PLEASE SCROLL DOWN FOR ARTICLE This article was downloaded by: [Crone, Eveline A.] On: 17 August 2010 Access details: Access Details: [subscription number 925873663] Publisher Psychology Press Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37- 41 Mortimer Street, London W1T 3JH, UK Social Neuroscience Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t741771143 Do you like me? Neural correlates of social evaluation and developmental trajectories Bregtje Gunther Moor abc ; Linda van Leijenhorst bc ; Serge A. R. B. Rombouts bc ; Eveline A. Crone bc ; Maurits W. Van der Molen a a University of Amsterdam, Amsterdam, The Netherlands b Leiden University, Leiden, The Netherlands c Leiden Institute for Brain and Cognition, Leiden, The Netherlands First published on: 17 August 2010 To cite this Article Moor, Bregtje Gunther , van Leijenhorst, Linda , Rombouts, Serge A. R. B. , Crone, Eveline A. and Van der Molen, Maurits W.(2010) 'Do you like me? Neural correlates of social evaluation and developmental trajectories', Social Neuroscience,, First published on: 17 August 2010 (iFirst) To link to this Article: DOI: 10.1080/17470910903526155 URL: http://dx.doi.org/10.1080/17470910903526155 Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
Transcript
PLEASE SCROLL DOWN FOR ARTICLE
This article was downloaded by: [Crone, Eveline A.]On: 17 August 2010Access details: Access Details: [subscription number 925873663]Publisher Psychology PressInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Social NeurosciencePublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t741771143
Do you like me? Neural correlates of social evaluation and developmentaltrajectoriesBregtje Gunther Moorabc; Linda van Leijenhorstbc; Serge A. R. B. Romboutsbc; Eveline A. Cronebc;Maurits W. Van der Molena
a University of Amsterdam, Amsterdam, The Netherlands b Leiden University, Leiden, The Netherlandsc Leiden Institute for Brain and Cognition, Leiden, The Netherlands
First published on: 17 August 2010
To cite this Article Moor, Bregtje Gunther , van Leijenhorst, Linda , Rombouts, Serge A. R. B. , Crone, Eveline A. and Vander Molen, Maurits W.(2010) 'Do you like me? Neural correlates of social evaluation and developmental trajectories',Social Neuroscience,, First published on: 17 August 2010 (iFirst)To link to this Article: DOI: 10.1080/17470910903526155URL: http://dx.doi.org/10.1080/17470910903526155
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.
PSNS Do you like me? Neural correlates of social evaluation and developmental trajectories
Neural Correlates of Peer Evaluation Bregtje Gunther MoorUniversity of Amsterdam, Amsterdam, Leiden University, Leiden, and Leiden Institute for Brain and
Cognition, Leiden, The Netherlands
Linda van Leijenhorst, Serge A.R.B. Rombouts, and Eveline A. CroneLeiden University, Leiden, and Leiden Institute for Brain and Cognition, Leiden, The Netherlands
Maurits W. Van der MolenUniversity of Amsterdam, Amsterdam, The Netherlands
Social acceptance is of key importance for healthy functioning. We used functional magnetic resonance imaging(fMRI) to examine age-related changes in the neural correlates of social acceptance and rejection processing.Participants from four age groups participated in the study: pre-pubertal children (8–10 years), early adolescents(12–14 years), older adolescents (16–17 years) and young adults (19–25 years). During the experiment, partici-pants were presented with unfamiliar faces of peers and were asked to predict whether they expected to be liked ordisliked by the other person, followed by feedback indicating acceptance or rejection. Results showed that activa-tion in the ventral mPFC and striatum to social feedback was context-dependent; there was increased activationwhen participants had positive expectations about social evaluation, and increased activation following socialacceptance feedback. Age-related comparisons revealed a linear increase in activity with age in these brainregions for positive expectations of social evaluation. Similarly, a linear increase with age was found for activa-tion in the striatum, ventral mPFC, OFC, and lateral PFC for rejection feedback. No age-related differences inneural activation were shown for social acceptance feedback. Together, these results provide important insights inthe developmental trajectories of brain regions implicated in social and affective behavior.
Keywords: Development; Neuroimaging; Medial prefrontal cortex; Social information processing; Striatum.
INTRODUCTION
An important hallmark of the human species is thesignificance of social interactions and relationships.Given the strong evolved motive of humans to formsocial bonds, social belonging and acceptance arepsychologically important events (Baumeister &Leary, 1995). There is a rapidly growing literature onthe use of neuroimaging methods to identify the neu-ral mechanisms underlying human social interac-
tions, with a specific focus on social feedbackprocessing (e.g., Amodio & Frith, 2006; Blakemore,Winston, & Frith, 2004; Rilling, King-Casas, & San-fey, 2008). Notably, research by Somerville, Heath-erton, and Kelley (2006) indicated that the ventralregion of the anterior cingulate cortex (v-ACC) issensitive to feedback information which signals thatan individual is liked by another individual. Socialacceptance processing is implicated in mentalizing,which refers to the ability to understand the intentions
Correspondence should be addressed to: Bregtje Gunther Moor, Department of Psychology, University of Amsterdam, Roetersstraat 15,1018 WB, Amsterdam, The Netherlands. E-mail: [email protected]
This work was supported by the NWO-MaGW Grant No. 400-07-066 from the Dutch Science Foundation. The authors would like toacknowledge Lottie Bullens for assistance in recruitment of participants and Zdena op de Macks for assistance during testing.
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
and mental states of others (Frith & Frith, 1999).Indeed, other fMRI studies using a wide range oftasks have consistently demonstrated involvement ofthe medial prefrontal cortex (mPFC), including the v-ACC, in social and affective information processing(Amodio & Frith, 2006; Gallagher & Frith, 2003).These studies have suggested that the mPFC isinvolved in self-referential processing, or the abilityto form judgments about other people and to makeinferences about how others view us (Sebastian, Bur-nett, & Blakemore, 2008).
However, prior studies have also suggested thatsocial feedback can have different informative valuedepending on prior expectations. For example, in anfMRI study by Delgado, Frank, and Phelps (2005),expectations about the social and moral characteris-tics of trading partners in a trust game modulatedbrain activity in neural structures underlying feed-back processing (i.e., reciprocity). These findingsare consistent with studies showing that prior know-ledge and person-related schemas are important inguiding social and emotional behavior (e.g., Num-menmaa, Peets, & Salmivalli, 2008). To date, it isunknown whether social acceptance or rejectionfeedback differ in informative value depending onexpectations of social evaluation. To our knowledge,the current study is the first to examine whether acti-vation in the mPFC (including v-ACC) and associ-ated brain regions are differentially involved ininterpersonal feedback processing depending onprior expectations to be liked or disliked. Therefore,the first goal of this study was to examine expecta-tion-related activation and context-dependency ofsocial feedback effects.
An important avenue for understanding socialaffective processing is tracking its neurobiologicalontogenetic emergence. Recently, it was demon-strated that brain regions that are involved in socialbehavior (mPFC and v-ACC) are highly sensitive toage-related change (e.g., Blakemore, 2008; Shaw et al.,2008). It is also well documented that children’s con-cerns about acceptance and rejection by peersincrease during late childhood and reach a peak insensitivity in adolescence (e.g., Kloep, 1999; Rose &Rudolph, 2006). Major changes in social sensitivity inadolescence co-occur with a shift in social orientationfrom parents to new social networks with peers(Steinberg & Morris, 2001). In addition, adolescentsbecome increasingly self-conscious and more awarethat they are subject to the evaluation of others(Sebastian et al., 2008). Despite these well-knownchanges in social behavior, the neural substrates thatsupport these developmental changes are still largelyunknown.
The few functional neuroimaging studies attempt-ing to study the neural mechanisms underlying socialand affective processes in adolescence, showedimmature prefrontal cortex activity and enhancedresponses in subcortical brain regions suggesting anintensification of emotional experience and an imma-ture capacity of affect regulation (e.g., Ernst et al.,2005; Galvan et al., 2006; Hare et al., 2008; Monket al., 2003; Yurgelun-Todd & Killgore, 2006). Thismismatch between early maturation of the affectivesystem and the protracted development of brainregions important for regulatory control could biasadolescents towards sensitivities in the social context,such as an increased sensitivity to social acceptanceand rejection by peers (e.g., Dahl, 2008; Nelson,Leibenluft, McClure, & Pine, 2005; Somerville,Jones, & Casey, 2010). In addition, studies demon-strated increased activity in the mPFC associated withself-processing and mentalizing in adolescents com-pared to adults (e.g., Blakemore, 2008; Burnett, Bird,Moll, Frith, & Blakemore, 2009; Sebastian et al.,2008).
There are hardly any studies examining the neuralcorrelates of adolescent sensitivity to peer evaluation.Guyer, McClure-Tone, Shiffrin, Pine and Nelson(2009) examined brain activation while 9–17-year-olds appraised how unfamiliar peers they previouslyhad identified as being of high or low interest wouldevaluate them for a future online chat session. Bothage- and sex-related differences were revealed inbrain regions known to be implicated in social cogni-tion and affective processing. More specifically,females displayed greater age-related increases inactivation in the nucleus accumbens, hypothalamus,hippocampus, and insula during appraisal of socialevaluation by high vs. low interest peers. In males,the activation patterns did not increase with age.Results were interpreted in terms of greater salienceof high interest peers in female adolescents. How-ever, in this study neural responses were associatedwith appraisals of social evaluation, but not to actualsocial feedback processing. The second goal of thecurrent study was to examine developmental differ-ences in neural activation related to social feedbackfrom peers, by including participants from childhoodto adulthood.
To pursue the goals of this study, a modified versionof the Social Judgment task, adopted from Somervilleet al. (2006), was designed. We used event-relatedfMRI in four age groups that have been associated withfour distinct phases of development: pre-pubertal chil-dren (8–10 years), early adolescents (12–14 years), lateadolescents (16–17 years) and young adults (19–25years). In this paradigm, participants were presented
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
NEURAL CORRELATES OF PEER EVALUATION 3
with a series of unfamiliar faces of age-matchedpeers, aiming to probe ecologically valid social inter-actions. Several weeks before testing, participantswere required to send a portrait photograph to theresearcher and were led to believe that others wouldbe forming impressions about them during the interimperiod. During the experiment, the participant viewedfaces and was asked to predict whether the otherperson would like them. The participant then receivedacceptance or rejection feedback from this personthat, unknown to the participant, was generated by thecomputer.
Previously, Somerville et al. (2006) observed that themPFC (particularly the v-ACC) was differentiallyengaged following social acceptance and rejection inadults. However, in this study participants’ judgmentswere collapsed across “like” and “dislike” expectationsto assess the general neural responses associated withfeedback processing. That is, social acceptance andrejection feedback was examined regardless of priorchoice behavior. In addition, violations in the expecta-tions of social feedback were examined regardless offeedback type (i.e., acceptance or rejection). The currentstudy aims to extend this prior study in two ways: (1)We tested neural activation related to expectationsabout social evaluation, and (2) we tested whether theneural response to social evaluative feedback is context-dependent; that is, we examined whether social accept-ance or rejection feedback would be differentially affec-ted by prior expectations to be liked or disliked. Wehypothesized that social feedback would be more salientwhen a participant expected to be liked by the otherperson. Therefore, by extending results from Somervilleet al. (2006), we expected increased activity in themPFC (particularly the v-ACC) to social acceptancefollowing a positive expectation of social evaluation.Furthermore, we hypothesized that regions that areimportant for mentalizing would be involved in expec-tations of social evaluation (Amodio & Frith, 2006).
Second, our analyses focused on developmentalchanges by using specific contrasts (linear and nonlin-ear) to test for periods of heightened social sensitivity.Given adolescents’ increased concern for social eval-uation by peers, we expected that the mPFC and sub-cortical regions would be hypersensitive to signals ofsocial acceptance by peers in mid-adolescence (e.g.,Blakemore, 2008; Nelson et al., 2005; Sebastian et al.,2008). These effects should be enlarged for thoseindividuals who are more sensitive to the influence ofpeers, and for those individuals with higher levels ofanxiety and lower levels of self-perceived socialacceptance and self-worth (e.g., Grosbas et al., 2007;Guyer et al., 2008). To this end, participants wereasked to complete self-report questionnaires.
METHODS
Participants
Sixty volunteers between 8 and 25 years of age wererecruited from the university and through localadvertisements. Three participants (ages 8, 10, and14) were excluded from analyses due to excessivehead movement (>3 mm translation in any direc-tion). In total, 57 participants were included in theanalyses, assigned to four age groups: 12 pre-puber-tal children (7 females; ages 8–10, mean age 9.7, SD= 0.9), 14 early adolescents (8 females; ages 12–14,mean age 13.3, SD = 0.8), 15 late adolescents (7females; ages 16–17, mean age 17.1, SD = 0.6) and16 young adults (8 females; ages 19–25, mean age21.7, SD = 1.9). A chi-square analysis revealed nosignificant differences in gender distributionbetween age groups (p > .9). All participants werehealthy right-handed volunteers with no history ofneurological or psychiatric disorders and receivedfixed payment for participation. For participants,aged 8–17 years, the Child Behavior Check List(Achenbach, 1991) was filled out by primary car-egivers to confirm the absence of behavioral prob-lems. All participants had total scores below theclinical range. Participants and primary caregivers(for minors) gave informed consent. All procedureswere approved by the medical ethical committee ofthe University Medical Center.
Behavioral assessment
In order to obtain an estimate of intellectual func-tioning (IQ), participants completed the Raven’sStandard Progressive Matrices test (Raven, 1941).Estimated mean IQs were 123 (SD = 8.1) for 8–10-year-olds, 122 (SD = 9.0) for 12–14-year-olds, 115(SD = 10.3) for 16–17-year-olds and 125 (SD = 7.5)for 19–25-year-olds. A one-way analysis of variance(ANOVA) revealed a difference in IQ scoresbetween age groups, F(3, 53) = 3.53, p < .05. Post-hoc comparisons revealed that 16–17-year-olds’average IQ was significantly lower relative to 19–25-year-olds (p < .05), and the other groups did notdiffer significantly from each other. Results of theanalyses reported below were corrected for differ-ences in IQ by adding IQ as a covariate factor to thedata abstracted from the spherical ROIs. That is,ANOVAs were performed to characterize activationpatterns in these ROIs. None of the effects wereinfluenced by IQ. Therefore, IQ differences are notdescribed further.
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
4 GUNTHER MOOR ET AL.
Participants of all age groups completed the Dutchversion of the Resistance to Peer Influence question-naire (RPI; Steinberg & Monahan, 2007), which con-sists of 10 pairs of opposite statements about inter-individual interactions. A one-way ANOVA revealeda difference between age groups, F(3, 49) = 7.3, p < .001.Averaged RPI scores were 2.59 (SD = .11) for 8–10-year-olds, 2.97 (SD = .09) for 12–14-year-olds, 3.01(SD = .09) for 16–17-year-olds and 3.25 (SD = .09)for 19–25-year-olds, indicating an age-relatedincrease in resistance to peer influence.
Self-report measures of anxiety and perceivedcompetence were administered in 8–17-year-olds.These questionnaires were not administered in adults,because these tests are validated for minors only.Level of anxiety was measured with the Multidimen-sional Anxiety Scale for Children (MASC; March,Parker, Sullivan, Stallings, & Conners, 1997; Dutchtranslation: Utens & Ferdinand, 2000). The MASCconsists of 39 statements indicating anxiety-arousingsituations and contains a total anxiety scale and foursubscales: physical symptoms, harm avoidance, socialanxiety, and separation anxiety. A one-way ANOVArevealed a main effect of Age group on the separationanxiety subscale, F(2, 38) = 6.7, p < .005, showing adecrease in separation anxiety with age. For all othersubscales no differences between age groups wereshown (all p values > .15). Self-perceived competencewas measured with the Self-Perception Profile forChildren in 8–10-year-olds (SPP-C; Harter, 1985;Dutch translation: Veerman, Straathof, Treffers, vanden Bergh, & Ten Brink, 1997) and the Self-percep-tion Profile for Adolescents in 12–17-year-olds (SPP-A:Harter, 1988; Dutch translation: Treffers et al., 2002).For the purpose of this study only the subscales ofsocial acceptance and global self-worth were used.One-way ANOVAs revealed no differences on thesesubscales between age groups (both p values > .15).
Experimental task
The task used was a modified version of the SocialJudgment task previously described by Somerville et al.(2006). Approximately 2 weeks prior to the experi-ment, participants and caregivers (for minors) werecontacted by telephone and were told that this was astudy about first impressions. For this reason, theywere instructed to send a photograph of their portraitor of their participating child to the researcher. Theywere told that this photograph would be rated by apanel of peers on first impressions. In addition, partic-ipants were informed that the faces of this panelwould be presented during the fMRI experiment and
that their task would be to decide whether theybelieved they were liked or disliked by the peers ofthe panel. Caregivers were explicitly instructed toexplain the procedure of this experiment to their childprior to the testing day.
During the fMRI experiment, participants per-formed an experimental task in which pictures of neu-tral faces of age-matched peers were presented. Fourdifferent versions of the task were created containingfaces of matching age groups. Each version consistedof a total of 120 different faces with an equal distribu-tion of male and female faces. Faces were presentedagainst a black background and were displayed onceduring the testing session. Facial stimuli of childrenwere obtained with the help of primary schools andhigh schools in different cities in the Netherlands.Photographs were taken after caregivers providedwritten approval to use the pictures for scientific pur-poses. Young adults were photographed after writtenapproval at the campus of another university. Meanages of the photographed individuals for the age-matched tasks were as follows: 10.2 for 8–10-year-olds; 13.0 for 12–14-year-olds; 15.9 for 16–17-year-olds; and 22.1 for 19–25-year-olds. Care was taken totest whether differences in brain activation betweenage groups could be due to differences in perceivedvalence of the faces. An independent sample of partic-ipants in the same age groups rated the valence of thepictures using the Self-Assessment Manikin (SAM)on a 9-point scale (Lang, Bradley, & Cuthbert, 2005).Averaged scores were 5.09 (SD = 0.65) for 19–25-year-olds, 5.05 (SD = .69) for 16–17-year-olds, 5.02(SD = .33) for 12–14-year-olds, and 5.48 (SD = 1.16)for 8–10-year-olds, revealing no differences betweenage groups, F(3, 158) = 2.63, p > .05.
The experimental task required participants tomake judgments about the presented faces. Partici-pants were instructed to predict whether the personon the picture would like or dislike them. On eachtrial, the participant was required to answer the ques-tion “Do you believe this person liked you?” Noexplicit instructions were given concerning on whatbasis participants should evaluate the faces. Thiswas done to avoid the induction of systematic strate-gies used for evaluating the faces, so as to mimic theintegrity of real-life social interactions. Judgmentswere followed by feedback indicating acceptance orrejection by the person on the picture (YES vs. NO).In reality, the participants were not judged by thepanel, but the feedback was selected by the com-puter and resulted in 50% acceptance and 50% rejec-tion feedback.
Trials started with a fixation cross, followed by a 3-scue display (see Figure 1). The duration of the fixation
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
NEURAL CORRELATES OF PEER EVALUATION 5
cross was jittered based on an optimalization program(optseq2, see http://surfer.nmr.mgh.harvard.edu/opt-seq), developed by Dale (1999). The onset of the cuewas indexed by the presentation of a face, whichremained on the screen until the end of the trial. Dur-ing the cue display participants were instructed togive a “yes” or “no” answer by giving a left-or right-button response, using the index and middle finger oftheir right hand within the 3-s timeframe. After the 3-scue-period, the choice of the subject appeared on theleft side of the face (YES/NO), during a 1-s delayperiod. The delay was followed by a 2-s feedback dis-play during which the feedback (YES/NO) was pre-sented on the right side of the face. Responses thatwere not made within the 3-s timeframe were fol-lowed by a 3-s “Too Slow” presentation, signaling theend of the trial. Mean percentages of “Too slow” trialswere as follows: 1.39 % for 8–10-year-olds, 1.07 %for 12–14-year-olds, 0.39 % for 16–17-year-olds, and0.37% for 19–25-year-olds.
Task design
Participants received positive feedback on half of thetrials (60 trials) and negative feedback on the otherhalf (60 trials). Feedback type was equally dividedwith regard to the gender of the faces. Both the orderof feedback type and the gender of the faces were deter-mined with an algorithm designed to maximize the effi-ciency of recovery of the blood-oxygen-level-dependent
(BOLD) response (optseq2). To eliminate possibleeffects of the order in which faces were presented,four different sequences of facial stimuli were used.The matching between faces and feedback types wasoutweighed in the sense that for half of the partici-pants a certain facial stimulus was coupled with nega-tive feedback and for the other half of the participantswith positive feedback. Feedback could be congruentor incongruent vis-à-vis the judgment made by theparticipant. This design resulted in the following feed-back conditions: congruent accepted (yes–yes), incon-gruent accepted (no–yes), congruent rejected (no–no)and incongruent rejected (yes–no).
Procedure
Prior to scanning, participants were reminded of thepurpose of the study by a rehearsal of the cover storyand received 10 practice trials. Children were famil-iarized with the scanner environment through the useof a mock scanner. During scanning, participantscompleted two runs of 60 trials with a short break inbetween. Self-report questionnaires were adminis-tered before the scanning session and the RavenMatrices test after the scanning session. At the end ofthe experiment, participants were asked to write downtheir experiences and thoughts about the experiment.None of the participants expressed doubts about thecover story. Participants were debriefed at the end ofthe experiment.
Figure 1. Example of a trial sequence (yes–no condition) in the Social Judgment task (adopted from Somerville et al., 2006). During the cueperiod participants responded to the question, “Do you think this person liked you?” During the delay period the choice of the participantappeared on the left side of the face. Following the delay, acceptance or rejection feedback was presented on the right side of the face. Trialswere separated by intertrial intervals (jittered), where a central fixation cross was shown.
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
Scanning was performed with a standard whole-headcoil on a 3-T Philips Achieva scanner at the Univer-sity Medical Center. Head motion was restricted,using foam inserts surrounding the head. The experi-mental task was projected onto a screen that partici-pants could view through a mirror connected on thehead coil. Functional data were acquired using T2*-weighted echo planar imaging (EPI) during two func-tional runs of 220 volumes each. The two first vol-umes of each run were discarded to allow forequilibration of T1 saturation effects. Each volumecovered the whole brain (38 slices of thickness 2.75mm, field of view 220 mm, 80 × 80 matrix, inplaneresolution 2.75 mm) and was acquired every 2200 ms(TE = 30 ms, descending acquisition). High-resolutionT2* weighted images and high-resolution T1 anatom-ical images were collected after the functional runs.All anatomical scans were reviewed by the radiologydepartment of the University Medical Center.
Data preprocessing and analysis were conductedusing SPM2 (Wellcome Department of CognitiveNeurology, London). Images were corrected for dif-ferences in timing of slice acquisition, followed byrigid body motion correction. Functional volumeswere spatially normalized to EPI templates. The nor-malization algorithm used a 12-parameter affine trans-formation together with a nonlinear transformationinvolving cosine basic functions and resampled thevolumes to 3 mm cubic voxels. Templates were basedon the MNI305 stereotaxic space. Functional volumeswere spatially smoothed with an 8- mm full width athalf maximum (FWHM) isotropic Gaussian kernel.
Statistical analyses were performed on individualsubjects’ data using the general linear model (GLM)in SPM2. The fMRI time series data were modeled bya series of events convolved with a canonical hemody-namic response function (HRF). Both the onset of thecue display and the onset of the feedback display ofeach trial were modeled as zero-duration events. Trialson which participants did not respond within the 3-scue period were not included in the contrasts of inter-est. The onset of the cue display related to the presen-tation of the faces was divided in two conditions: a“yes” condition (“like” expectation) and a “no” condi-tion (“dislike” expectation). Because the task requireda similar judgment on each trial, these event typeswere modeled at the onset of the cue display, sinceparticipants are likely to begin assessing the facesbefore rating them. The feedback display was dividedin four conditions: congruent accepted (yes–yes),incongruent accepted (no–yes), congruent rejected(no–no) and incongruent rejected (yes–no). Feedback
effects were examined for “yes” and “no” judgmentstrials separately. The modeled events were used ascovariates in a general linear model, along with a basicset of cosine functions that high-pass filtered the data,and a covariate for session effects. The least-squaresparameter estimates of height of the best-fittingcanonical HRF for each condition were used in pair-wise contrasts. The resulting contrast images, com-puted on a subject-by-subject basis, were submitted togroup analyses. At the group level, whole-brain con-trasts were computed by performing one-tailed t-tests,treating participants as a random effect. Task-relatedresponses were considered significant if they con-sisted of at least 10 contiguous voxels that exceededan uncorrected threshold of p < .001. This thresholdwas based on other neuroimaging studies involvingcomparisons between different age groups (e.g., VanDuijvenvoorde, Zanolie, Rombouts, Raijmakers, &Crone, 2008; Van Leijenhorst et al., 2010).
We performed voxelwise ANOVAs to identifyregions that showed age-related differences in activa-tion to the judgment of the faces and during feedbackprocessing. Both linear and quadratic age-relatedtrends were tested. In addition, it was examinedwhether activation peaked specifically in one agegroup relative to the other three groups (see Table 1).ANOVAs were considered significant if they con-sisted of at least 10 contiguous voxels that exceededan uncorrected threshold of p < .001. In addition, weused the MARSBAR toolbox (Brett, Anton, Vala-bregue, & Poline, 2002) to perform region of interest(ROI) analyses in brain regions that were identified inthe ANOVAs. We created 6-mm spherical ROIs at thepeak activity voxel of these regions to further charac-terize patterns of activation. Correlations with age onthe data abstracted from these ROIs are presented forillustrative purposes only (Vul, Harris, Winkielman,& Pashler, 2009). Finally, correlational analyses (two-tailed; Pearson’s correlation) were performed foractivation in the spherical ROIs and (1) behavior onthe task and (2) self-report questionnaires. This was
TABLE 1 Age-related trends that were tested at whole
brain level using voxelwise ANOVAs
Age-related trend Contrast
Linear increase –3 –1 1 3Linear decrease 3 1 –1 –3Quadratic trend –1 1 1 –1Peak in 8–10-year olds 3 –1 –1 –1Peak in 12–14-year olds –1 3 –1 –1Peak in 16–17-year olds –1 –1 3 –1Peak in 19–25-year olds –1 –1 –1 3
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
NEURAL CORRELATES OF PEER EVALUATION 7
done for the contrasts yes > no judgments and rejec-tion > acceptance following a “no” judgment sinceother contrasts did not reveal any age-related trends(see “Results” section). These correlational analysesare presented with exact p-values.
RESULTS
Behavioral results
The mean numbers of trials per feedback condition foreach age group are presented in Table 2. A one-wayANOVA with age group as between-subjects factordemonstrated a main effect of Age group, F(3, 53) =6.39, p = .001, on difference scores for “yes” vs. “no”choices. Differences were further explored by post-hocTukey comparisons, revealing a difference between19–25-year-olds and 8–10-year-olds, and between 19–25-year-olds and 12–14-year-olds (p < .05, p < .001respectively). That is, 19–25-year-olds made signifi-cantly more “yes” choices than 8–10-year-olds and 12–14-year-olds. 16–17-year-olds did not differ signifi-cantly from adults or from the two younger age groups.One-sample t-tests confirmed that 19–25-year-oldsmade more “yes” judgments (60 %) relative to a 50 %baseline and fewer “no” (39.64 %) judgments (both pvalues < .001). For all other age groups no significantdifferences between “yes” and “no” judgments vs. 50%baseline were found.
Further, age-related differences were examined forreaction times (RTs) for “yes” and “no” judgments(see Table 2). An ANOVA with age group asbetween-subjects factor revealed no significant differ-ence in RT between “yes” and “no” choices, F(1, 53)
= 0.44, p > .5, and the interaction effect with Agegroup failed to reach significance, F(3, 53) = 1.85,p > .15. There was, however, a main effect of Agegroup, F(3, 53) = 3.5, p < .05, showing that 19–25-year-olds (M = 1436, SD = 283) generally respondedmore slowly than 8–10-year-olds (M = 1126, SD =174). RTs for the other age groups did not differ fromeach other.
fMRI results
The results are presented in two sections. The firstsection contains whole-brain comparisons in adults,followed by a second section containing an overviewof analyses examining age-related differences. Bothsections describe two sets of analyses. First, neuralresponses will be presented relating to the onset ofstimulus presentation and associated expectationsabout social evaluation. Second, brain analyses willbe reported examining neural activation related tosocial feedback processing.
Whole-brain comparisons: adults
Judgment of faces. A (GLM) analysis was per-formed on the functional data modeled at the onset ofthe presentation of the faces. Contrasts of interestwere yes > no judgments (i.e., expectation to be likedvs. expectation to be disliked) and no > yes judg-ments. In adults, the comparison yes > no judgmentsresulted in activation in the ventral mPFC, left subcal-losal cortex, left anterior and middle cingulate cortex,right putamen, right amygdala, left hippocampus, andright parahippocampal gyrus (see Figure 2A, p < .001,
TABLE 2 Performance of age groups on the social judgment task
Age group (years)
8–10 12–14 16–17 19–25
Judgment participantYes choice
Mean frequency 58.6 (12.2) 51.9 (15.6) 63.1 (11.6) 72.0 (10.2)Mean RT 1123.9 (162.1) 1281.3 (261.5) 1271.2 (251.9) 1384.0 (342.8)
No choiceMean frequency 59.7 (12.8) 66.8 (15.9) 56.4 (11.9) 47.6 (9.9)Mean RT 1131.9 (194.8) 1261.7 (285.9) 1251.9 (274.5) 1462.2 (285.2)
Notes: Mean frequencies in number of trials for “yes” and “no” judgments and per feedback condition for each age group. Mean reactiontimes (RT) are presented for “yes” and “no” judgments (SD in parentheses).
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
8 GUNTHER MOOR ET AL.
at least 10 contiguous voxels). Significant clusters andcorresponding MNI coordinates are reported in sup-plementary Table 1. The reversed contrast no > yesjudgments did not result in significant activation at athreshold of p < .001.
Social feedback. To examine neural activationrelated to social feedback processing, a GLM analysiswas performed on the functional data modeled at theonset of the feedback. These analyses focused specifi-cally on acceptance vs. rejection feedback for “yes”and “no” judgment trials separately. Prior studies havecollapsed across “yes” and “no” judgments to assessthe neural responses associated with feedback process-ing (Somerville et al., 2006). Here, we tested whether
social feedback was differentially affected by priorchoice behavior. In addition, comparisons revealed dif-ferences in neural responses following “yes” and “no”judgments prior to feedback presentation. Therefore,these choice conditions were analyzed separately.
The first set of whole-brain comparisons tested thecontrasts: acceptance > rejection (yes–yes > yes–no)and rejection > acceptance (yes–no > yes–yes) for“yes” judgments trials. In adults, the comparisonacceptance > rejection resulted in activation in theventral mPFC, right subcallosal cortex, right posteriorcingulate cortex, left orbital frontal cortex (OFC),caudate nucleus, left precuneus, and left thalamus. Inaddition, activation was shown in bilateral middlefrontal gyrus (i.e., BA 46), bilateral superior frontalgyrus and fusiform gyrus (see Figure 3A, p < .001, at
Figure 2. Whole brain results for the contrast YES > NO judg-ments. Results for 19–25-year-olds are presented in panel A.Results for 8–10-year-olds, 12–14-year-olds and 16–17-year-oldsare presented in panel B (uncorrected, p < .001, >10 contiguousvoxels). The reversed contrast NO > YES judgments did not resultin significant activation for any of the four age groups.
Figure 3. Whole brain results for the contrast acceptance > rejec-tion following “yes” choices. Results for 19–25-year-olds are pre-sented in panel A. Results for 8–10-year-olds, 12–14-year-olds and16–17-year-olds are depicted in panel B (uncorrected, p < .001, >10contiguous voxels). The reversed contrast YES–NO > YES–YESdid not result in significant activation for any of the age groups.
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
NEURAL CORRELATES OF PEER EVALUATION 9
least 10 contiguous voxels). Significant clusters andcorresponding MNI coordinates are reported in sup-plementary Table 2. The reversed contrast rejection >acceptance did not result in significant activation at athreshold of p < .001.
The second set of whole-brain comparisons testedthe contrasts acceptance > rejection (no–yes > no–no) and rejection > acceptance (no–no > no–yes) for“no” judgments trials. The contrast acceptance >rejection did not result in significant clusters at athreshold of p < .001. Regions that were active forthe reversed contrast rejection > acceptance are pre-sented in Figure 4A (p < .001, at least 10 contiguousvoxels). Adults showed activation in the right sub-callosal cortex, left caudate, right putamen, rightmiddle frontal gyrus, and right inferior frontalgyrus (BA 46). Significant clusters and correspond-ing MNI coordinates are reported in supplementaryTable 3.
Age comparisons
Judgment of faces. Whole brain results for the con-trast yes > no judgments for the youngest three agegroups are presented in Figure 2B. Sixteen- to seven-teen-year-olds activated a similar set of brain regionsas adults, whereas fewer activations are observed for12–14-year-olds and 8–10-year-olds (coordinates arereported in supplementary Table 1, p < .001, at least10 contiguous voxels). The reversed contrast no > yesjudgments did not result in significant activation forany of the four age groups. The voxelwise ANOVAstesting for age-related changes for the yes > no judg-ments contrast revealed a linear change with age inseveral regions including the right putamen, peak at:30, –3, 3, z = 4.17, t(1, 53) = 4.56, p < .001; ventralmPFC, peak at: 6, 48, –12, z = 3.47, t(1, 53) = 3.69,p < .001; left precuneus, peak at: –12, –45, 39, z =3.57, t(1, 53) = 3.81, p < .001; and parahippocampalgyrus, peak at: 30, 6, –21, z = 4.27, t(1, 53) = 4.69, p <.001. Other significant clusters and correspondingMNI coordinates are reported in supplementary Table 4.We created 6-mm spherical ROIs at the peak activityvoxel of the right putamen and the ventral mPFC tofurther characterize patterns of activation. In Figure 5scatterplots for these regions with age are presentedfor illustrative purposes.
Social feedback. Whole brain results for the con-trast acceptance > rejection (yes–yes > yes–no) for“yes” judgment trials are presented in Figure 3 (coordi-nates are reported in supplementary Table 2, p < .001,at least 10 contiguous voxels). The voxelwise ANO-VAs testing for age-related changes did not result inany significant clusters at a threshold of p < .001. Simi-larly, the reversed contrast rejection > acceptance (yes–no > yes–yes) for “yes” judgment trials did not result insignificant activation for any of the four age groups.
Similar analyses were performed testing for age-related changes in the contrast rejection > accept-ance (no–no > no–yes) following a “no” judgment.In Figure 4, whole brain results for the four agegroups are presented. Adults showed several activa-tion clusters, whereas less activation is observed forthe youngest three age groups (coordinates arereported in supplementary Table 3, p < .001, at least10 contiguous voxels). The voxelwise ANOVAsconfirmed the whole brain findings by showing lin-ear changes with age in regions including the leftsubcallosal cortex/OFC, peak at: –15, 18, –12, z =4.35, t(1, 53) = 4.79, p < .001; left paracingulate cor-tex, peak at: –15, 30, 27, z = 4.13, t(1, 53) = 4.51, p <.001; right OFC, peak at: 24, 18, –21, z = 3.87, t(1, 53)= 4.18, p < .001; left lateral PFC (LPFC, BA 47),
Figure 4. Whole brain results for the contrast rejection > accept-ance following “no” choices. Results for 19–25-year-olds are pre-sented in panel A. Results for 8–10-year-olds, 12–14-year-olds and16–17-year-olds are depicted in panel B (uncorrected, p < .001, >10contiguous voxels). The reversed contrast NO–YES> NO–NO didnot result in significant activation for any of the age groups.
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
10 GUNTHER MOOR ET AL.
peak at: –30, 21, 15, z = 4.01, t(1, 53) = 4.35, p <.001; and left putamen/globus pallidus, peak at: –24,3, 0, z = 3.49, t(1, 53) = 3.71, p < .001. Other signi-ficant clusters and corresponding MNI coordinatesare reported in supplementary Table 5. We created6-mm spherical ROIs at the peak activity voxel ofthe left subcallosal cortex, right OFC, left putamen,left paracingulate cortex and left lateral PFC tofurther characterize patterns of activation. In Figure6 scatterplots for these regions with age are pre-sented. The reversed contrast acceptance > rejection(no–yes > no–no) for “no” judgment trials did notresult in significant activation for any of the four agegroups.
Correlations
Finally, we performed correlational analyses foractivation in the spherical ROIs and (1) behavior onthe task, (2) resistance to peer influence, (3) levels ofanxiety and (4) self-perceived social acceptance andself-worth, within and across age groups. No signi-ficant correlations were found for activation in thespherical ROIs and behavior on the task. Withregard to resistance to peer influence, significantpositive correlations were present between activityin the left putamen (r = .52, p < .001) and the leftLPFC (r = .54, p < .001) for the contrast rejection >acceptance following a “no” judgment across agegroups (see supplementary Figure 1). That is, indi-viduals who were more resistant to peer influenceshowed larger activation in these regions to socialrejection following a “no” judgment. After control-ling for the observed age-related increase in resist-
ance to peer influence, the correlations remainedsignificant (r = .37, p = .01 and r = .37, p = .007,respectively). No significant correlations were foundfor levels of resistance to peer influence within agegroups.
In 8–17-year-olds correlational analyses were per-formed for activation in the spherical ROIs and levelsof anxiety and self-perceived social acceptance andself-worth. Significant positive correlations werepresent between social anxiety and activity in the leftsubcallosal cortex (r = .39, p = .01), left paracingulatecortex (r = .45, p = .003), right OFC (r = .38, p =.014), left putamen (r = .39, p = .012), and left LPFC(r = .38, p = .013) for the contrast rejection > accept-ance following a “no” judgment (see supplementaryFigure 2). That is, those 8–17-year-olds who reportedhigher levels of social anxiety showed larger activa-tion in these regions to social rejection following a“no” judgment. For the same contrast significant neg-ative correlations were present between self-worthand activity in the left subcallosal cortex (r = –.39, p =.03) and the left putamen (r = –.42, p = .006). Individ-uals with lower levels of self-worth showed largeractivation in these regions to social rejection follow-ing a “no” judgment (see supplementary Figure 3). Nosignificant correlations were found for self-perceivedsocial acceptance. Finally, similar correlations werecalculated for each age group separately. Only in 16–17-year-olds was self-worth negatively correlatedwith activation in the left subcallosal cortex (r = –.75,p = .001), right OFC (r = –.625, p = .013), left puta-men (r = –.55, p = .034), and LPFC (r = –.65, p =.009) for the contrast rejection > acceptance followinga “no” judgment.
Figure 5. Scatterplots for the difference in activation between “yes” and “no” judgments against age. Spherical ROIs for the right putamen(30, –3, 3) and ventral mPFC (6, 48, –12).
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
NEURAL CORRELATES OF PEER EVALUATION 11
DISCUSSION
The goal of this study was to investigate the neuralcorrelates of expectations about social evaluationand social feedback processing. In addition, neu-rodevelopmental changes of concern about socialacceptance and rejection were examined. Brainimaging data in adults yielded two main results: (1)
The ventral mPFC was active when participantsexpected positive evaluations by peers. (2) The ven-tral mPFC, in particular the subcallosal cortex, wasdifferentially involved in social acceptance andrejection feedback depending on prior expectationsof social evaluation. In addition, expectation andfeedback-related activation was associated withactivity in the striatum.
Figure 6. Scatterplots for the difference in activation between rejection and acceptance feedback following a “no” judgment against age.Spherical ROIs for the left subcallosal cortex/OFC (–15, 18, –12), left paracingulate cortex (–15, 30, 27), left putamen (–24, 3, 0), right OFC(24, 18, –21), and left lateral PFC (–30, 21, 15).
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
12 GUNTHER MOOR ET AL.
Age-related comparisons revealed four importantresults: (1) There was increased activation with age inthe ventral mPFC and striatum associated with theexpectation to be liked by a peer. (2) Acceptancefeedback following the expectation of positive socialevaluation resulted in overlapping patterns of activa-tion in the ventral mPFC and striatum across agegroups. (3) Rejection feedback following a negativeexpectation of social evaluation resulted in activationin the striatum, ventral mPFC, OFC and LPFC inadults, but not in younger participants. (4) The lattereffect was generally stronger for those individualswho reported high resistance to peers and for those 8–17-year-olds who scored higher on scales of socialanxiety and lower on self-worth.
The discussion is organized around the two maingoals of this study. First, brain imaging data will bediscussed addressing the expectation-related activa-tion and context dependency of social feedbackeffects in adults. This will be followed by a discussionof results examining age-related differences.
Neural correlates of expectations about social evaluation in adults
The results of the current study show activity in boththe ventral mPFC and striatum for “like” compared to“dislike” expectations of social evaluation. Thesefindings are consistent with studies showing that thestriatum responds to anticipation of potential rewards(e.g., Knutson, Adams, Fong, & Hommer, 2001).While activation in the striatum has been associatedwith both social and non-social rewards, activity inthe mPFC is most consistently reported for socialrewards and social decision-making (Izuma, Saito, &Sadato, 2008). Together, these lines of evidence couldbe taken to suggest that the findings of the currentstudy are specific for social interactions. That is, themPFC is believed to be implicated in self-referentialprocessing and mentalizing, which are also likely tobe important for making inferences about whetherother people would like us (Amodio & Frith, 2006).The involvement of the mPFC for positive expecta-tions of social evaluation could suggest that partici-pants use different strategies for “like” vs. “dislike”expectations. Possibly, positive expectations of peersocial evaluation rely more on self-knowledge andself-consciousness, which could be more adaptivewhen taking a risk in predicting to be liked, whereasthe intentions of the other person are unknown. Previ-ous imaging studies highlighted a functional dissocia-tion between dorsal and ventral regions of the mPFCas a function of how similar one perceives another
person to be to oneself (Mitchell, Macrae, & Banaji,2006). These studies have shown that the ventralmPFC is important for mentalizing about a similarother, whereas mentalizing about a dissimilar otherengages a more dorsal section of the mPFC. In thecurrent study activation was predominantly located inthe ventral regions of mPFC, which overlap consider-ably with the region reported by Mitchell et al.(2006). This could suggest that adults had a higherexpectation to be liked by individuals whom they per-ceived as more similar to themselves. This suggestionshould be examined more closely in future research.
Neural correlates of context-dependency of social feedback effects in adults
Previously, it was shown that the ventral part of themPFC, in particular the v-ACC, is sensitive to socialacceptance feedback (Somerville et al., 2006). Here,we extend this finding by showing that these effectsare specific for situations in which participants have apositive expectation of social evaluation. Further, thisacceptance feedback is again accompanied by activa-tion in the striatum (caudate nucleus). Possibly, socialacceptance is more salient when participants expect tobe liked. This neural response may be highly sensitiveto social expectations and demonstrates the rewardvalue of social acceptance that is aligned with an indi-vidual’s own expectation of social evaluation.
This hypothesis is reinforced by the findings onsocial rejection feedback. In adults, rejection com-pared to acceptance feedback following a “dislike”expectation resulted in activation in the subcallosalcortex, striatum and left LPFC. No significant activa-tion was found for rejection relative to acceptancefeedback following the expectation to be liked by theother person. These results may indicate the import-ance of learning from feedback from others thatmatches prior expectations in a social context. Resultsof the current study are consistent with previouslyreported striatum activation in a social learning con-text (Klucharev, Hytönen, Rijpkema, Smidts, & Fern-ández, 2009; Rilling et al., 2002). For instance, in thestudy by Klucharev et al. (2009) the ventral striatumwas found to be more active when there was no con-flict with group opinion in a task designed to measuresocial conformity. Another interpretation that needs tobe considered is that the observed pattern of neuralactivation to social rejection feedback could reflectstrategies to regulate the negative thoughts and feel-ings associated with expected social rejection. Indeed,other studies have shown that regions of the mPFC,
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
NEURAL CORRELATES OF PEER EVALUATION 13
LPFC and striatum are involved in affect regulation,such as tasks requiring cognitive reappraisal of nega-tive emotional images (e.g., Ochsner et al., 2004;Wager, Davidson, Hughes, Lindquist, & Ochsner,2008). Note that affect regulation seems to be specificfor those situations where participants had negativeexpectations of social evaluation. It was surprisingthat these effects were not observed for the feedbackcondition where the negative impact of social rejec-tion was expected to be the highest (Yes–No condi-tion). One method to explore the absence of thisrelation in more detail is by the use of time-specificmeasures, such as event-related potentials. In futurestudies it will also be important to include an explicitcontrol condition. Overall, the results of this studyadd to prior studies by showing that the brain proc-esses to social feedback differ depending on expecta-tions of social evaluation.
Developmental changes
At a behavioral level, 8–10-year-olds respondedfaster for both “like” and “dislike” expectations ofpeer social evaluation compared to adults. Theobserved increase in reaction times with age couldsuggest that children, adolescents and adults use dif-ferential strategies for evaluating the faces. In addi-tion, the proportions of positive vs. negativejudgments about peer evaluation differed betweenage groups, in such a way that adults more oftenexpected to be liked than the younger age groups.Possibly, adults have a greater tendency to expect beliked by peers. The increase in like expectations withage could also reflect developmental changes in self-focus and self-consciousness (Sebastian et al.,2008). Interestingly, the brain imaging results showthat in 16–17-year-olds a similar set of brain regionswas active as in adults for positive compared to neg-ative expectations of social evaluation, while lessactivation was observed in the two youngest agegroups. More specifically, activation in the ventralmPFC and right putamen revealed a linear increasewith age: These brain regions were most active for“like” compared to “dislike” expectations in olderparticipants. These results suggest that regions thatare involved in mentalizing and self-processing (par-ticularly the mPFC) show an age-related increase insensitivity to positive expectations of social evalua-tion (Amodio & Frith, 2006). The absence of signi-ficant positive correlations between the number of“like” expectations and activation in the selectedROIs indicates that these linear increases with ageare not likely to be due to age-related differences in
behavior. In future studies it will be important toexamine this issue in more detail.
Second, age groups were similarly sensitive tosocial acceptance feedback. That is, acceptance feed-back following the expectation to be liked resulted insimilar ventral mPFC and striatum activation acrossage groups. These results indicate that social accept-ance is salient in all age groups. It is possible that inyoung children the system is already wired to respondto social acceptance, which could reflect the highevolutionary value of social acceptance. Recently,Lieberman and Eisenberger (2009) highlighted theoverlap between brain regions that respond to phys-ical pain and pleasures and more abstract social expe-riences such as social rejection and acceptance. Thealleged overlap in neural circuitry of physical andsocial needs is consistent with the notion concerningthe high evolutionary value of social acceptance inpromoting survival and well-being in humans (Pank-sepp, 2003).
Third, age-related differences were observed inneural responses to social rejection feedback follow-ing a negative expectation of social evaluation. Thatis, activation in the striatum (e.g., putamen/globuspallidus), subcallosal cortex, paracingulate cortex,LPFC, and OFC showed a linear increase with age:These regions were most active in adults. One expla-nation that has to be considered is that adults are bet-ter able to regulate the negative feelings associatedwith expected social rejection than the younger agegroups. In the light of results of prior studies, thiswould fit the hypothesis of immature affect regulationand self-control in children and adolescents (e.g., Nel-son et al., 2005). This interpretation is further sup-ported by our results showing that activity in thestriatum and LPFC is sensitive to individual differ-ences in resistance to peer influence. Individuals whoshow high resistance to peers showed larger activa-tion to social rejection feedback, even when control-ling for age.
Finally, in 8–17-year-olds, brain activity related tosocial feedback processing was sensitive to individualdifferences in social anxiety and self-perceived self-worth. That is, those 8–17-year-olds with higher lev-els of social anxiety showed larger activation in thesubcallosal cortex, paracingulate cortex, OFC, LPFC,and putamen to social rejection when having a nega-tive expectation of social evaluation. Individuals withlower levels of global self-worth showed larger acti-vation in the subcallosal cortex and putamen. Itshould be mentioned that the ROIs that were used forthese correlation analyses were based on whole-braincomparisons with age as predicting factor. However, allthese effects remained significant also when controlling
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
14 GUNTHER MOOR ET AL.
for age-related variance. Thus, these regions are sen-sitive to both age-related and individual differences.Possibly, these results suggest a strong impact ofsocial rejection feedback in 8–17-year-olds withhigher levels of social anxiety and lower levels ofself-worth. A limitation of the current study is that notall correlational analyses would survive Bonferonnicorrection for multiple comparisons, increasing theprobability of Type 1 errors. Therefore, caution in theinterpretation of these findings is warranted.
The results of the current study do not fit well withthe hypothesis that the mPFC and subcortical brainregions would be hypersensitive to peer social evalua-tion in adolescence. Given adolescents’ increasedself-consciousness and concerns about social evalua-tion, social sensitivity was expected to peak in earlyand middle adolescence (Kloep, 1999). The few func-tional neuroimaging studies so far showed immatureprefrontal cortex activity and enhanced responses insubcortical brain regions (e.g., Blakemore, 2008;Nelson et al., 2005), which could bias adolescentstowards heightened sensitivity in a social context. Ourresults do not reveal a peak in activation in adoles-cence, which was tested in terms of both anticipationof social evaluation and social feedback processing.The results of the current study, however, indicatethat social acceptance feedback is salient from child-hood into adulthood. While the adolescent period ischaracterized by an increased focus on peers, thisdoes not necessarily imply a heightened sensitivity inbrain regions implicated in social and affective behav-ior to peer interactions.
It should be noted that the inclusion of participantsfrom a broad age range poses several challenges foran optimal study design. A limitation of the currentstudy is the use of different photos in different agegroups. Our reasoning behind the use of photos ofage-matched peers was driven by the goal of the studyto probe ecologically valid peer interactions. How-ever, if participants were presented with a similar setof photos of a certain age group (for instance, onlyphotos of adolescents), the experience of viewing thefaces could also be different for different age groups.In future research, photos of faces of a broad agerange could be used in a single task version, whichhowever would lengthen the task significantly. Thesecond limitation of the current study is its cross-sectional design. Longitudinal studies should beundertaken to track changes in social behavior andbrain development. In addition, in future studies itwill be important to examine cross-cultural differ-ences in order to investigate whether cultural back-ground would have an impact on the experience ofsocial-evaluative feedback.
CONCLUSIONS
Together, the results of this study demonstrate the con-text-dependency of neural activation related to socialfeedback effects. That is, the ventral mPFC and striatumwere differentially involved in social acceptance andrejection feedback depending on prior expectations con-cerning social evaluation in adults. In addition, resultsdemonstrated increased activity in these brain regionsfor positive compared to negative expectations of socialevaluation. To our knowledge, this is the first studyexamining neurodevelopmental changes in brain activ-ity underlying appraisals of social evaluation and socialfeedback processing by including participants from abroad age range. Age differences were found in neuralresponses associated with positive expectations of socialevaluation and social rejection feedback. No age-relateddifferences were shown to social acceptance feedback.Finally, although the current study focused on psychiat-rically healthy participants, it may provide importantinsights in the neural correlates of individual differ-ences. It would be of considerable interest to examinethe neural correlates of social feedback processing inchildren and adults with psychopathology.
REFERENCES
Achenbach, T. M. (1991). Manual for Child BehaviorChecklist/4-18 and 1991 Profile. Burlington, VT:University of Vermont.
Amodio, D. M., & Frith, C. D. (2006). Meeting of minds:The medial frontal cortex and social cognition. NatureReviews Neuroscience, 7(4), 268–277.
Baumeister, R. F., & Leary, M. R. (1995). The need tobelong: Desire for interpersonal attachments as a funda-mental human motivation. Psychological Bulletin,117(3), 497–529.
Blakemore, S. J. (2008). The social brain in adolescence.Nature Reviews Neuroscience, 9(4), 267–277.
Blakemore, S. J., Winston, J., & Frith, U. (2004). Socialcognitive neuroscience: Where are we heading? Trendsin Cognitive Science, 8(5), 216–222.
Brett, M., Anton, J. L., Valabregue, R., & Poline, J. B.(2002). Region of interest analysis using an SPM toolbox.NeuroImage, 16(2), abstract 497.
Burnett, S., Bird, G., Moll, J., Frith, C., & Blakemore, S. J.(2009). Development during adolescence of the neuralprocessing of social emotion. Journal of Cognitive Neu-roscience, 21(9), 1735–1750.
Dahl, R. E. (2008). Biological, developmental, and neurobe-havioral factors relevant to adolescent driving risks. Ameri-can Journal of Preventive Medicine, 35, S278–S284.
Dale A. M. (1999). Optimal experimental design for event-related fMRI. Human Brain Mapping, 8(23), 109–114.
Delgado, M. R., Frank, R. H., & Phelps, E. A. (2005). Per-ceptions of moral character modulate the neural systemsof reward during the trust game. Nature Neuroscience,8(11), 1611–1618.
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
NEURAL CORRELATES OF PEER EVALUATION 15
Ernst, M., Nelson, E. E., Jazbec, S., McClure, E. B., Monk, C. S.,Leibenluft, E., et al. (2005). Amygdala and nucleusaccumbens in response to receipt and omission of gains inadults and adolescents. NeuroImage, 25(4), 1279–1291.
Frith, C. D., & Frith, U. (1999). Interacting minds: A bio-logical basis. Science, 286(5445), 1691–1695.
Gallagher, H. L., & Frith, C. D. (2003). Functional imagingof ‘theory of mind’. Trends in Cognitive Science, 7(2),77–83.
Galvan, A., Hare, T. A., Parra, C. E., Penn, J., Voss, H.,Glover, G., et al. (2006). Earlier development of theaccumbens relative to orbitofrontal cortex might under-lie risk-taking behavior in adolescents. Journal of Cog-nitive Neuroscience, 26(25), 6885–6892.
Grosbas, M. H., Jansen, M., Leonard, G., McIntosh, A.,Osswald, K., Poulsen, C., et al. (2007). Neural mecha-nisms of resistance to peer influence in early adoles-cence. Journal of Neuroscience, 27(30), 8040–8045.
Guyer, A. E., Lau, J. Y., McClure-Tone, E. B., Parrish, J.,Shiffrin, N. D., Reynolds, R. C., et al. (2008). Amygdalaand ventrolateral prefrontal cortex function during antic-ipated peer evaluation in pediatric social anxiety.Archives of General Psychiatry, 65(11), 1303–1312.
Guyer, A. E., McClure-Tone, E. B., Shiffrin, N. D., Pine, D. S.,& Nelson, E. E. (2009). Probing the neural correlates ofanticipated peer evaluation in adolescence. Child Devel-opment, 80(4), 1000–1015.
Hare, T. A., Tottenham, N., Galvan, A., Voss, H. U.,Glover, G. H., & Casey, B. J. (2008). Biological sub-strates of emotional reactivity and regulation in adoles-cence during an emotional go–nogo task. BiologicalPsychiatry, 63(10), 927–934.
Harter, S. (1985). Manual for the Self-Perception Profile forChildren. Denver, CO: University of Denver.
Harter, S. (1988). Manual for the Self-Perception Profile forAdolescents. Denver, CO: University of Denver.
Izuma, K., Saito, D. N., & Sadato, N. (2008). Processing ofsocial and monetary rewards in the human striatum.Neuron, 58(2), 284–294.
Kloep, M. (1999). Love is all you need? Focusing on ado-lescents’ life concerns from an ecologically point ofview. Journal of Adolescence, 22(1), 49–63.
Klucharev, V., Hytönen, K., Rijpkema, M., Smidts, A., &Fernández, G. (2009). Reinforcement learning signalpredicts social conformity. Neuron, 61(1), 140–151.
Knutson, B., Adams, C. M., Fong, G. W., & Hommer, D.(2001). Anticipation of increasing monetary rewardselectively recruits nucleus accumbens. Journal of Cog-nitive Neuroscience, 21(16), 1–5.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2005). Inter-national affective picture system (IAPS): Affective rat-ings of pictures and instruction manual. Gainesville, FL,University of Florida.
Lieberman, M. D., & Eisenberger, N. I. (2009). Pains andpleasures of social life. Science, 323(5916), 890–891.
March, J. S., Parker, J. D. A., Sullivan, K., Stallings, P., &Conners, C. K. (1997). The Multidimensional AnxietyScale for Children (MASC): Factor structure, reliability,and validity. Journal of American Academy of Child andAdolescent Psychiatry, 36(4), 554–565.
Mitchell, J. P., Macrae, C. N., & Banaji, M. R. (2006).Dissociable medial prefrontal contributions to judg-ments of similar and dissimilar others. Neuron, 50(4),655–663.
Monk, C. S., McClure, E. B., Nelson, E. E., Zarahn, E.,Bilder, R. M., Leibenluft, E., et al. (2003). Adolescentimmaturity in attention-related brain engagement toemotional facial expressions. NeuroImage, 20(1),420–428.
Nelson, E. E., Leibenluft, E., McClure, E. B., & Pine, D. S.(2005). The social re-orientation of adolescence: A neuro-science perspective on the process and its relation to psy-chopathology. Psychological Medicine, 35(2), 163–174.
Nummenmaa, L., Peets, K., & Salmivalli, C. (2008). Auto-matic activation of adolescents’ peer-relational schemas:Evidence from priming with facial identity. Child Devel-opment, 79(6), 1659–1675.
Ochsner, K. N., Ray, R. D., Cooper, J. C., Robertson, E. R.,Chopra, S., Gabrieli, J. D. E., et al. (2004). For better orfor worse: Neural systems supporting the cognitivedown- and up-regulation of negative emotion. NeuroIm-age, 23(2), 483–499.
Panksepp, J. (2003). Feeling the pain of social loss. Science,302(5643), 237–239.
Raven J. C. (1941). Standardization of progressive matrices.British Journal of Medical Psychology, 19, 137–150.
Rilling, J. K., Gutman, D. A., Zeh, T. R., Pagnoni, G.,Berns, G. S., & Kilts, C. D. (2002). A neural basis forsocial cooperation. Neuron, 35(2), 395–405.
Rilling, J. K., King-Casas, B., & Sanfey, A. G. (2008). Theneurobiology of social decision- making. Current Opin-ion in Neurobiology, 18(2), 159–165.
Rose, A. J., & Rudolph, K. D. (2006). A review of sexdifferences in peer relationship processes: Potentialtrade-offs for the emotional and behavioral develop-ment of girls and boys. Psychological Bulletin,132(1), 98–131.
Sebastian, C., Burnett, S., & Blakemore, S. J. (2008).Development of the self-concept during adolescence.Trends in Cognitive Sciences, 12(11), 441–446.
Shaw, P., Kabani, N. J., Lerch, J. P., Eckstrand, K., Lenroot,R., Gogtay, N., et al. (2008). Neurodevelopmental tra-jectories of the human cerebral cortex. Journal of Neuro-science, 28(14), 3586–3594.
Somerville, L. H., Heatherton, T. F., & Kelley, W. M.(2006). Anterior cingulate cortex responds differentiallyto expectancy violation and social rejection. Nature Neu-roscience, 9(8), 1007–1008.
Somerville, L. H., Jones, R. M., & Casey, B. J. (2010). A timeof change: Behavioral and neural correlates of adolescentsensitivity to appetitive and aversive environmental cues.Brain and Cognition, 71(1), 124–133. Retrieved July 31,2009. doi:10.1016/j.bandc.2009.07.003
Steinberg, L., & Monahan, K. C. (2007). Age differences inresistance to peer influence. Developmental Psychology,43(6), 1531–1543.
Steinberg, L., & Morris, A. M. (2001). Adolescent develop-ment. Annual Review of Psychology, 52, 83–110.
Treffers, D. A., Goedhart, A. W., Veerman, J. W., Van denBergh, B. R. H., Ackheart, L., & De Rijcke, L. (2002).Manual of the Self-perception Scale for Adolescents.Lisse, The Netherlands: Swets and Zeitlinger.
Utens, E. M. W. J., & Ferdinand, R. F. (2000). Dutch trans-lation of the March’s Multidimensional Anxiety Scale forChildren (MASC-NL). Rotterdam, The Netherlands:AZR-Sophia/Erasmus University.
Van Duijvenvoorde, A. C., Zanolie, K., Rombouts, S. A.,Raijmakers, M. E., & Crone, E. A. (2008). Evaluating
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
16 GUNTHER MOOR ET AL.
the negative or valuing the positive? Neural mechanismssupporting feedback-based learning across development.Journal of Neuroscience, 28(38), 9495–9503.
Van Leijenhorst, L., Zanolie, K., Van Meel, C. S., Westen-berg, P. M., Rombouts, S. A., & Crone, E. A. (2010).What motivates the adolescent? Brain regions mediatingreward sensitivity across adolescence. Cerebral Cortex,20(1), 61–69.
Veerman, J. W., Straathof, M. A. E., Treffers, P. D. A., vanden Bergh, B. R. H., & Ten Brink, L. T. (1997). Dutchtranslation of the Self Perception Profile for Children.Lisse, The Netherlands: Swets and Zeitlinger.
Vul, E., Harris, C., Winkielman, P., & Pashler, H. (2009).Puzzling high correlations in fMRI studies of emotion,personality, and social cognition. Perspectives on Psy-chological Science, 4, 274–290.
Wager, T. D., Davidson, M. L., Hughes, B. L., Lindquist, M.A., & Ochsner, K. N. (2008). Prefrontal–subcortical path-ways mediating successful emotion regulation. Neuron,59(6), 1037–1050.
Yurgelun-Todd, D. A., & Killgore, W. D. S. (2006). Fear-related activity in the prefrontal cortex increases withage during adolescence: A preliminary fMRI study. Neu-roscience Letters, 406(3), 194–199.
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
NEURAL CORRELATES OF PEER EVALUATION 17
SUPPLEMENTARY TABLE 1 Coordinates for the contrast YES > NO judgments for each age group, uncorrected for multiple comparisons
(p < .001, minimal 10 contiguous voxels)
Anatomical region L/R Voxels Z-value
MNI coordinates
x y z
19–25 year oldsPrefrontal cortex VMPFC R 159 4.11 9 48 –6
Subcallosal cortex/VMPFC L 22 3.75 –6 21 –15Cingulate cortex Middle cingulate cortex L 215 5.09 –3 –45 39
Anterior cingulate cortex L 42 3.62 –6 45 9Basal ganglia Putamen R 16 3.73 33 0 0Temporal lobe Amygdala R 28 3.78 36 3 –27
Hippocampus L 15 4.34 –21 –12 –21Parahippocampal gyrus R 17 4.07 21 –12 –24Middle temporal gyrus R 121 4.14 45 –60 0Inferior temporal gyrus L 18 3.60 –54 –45 –18
Parietal lobe Superior parietal gyrus R 40 4.01 18 –54 60Lingual gyrus L 10 3.26 –6 –57 3
Occipital lobe Middle occipital gyrus R 26 3.81 42 –78 24Angular gyrus L 16 3.33 –42 –72 30
16–17 year oldsPrefrontal cortex VMPFC R 39 3.76 9 27 –9
VMPFC R 14 3.42 3 63 12Middle orbital gyrus R 35 3.95 12 57 –12Inferior frontal gyrus L 21 3.81 –33 33 9
Prefrontal cortex Subcallosal cortex L 64 4.35 –15 18 –12Inferior frontal gyrus L 32 4.01 –30 21 15
R 37 3.87 24 18 –21Cingulate cortex Paracingulate cortex L 26 4.13 –15 30 27Basal ganglia Putamen L 10 3.49 –24 3 0
Insula R 10 3.53 30 21 –6Temporal lobe Parahippocampal gyrus R 12 4.06 33 –15 –30Occipital lobe Occipital fusiform gyrus L 28 4.02 –39 –51 –15
R 18 3.95 30 –63 –15
Supplementary Figure 1. Correlations between resistance to peer influence (RPI) in the left putamen (−24, 3, 0) and the left lateral PFC(−30, 21, 15) for the contrast rejection > acceptance following a “no” judgment across all participants.
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
NEURAL CORRELATES OF PEER EVALUATION 21
Supplementary Figure 2. Correlations between levels of social anxiety and activity in spherical ROIs for the contrast rejection > acceptancefollowing a “no” judgment across 8–17-year-olds. Spherical ROIs for the left subcallosal cortex/OFC (−15, 18, −12), left paracingulate cortex(–15, 30, 27), left putamen (–24, 3, 0), right OFC (24, 18, –21), and left lateral PFC (–30, 21, 15).
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
22 GUNTHER MOOR ET AL.
Supplementary Figure 3. Correlations between self-perceived global self-worth in the left subcallosal cortex/OFC (–15, 18, –12) and leftputamen (–24, 3, 0) for the contrast rejection > acceptance following a “no” judgment across 8–17 year olds.
Downloaded By: [Crone, Eveline A.] At: 08:09 17 August 2010
