai, 2007), and the striatum, a region involvedin response control (Atallah et al., 2007; Casey et al., 2004; Loose etal., 2006).

The current study found these regions to be more active inthe ES group than in controls during correct change trials rela-tive to go trials (BA44/45, striatum) as well as during correct vs.incorrect change responses (BA44/46). Comparisons against fixa-tion confirmed that the greater activation seen in the ES group vs.controls occurred during the change trials and not during the go tri-als, suggesting an impaired behavioural control aspect of the taskrather than impaired simple motor response. In addition, IFG, dACCand postcentral gyrus findings remained robust, while the remain-ing regions were trending, when controlling for CSRT, suggestingthat differences in neural activation were not due to the slowerexecution of the alternative response in the ES group.

Studies in psychiatric samples indicate that psychopathologymay disrupt the same prefrontal-striatal circuitry whose functiondiffered between our ES sample and controls (Carrion et al., 2008;Durston et al., 2006; Leibenluft et al., 2007; Nelson, Vinton et al.,2007; Schulz et al., 2004). Indeed, the neural pattern of our datais consistent with neuroimaging findings of disorders for which ESpopulations show enhanced risk. Findings in adults, for instance,have associated depression with distinct patterns of activation inthe inferior frontal and anterior cingulate regions during inhibitorycontrol tasks (Langenecker et al., 2007). Likewise, fronto-striataland somatosensory network alterations during cognitive inhibition

have been reported in individuals with PTSD (Falconer et al., 2008).In youths, prior studies of ES have focused primarily on morpho-logical differences (De Bellis & Kuchibhatla, 2006; Eluvathingal etal., 2006; Mehta et al., 2009; Teicher et al., 2004); nevertheless,preliminary investigations using functional imaging have yielded

Page 6

3 sychol

ilssri

a(tsfiaa(o(ibalp

sii(irbw(rpri

iiHwt2Pisu(pvi&dept

agpi&Ddb

042 S.C. Mueller et al. / Neurop

nteresting patterns. Maheu et al. (2010) reported medial temporalobe dysfunction during emotional face processing in ES subjects,ome of whom were included in the current sample. In anotherample of maltreated youth, Dillon et al. (2009) found decreasedesponding of the basal ganglia to reward stimuli. The current find-ngs extend prior reports to a clinical non-PTSD sample.

Notably, current IFG activations were left-lateralized. Studies indults commonly associate the right IFG with inhibitory processese.g., Aron et al., 2004). Three possible explanations can account forhis discrepancy. First, a recent meta-analysis of inhibitory controltudies that included a set of patients with selective left inferiorrontal gyrus lesions reported a critical involvement of this regionn inhibitory control (Swick, Ashley, & Turken, 2008), suggesting

bilateral contribution of the IFG to inhibitory control. Second,ctivation in the left IFG could also account for the shift processpressing button 3 during the change signal instead of buttons 1r 2), in addition to the inhibitory process. Indeed, Loose et al.2006) reported left-sided IFG activations during a response switch-ng task. Finally, significantly activated clusters in left IFG have alsoeen reported in children during successful resolution of responsembiguity (Bunge et al., 2002). More research in pediatric popu-ations is needed to disentangle the contribution of each of theserocesses to the current findings.

Similar to the group differences in inferior frontal gyrus andtriatum activation, the ES group showed heightened activationsn motor and sensorimotor cortices than controls. Prior studiesn youths with ADHD (Schulz et al., 2004) and healthy volunteersMenon et al., 2001) have also documented activations in BAs 3/4n go/no-go studies. Given that fine motor skills of the fingers areequired to flexibly select and press the correct out of 3 possibleuttons on the current task, our findings are further consistentith a report of impaired manual dexterity in neglected youth

Chugani et al., 2001). However, the absence of group differences ineaction times to go stimuli in the present study complicate inter-retation. Regardless, heightened activation of these regions couldeflect developmental disruptions of primary sensorimotor areasn ES.

Likewise, the anterior cingulate cortex has also been implicatedn response conflict but not in stimulus conflict during task switch-ng (Liston, Matalon, Hare, Davidson, & Casey, 2006; Rushworth,adland, Paus, & Sipila, 2002). In this study, increased activationas present in two regions of the dACC (BA24 and BA32). Prior func-

ional studies of youths with cognitive control deficits (Pliszka et al.,006; Rubia et al., 1999; Tamm, Menon, Ringel, & Reiss, 2004) andTSD after maltreatment (Carrion et al., 2008) documented changesn dACC function during inhibitory control tasks. In particular, in aimilar study in youths with post-traumatic stress symptoms thatsed a block design to compare go and no-go trials, Carrion et al.2008) reported hyperactivations in both BAs 24 and 32. Likewise,revious morphological studies documented smaller grey matterolumes in the dACC of adults who experienced adverse life eventsn childhood (Cohen et al., 2006) or adulthood (Ganzel, Kim, Glover,

Temple, 2008). Consequently, our findings are in-line with priorata linking this region with cognitive control and adverse experi-nces; they also extend previous research and suggest that ES mighterturb a network involved in motor control aspects of executiveasks leading to a difficulty in resolving response conflict.

Finally, an interesting but unexpected finding was the hyper-ctivation of the posterior insula/claustrum observed in the ESroup. Although the posterior insula is commonly associated withain and visceral sensory functions (Augustine, 1996), other stud-

es have documented a role in working memory (Paulesu, Frith,Frackowiak, 1993) and divided attention (Corbetta, Miezin,

obmeyer, Shulman, & Petersen, 1991). The current task requiredivided attention given that participants had to distinguishetween letters as well as pay attention to a sudden change in

ogia 48 (2010) 3037–3044

background colour when a switch in responses was required.Moreover, previous studies have reported perturbed attentionalemotional processing in maltreated youth (Pine et al., 2005), sup-porting the idea of impaired attention processing following ES.Taken together, the present study demonstrates the impact of ES onseveral brain regions including fronto-striatal circuitry, dACC, andinsula/claustrum during cognitive control. However, the small sam-ple precludes us from examining the contribution of the specifictypes of stress experienced. Future studies using animal modelscould further explore the vulnerability of particular brain regionsto specific stressors.

One strength of this sample is that participants were recruitedfrom a very well-characterized and longitudinally followed cohort(Dozier et al., 2008; Dozier, Peloso, et al., 2006). However, a few lim-itations of the study also need to be addressed. One concern is theheterogeneity of the relatively small sample. A common problemin conducting studies in ES youth is recruitment; indeed, previ-ous studies have suffered from similar limitations in their ability toreport differences in the type of maltreatment experienced in theirsample (Carrion et al., 2008; Dillon et al., 2009) or in samples com-ing from different orphanages (Chugani et al., 2001; Eluvathingalet al., 2006). Another problem is the extent to which prenatal fac-tors, such as malnutrition or prenatal exposure to drugs, may havecontributed to the present findings. This issue is very difficult toresolve because of poor historical information regarding drug usein the biological parents. The use of comparison groups of low birthweight or those with known prenatal exposure to drugs mighthelp to clarify these factors. It is important to note, however, thatalthough youths may have experienced different types of stres-sors, one commonality is that all stressors are likely to trigger thesame determinants of the stress response such as stress-relatedhormones (e.g., cortisol) and neurotransmitters (e.g., serotonin). Afinal issue concerns the fact that four adopted participants sufferedfrom a psychiatric disorder. However, when these four subjectswere removed from the analysis (data not shown), the pattern ofresults was unchanged, suggesting that psychiatric morbidity hadlittle impact on the present data.

In summary, the present study extends previous behaviouralevidence and demonstrates perturbations in the neural circuitscritical for cognitive and motor control in youths with ES.

ES, here, included early separation from the biological par-ent, several separations from caregivers before adoption, and earlyneglect and abuse. Future studies should replicate and refine thesefindings. In particular, critical periods of development, the typeof maltreatment experienced, and other prenatal factors such asexposure to drugs will need to be investigated.

Conflict of interest

Dr. Pine has received compensation for activities related toteaching, editing, and clinical care that pose no conflict. All otherauthors declare that, except for income received from their primaryemployer, no financial support or compensation has been receivedfrom any individual or corporate entity over the past 3 years forresearch or professional service and there are no personal financialholdings that could be perceived as constituting a potential conflictof interest.

Acknowledgements

This study was supported by a National Alliance for Researchon Schizophrenia and Depression (NARSAD) Young Investiga-tor Award and a postdoctoral fellowship from the Fonds de laRecherche en Santé du Québec (FRSQ) to FSM, and by the NIMHIntramural Research Program. Ioline Henter provided invaluable

Page 7

sychol

emi

R

A

A

A

B

B

B

B

B

C

C

C

C

C

C

C

C

D

D

D

D

D

D

D

E

E

S.C. Mueller et al. / Neurop

ditorial assistance. We would also like to thank the three anony-ous reviewers for their very helpful and constructive comments

n improving earlier versions of this manuscript.

eferences

ron, A. R., Robbins, T. W., & Poldrack, R. A. (2004). Inhibition and the right inferiorfrontal cortex. Trends in Cognitive Sciences, 8(4), 170–177.

tallah, H. E., Lopez-Paniagua, D., Rudy, J. W., & O’Reilly, R. C. (2007). Separate neuralsubstrates for skill learning and performance in the ventral and dorsal striatum.Nature Neuroscience, 10(1), 126–131.

ugustine, J. R. (1996). Circuitry and functional aspects of the insular lobe in primatesincluding humans. Brain Research. Brain Research Reviews, 22(3), 229–244.

otvinick, M., Nystrom, L. E., Fissell, K., Carter, C. S., & Cohen, J. D. (1999). Con-flict monitoring versus selection-for-action in anterior cingulate cortex. Nature,402(6758), 179–181.

remner, J. D., Southwick, S. M., Johnson, D. R., Yehuda, R., & Charney, D. S. (1993).Childhood physical abuse and combat-related posttraumatic stress disorder inVietnam veterans. The American Journal of Psychiatry, 150(2), 235–239.

remner, J. D., & Vermetten, E. (2001). Stress and development: Behav-ioral and biological consequences. Development and Psychopathology, 13,473–489.

unge, S. A., Dudukovic, N. M., Thomason, M. E., Vaidya, C. J., & Gabrieli, J. D. (2002).Immature frontal lobe contributions to cognitive control in children: Evidencefrom fMRI. Neuron, 33(2), 301–311.

unge, S. A., & Wright, S. B. (2007). Neurodevelopmental changes in working memoryand cognitive control. Current Opinion in Neurobiology, 17(2), 243–250.

arrion, V. G., Garrett, A., Menon, V., Weems, C. F., & Reiss, A. L. (2008). Posttraumaticstress symptoms and brain function during a response-inhibition task: An fMRIstudy in youth. Depression and Anxiety, 25(6), 514–526.

arrion, V. G., Haas, B. W., Garrett, A., Song, S., & Reiss, A. L. (2009). Reduced hip-pocampal activity in youth with posttraumatic stress symptoms: An fMRI study.Journal of Pediatric Psychology, 35(5), 559–569.

arrion, V. G., Weems, C. F., Watson, C., Eliez, S., Menon, V., & Reiss, A. L. (2009).Converging evidence for abnormalities of the prefrontal cortex and evaluation ofmidsagittal structures in pediatric posttraumatic stress disorder: An MRI study.Psychiatry Research, 172(3), 226–234.

asey, B. J., Castellanos, F. X., Giedd, J. N., Marsh, W. L., Hamburger, S. D., Schubert, A.B., et al. (1997). Implication of right frontostriatal circuitry in response inhibitionand attention-deficit/hyperactivity disorder. Journal of the American Academy ofChild and Adolescent Psychiatry, 36(3), 374–383.

asey, B. J., Davidson, M. C., Hara, Y., Thomas, K. M., Martinez, A., Galvan, A., et al.(2004). Early development of subcortical regions involved in non-cued attentionswitching. Developmental Science, 7(5), 534–542.

hugani, H. T., Behen, M. E., Muzik, O., Juhasz, C., Nagy, F., & Chugani, D. C. (2001).Local brain functional activity following early deprivation: A study of postinsti-tutionalized Romanian orphans. Neuroimage, 14(6), 1290–1301.

ohen, R. A., Grieve, S., Hoth, K. F., Paul, R. H., Sweet, L., Tate, D., et al. (2006). Earlylife stress and morphometry of the adult anterior cingulate cortex and caudatenuclei. Biological Psychiatry, 59(10), 975–982.

orbetta, M., Miezin, F. M., Dobmeyer, S., Shulman, G. L., & Petersen, S. E. (1991).Selective and divided attention during visual discriminations of shape, color,and speed: Functional anatomy by positron emission tomography. The Journalof Neuroscience, 11(8), 2383–2402.

avidson, M. C., Amso, D., Anderson, L. C., & Diamond, A. (2006). Development ofcognitive control and executive functions from 4 to 13 years: Evidence frommanipulations of memory, inhibition, and task switching. Neuropsychologia,44(11), 2037–2078.

e Bellis, M. D., & Kuchibhatla, M. (2006). Cerebellar volumes in pediatricmaltreatment-related posttraumatic stress disorder. Biological Psychiatry, 60(7),697–703.

illon, D. G., Holmes, A. J., Birk, J. L., Brooks, N., Lyons-Ruth, K., & Pizzagalli, D. A.(2009). Childhood adversity is associated with left basal ganglia dysfunctionduring reward anticipation in adulthood. Biological Psychiatry, 66(3), 206–213.

ozier, M., Manni, M., Gordon, M. K., Peloso, E., Gunnar, M. R., Stovall-McClough, K.C., et al. (2006). Foster children’s diurnal production of cortisol: An exploratorystudy. Child Maltreatment, 11(2), 189–197.

ozier, M., Peloso, E., Lewis, E., Laurenceau, J., & Levine, S. (2008). Effects of anattachment-based intervention on the cortisol production of infants and tod-dlers in foster care. Development and Psychopathology, 20, 845–859.

ozier, M., Peloso, E., Lindhiem, O., Gordon, M. K., Manni, M., Sepulveda, S., etal. (2006). Preliminary evidence from a randomized clinical trial: Interventioneffects on foster children’s behavioral and biological regulation. Journal of SocialIssues, 62, 767–785.

urston, S., Mulder, M., Casey, B. J., Ziermans, T., & van Engeland, H. (2006). Activationin ventral prefrontal cortex is sensitive to genetic vulnerability for attention-deficit hyperactivity disorder. Biological Psychiatry, 60(10), 1062–1070.

luvathingal, T. J., Chugani, H. T., Behen, M. E., Juhasz, C., Muzik, O., Maqbool,M., et al. (2006). Abnormal brain connectivity in children after early severesocioemotional deprivation: A diffusion tensor imaging study. Pediatrics, 117(6),2093–2100.

rnst, M., & Mueller, S. C. (2008). The adolescent brain: Insights from functionalneuroimaging research. Developmental Neurobiology, 68(6), 729–743.

ogia 48 (2010) 3037–3044 3043

Ernst, M., Pine, D. S., & Hardin, M. (2006). Triadic model of the neurobiology ofmotivated behavior in adolescence. Psychological Medicine, 36(3), 299–312.

Falconer, E., Bryant, R., Felmingham, K. L., Kemp, A. H., Gordon, E., Peduto, A., etal. (2008). The neural networks of inhibitory control in posttraumatic stressdisorder. Journal of Psychiatry & Neuroscience, 33(5), 413–422.

Fisher, H., Morgan, C., Dazzan, P., Craig, T. K., Morgan, K., Hutchinson, G., et al. (2009).Gender differences in the association between childhood abuse and psychosis.The British Journal of Psychiatry, 194(4), 319–325.

Friston, K. J., Fletcher, P., Josephs, O., Holmes, A., Rugg, M. D., & Turner, R. (1998).Event-related fMRI: Characterizing differential responses. Neuroimage, 7(1),30–40.

Ganzel, B. L., Kim, P., Glover, G. H., & Temple, E. (2008). Resilience after 9/11: Mul-timodal neuroimaging evidence for stress-related change in the healthy adultbrain. Neuroimage, 40(2), 788–795.

Hallett, P. E., & Adams, B. D. (1980). The predictability of saccadic latency in a novelvoluntary oculomotor task. Vision Research, 20(4), 329–339.

Hikosaka, O., & Isoda, M. (2010). Switching from automatic to controlled behavior:Cortico-basal ganglia mechanisms. Trends in Cognitive Sciences, 14(4), 154–161.

Hollingshead, A. B. (1975). Four-factor index of social status. New Haven: Yale Uni-versity Press.

Kaufman, J., Birmaher, B., Brent, D., Rao, U., Flynn, C., Moreci, P., et al. (1997). Schedulefor affective disorders and schizophrenia for school-age children-Present andLifetime Version (K-SADS-PL): Initial reliability and validity data. Journal of theAmerican Academy of Child and Adolescent Psychiatry, 36(7), 980–988.

Kilpatrick, D. G., Ruggiero, K. J., Acierno, R., Saunders, B. E., Resnick, H. S., &Best, C. L. (2003). Violence and risk of PTSD, major depression, substanceabuse/dependence, and comorbidity: Results from the National Survey of Ado-lescents. Journal of Consulting and Clinical Psychology, 71(4), 692–700.

Langenecker, S. A., Kennedy, S. E., Guidotti, L. M., Briceno, E. M., Own, L. S., Hooven,T., et al. (2007). Frontal and limbic activation during inhibitory control predictstreatment response in major depressive disorder. Biological Psychiatry, 62(11),1272–1280.

Leibenluft, E., Rich, B. A., Vinton, D. T., Nelson, E. E., Fromm, S. J., Berghorst, L. H.,et al. (2007). Neural circuitry engaged during unsuccessful motor inhibition inpediatric bipolar disorder. The American Journal of Psychiatry, 164(1), 52–60.

Leung, H. C., & Cai, W. (2007). Common and differential ventrolateral prefrontalactivity during inhibition of hand and eye movements. The Journal of Neuro-science, 27(37), 9893–9900.

Lewis, E. E., Dozier, M., Ackerman, J., & Sepulveda-Kozakowski, S. (2007). The effectof placement instability on adopted children’s inhibitory control abilities andoppositional behavior. Developmental Psychology, 43(6), 1415–1427.

Liston, C., Matalon, S., Hare, T. A., Davidson, M. C., & Casey, B. J. (2006). Anteriorcingulate and posterior parietal cortices are sensitive to dissociable forms ofconflict in a task-switching paradigm. Neuron, 50(4), 643–653.

Logan, G. D., Schachar, R., & Tannock, R. (1997). Impulsivity and inhibitory control.Psychological Science, 8, 60–64.

Loose, R., Kaufmann, C., Tucha, O., Auer, D. P., & Lange, K. W. (2006). Neural net-works of response shifting: Influence of task speed and stimulus material. BrainResearch, 1090(1), 146–155.

Luna, B., Thulborn, K. R., Munoz, D. P., Merriam, E. P., Garver, K. E., Minshew, N. J., etal. (2001). Maturation of widely distributed brain function subserves cognitivedevelopment. Neuroimage, 13(5), 786–793.

Lyons, D. M., Lopez, J. M., Yang, C., & Schatzberg, A. F. (2000). Stress-level corti-sol treatment impairs inhibitory control of behavior in monkeys. The Journal ofNeuroscience, 20(20), 7816–7821.

Maheu, F., Dozier, M., Mandell, D., Peloso, E., Poeth, K., Jenness, J., et al. (2010).Enhanced amygdala and hippocampal responses to fear faces among Youthwith a History of Early Caregiver Deprivation. Cognitive, Affective, & BehavioralNeuroscience., 10(1), 34–49.

McClure, E. B., Treland, J. E., Snow, J., Schmajuk, M., Dickstein, D. P., Towbin, K. E., etal. (2005). Deficits in social cognition and response flexibility in pediatric bipolardisorder. The American Journal of Psychiatry, 162(9), 1644–1651.

McEwen, B. S. (1998). Protective and damaging effects of stress mediators. The NewEngland Journal of Medicine, 338(3), 171–179.

Mehta, M. A., Golembo, N. I., Nosarti, C., Colvert, E., Mota, A., Williams, S. C., etal. (2009). Amygdala, hippocampal and corpus callosum size following severeearly institutional deprivation: The English and Romanian Adoptees studypilot. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 50(8),943–951.

Menon, V., Adleman, N. E., White, C. D., Glover, G. H., & Reiss, A. L. (2001). Error-related brain activation during a Go/NoGo response inhibition task. Human BrainMapping, 12(3), 131–143.

Mueller, S. C., Swainson, R., & Jackson, G. M. (2007). Behavioural and neurophysi-ological correlates of bivalent and univalent responses during task switching.Brain Research, 1157, 56–65.

Mueller, S. C., Swainson, R., & Jackson, G. M. (2009). ERP indices of persisting andcurrent inhibitory control: A study of saccadic task switching. Neuroimage, 45(1),191–197.

Nelson, C. A., 3rd, Zeanah, C. H., Fox, N. A., Marshall, P. J., Smyke, A. T., & Guthrie, D.(2007). Cognitive recovery in socially deprived young children: The Bucharest

Early Intervention Project. Science, 318(5858), 1937–1940.

Nelson, E. E., Vinton, D. T., Berghorst, L., Towbin, K. E., Hommer, R. E., Dickstein, D. P.,et al. (2007). Brain systems underlying response flexibility in healthy and bipolaradolescents: An event-related fMRI study. Bipolar Disorders, 9(8), 810–819.

Paulesu, E., Frith, C. D., & Frackowiak, R. S. (1993). The neural correlates of the verbalcomponent of working memory. Nature, 362(6418), 342–345.

Page 8

3 sychol

P

P

P

R

R

R

R

R

S

S

TX: Harcourt Assessment.Zarahn, E., Aguirre, G. K., & D’Esposito, M. (1997). Empirical analyses of BOLD fMRI

044 S.C. Mueller et al. / Neurop

ears, K., & Fisher, P. A. (2005). Developmental, cognitive, and neuropsychologicalfunctioning in preschool-aged foster children: Associations with prior maltreat-ment and placement history. Journal of Developmental and Behavioral Pediatrics,26(2), 112–122.

ine, D. S., Mogg, K., Bradley, B. P., Montgomery, L., Monk, C. S., McClure, E., et al.(2005). Attention bias to threat in maltreated children: Implications for vul-nerability to stress-related psychopathology. The American Journal of Psychiatry,162(2), 291–296.

liszka, S. R., Glahn, D. C., Semrud-Clikeman, M., Franklin, C., Perez, R., 3rd, Xiong, J.,et al. (2006). Neuroimaging of inhibitory control areas in children with atten-tion deficit hyperactivity disorder who were treatment naive or in long-termtreatment. The American Journal of Psychiatry, 163(6), 1052–1060.

itchie, K., Jaussent, I., Stewart, R., Dupuy, A. M., Courtet, P., Ancelin, M. L., et al.(2009). Association of adverse childhood environment with late-life depression.The Journal of Clinical Psychiatry, 70(9), 1281–1288.

ogers, R. D., & Monsell, S. (1995). Costs of a predictable switch between simplecognitive tasks. Journal of Experimental Psychology: General, 124(2), 207–231.

ubia, K., Overmeyer, S., Taylor, E., Brammer, M., Williams, S. C., Simmons, A.,et al. (1999). Hypofrontality in attention deficit hyperactivity disorder duringhigher-order motor control: A study with functional MRI. The American Journalof Psychiatry, 156(6), 891–896.

ubia, K., Smith, A. B., Woolley, J., Nosarti, C., Heyman, I., Taylor, E., et al. (2006). Pro-gressive increase of frontostriatal brain activation from childhood to adulthoodduring event-related tasks of cognitive control. Human Brain Mapping, 27(12),973–993.

ushworth, M. F. S., Hadland, K. A., Paus, T., & Sipila, P. K. (2002). Role of the humanmedial frontal cortex in task switching: A combined fMRI and TMS study. Journalof Neurophysiology, 87, 2577–2592.

anchez, M. M., McCormack, K., Grand, A. P., Fulks, R., Graff, A., & Maestripieri, D.

(2010). Effects of sex and early maternal abuse on adrenocorticotropin hor-mone and cortisol responses to the corticotropin-releasing hormone challengeduring the first 3 years of life in group-living rhesus monkeys. Development andPsychopathology, 22(1), 45–53.

chenkel, L. S., Spaulding, W. D., DiLillo, D., & Silverstein, S. M. (2005). Historiesof childhood maltreatment in schizophrenia: Relationships with premorbid

ogia 48 (2010) 3037–3044

functioning, symptomatology, and cognitive deficits. Schizophrenia Research,76(2–3), 273–286.

Schulz, K. P., Fan, J., Tang, C. Y., Newcorn, J. H., Buchsbaum, M. S., Cheung, A.M., et al. (2004). Response inhibition in adolescents diagnosed with attentiondeficit hyperactivity disorder during childhood: An event-related FMRI study.The American Journal of Psychiatry, 161(9), 1650–1657.

Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1970). Manual for the state-traitanxiety inventory. Palo Alto, CA: Consulting Psychologists Press.

Stein, M. B., Walker, J. R., Anderson, G., Hazen, A. L., Ross, C. A., Eldridge, G., et al.(1996). Childhood physical and sexual abuse in patients with anxiety disor-ders and in a community sample. The American Journal of Psychiatry, 153(2),275–277.

Suskauer, S. J., Simmonds, D. J., Fotedar, S., Blankner, J. G., Pekar, J. J., Denckla, M. B., etal. (2008). Functional magnetic resonance imaging evidence for abnormalitiesin response selection in attention deficit hyperactivity disorder: Differences inactivation associated with response inhibition but not habitual motor response.Journal of Cognitive Neuroscience, 20(3), 478–493.

Swick, D., Ashley, V., & Turken, A. U. (2008). Left inferior frontal gyrus is critical forresponse inhibition. BMC Neuroscience, 9, 102.

Tamm, L., Menon, V., Ringel, J., & Reiss, A. L. (2004). Event-related FMRI evidence offrontotemporal involvement in aberrant response inhibition and task switchingin attention-deficit/hyperactivity disorder. Journal of the American Academy ofChild and Adolescent Psychiatry, 43(11), 1430–1440.

Teicher, M. H., Dumont, N. L., Ito, Y., Vaituzis, C., Giedd, J. N., & Andersen, S. L. (2004).Childhood neglect is associated with reduced corpus callosum area. BiologicalPsychiatry, 56(2), 80–85.

Wechsler, D. (1999). Wechsler abbreviated scale of intelligence (WASI). San Antonio,

statistics. I. Spatially unsmoothed data collected under null-hypothesis condi-tions. Neuroimage, 5(3), 179–197.

Zarahn, E., & Slifstein, M. (2001). A reference effect approach for power analysis infMRI. Neuroimage, 14(3), 768–779.