Isolating N400 as neural marker of vocal anger processing in 6-11-year old children

(1)

Article

Isolating N400 as neural marker of vocal anger

processing in 6-11-year old children

Chronaki, G, Broyd, S, Garner, M, Hadwin, J, Thompson, M and

Sonuga-Barke, E

Available at http://clok.uclan.ac.uk/14179/

Chronaki, G, Broyd, S, Garner, M, Hadwin, J, Thompson, M and Sonuga-Barke, E (2011)

Isolating N400 as neural marker of vocal anger processing in 6-11-year old children.

Developmental Cognitive Neuroscience, 2 (2). pp. 268-276. ISSN 1878-9293

It is advisable to refer to the publisher’s version if you intend to cite from the work.

http://dx.doi.org/10.1016/j.dcn.2011.11.007

For more information about UCLan’s research in this area go to

http://www.uclan.ac.uk/researchgroups/

and search for <name of research Group>.

For information about Research generally at UCLan please go to

http://www.uclan.ac.uk/research/

All outputs in CLoK are protected by Intellectual Property Rights law, including

Copyright law. Copyright, IPR and Moral Rights for the works on this site are retained

by the individual authors and/or other copyright owners. Terms and conditions for use

of this material are defined in the

http://clok.uclan.ac.uk/policies/

CLoK

Central Lancashire online Knowledge

(2)

ContentslistsavailableatSciVerseScienceDirect

Developmental

Cognitive

Neuroscience

jou rn a l h om epa g e:h t t p : / / w w w . e l s e v i e r . c o m / l o c a te / d c n

Isolating

N400

as

neural

marker

of

vocal

anger

processing

in

6–11-year

old

children

Georgia

Chronaki

a,∗

_,

_Samantha

_Broyd

a

_,

_Matthew

_Garner

c

_,

_Julie

_A.

_Hadwin

a

_,

Margaret

J.J.

Thompson

a

_,

_Edmund

_J.S.

_Sonuga-Barke

a,b,∗∗

a_{Developmental}_Brain_Behaviour_Laboratory,_Psychology,_University_of_Southampton,_UK b_Department_of_Experimental_Clinical_and_Health_Psychology,_Ghent_University,_Belgium c_Psychology_&_Division_of_Clinical_{Neuroscience,}_School_of_Medicine,_University_of_Southampton,_UK

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received15August2011

Receivedinrevisedform28October2011 Accepted25November2011 Keywords: Vocal ERP N400 Children

a

b

s

t

r

a

c

t

Introduction:Vocalangerisasalientsocialsignalservingadaptivefunctionsintypicalchild development.Despiterecentadvancesinthedevelopmentalneuroscienceofemotion pro-cessingwithregardtovisualstimuli,littleremainsknownabouttheneuralcorrelatesof vocalangerprocessinginchildhood.Thisstudyrepresentsthefirstattempttoisolatea neuralmarkerofvocalangerprocessinginchildrenusingelectrophysiologicalmethods.

Methods:WecomparedERPwaveformsduringtheprocessingofnon-wordemotional vocalstimuliinapopulationsampleof556–11-year-oldtypicallydevelopingchildren. Childrenlistenedtothreetypesofstimuliexpressingangry,happy,andneutralprosody andcompletedanemotionidentificationtaskwiththreeresponseoptions(angry,happy andneutral/‘ok’).

Results:AdistinctiveN400componentwhichwasmodulatedbyemotionalcontentofvocal stimuluswasobservedinchildrenoverparietalandoccipitalscalpregions—amplitudes weresignificantlyattenuatedtoangrycomparedtohappyandneutralvoices.

Discussion:FindingsofthepresentstudyregardingtheN400arecompatiblewithadult studiesshowingreducedN400amplitudestonegativecomparedtoneutralemotional stim-uli.Implicationsforstudiesoftheneuralbasisofvocalangerprocessinginchildrenare discussed.

1. Introduction

Human voicesareauditorystimuli processedin spe-cialisedbrainregionsinhealthyadults(Belinetal.,2000),

∗ Correspondingauthorat:UniversityofSouthampton,Developmental Brain-BehaviourLaboratory,Psychology,HighfieldCampus,Southampton SO171BJ,UK.Tel.:+4402380594586;fax:+4402380594597.

∗∗ Correspondingauthorat:UniversityofSouthampton,Developmental Brain-BehaviourLaboratory,Psychology,HighfieldCampus,Southampton SO171BJ,UKandDepartmentofExperimentalClinicalandHealth Psy-chology,GhentUniversity,Belgium.Tel.:+4402380595785; fax:+4402380594597.

E-mailaddresses:[email protected] (G.Chronaki), [email protected]

(E.J.S.Sonuga-Barke).

aswellasinfants(Grossmannetal.,2010)and 4–5-year-oldchildren(Rogieretal.,2010).Furtherresearchsuggests thatthehumanbrainbeginstobecomesensitiveto emo-tionalinformationexpressedinvocaltoneatanearlystage indevelopment(Grossmannetal.,2010,2005).The abil-itytoreadothers’non-verbalsocial signalsisarguedto allowhumanstonavigatetheirsocialenvironment suc-cessfullyandmodifytheirbehaviouraccordingtoothers’ feelings and intentions (Frith and Frith, 2007). Further-more,sensitivitytoothers’vocalemotionalexpressionshas beenassociatedwithsocialcompetenceinchildhoodand adolescence(GoodfellowandNowicki,2009;Rothmanand Nowicki,2004;TrentacostaandFine,2010).

Emotionalprosody(orspeechmelody)referstochanges intheintonationofthevoiceaccordingtothespeaker’s

Open access under CC BY-NC-ND license.

(3)

emotionalstate(BanseandScherer,1996;Hargrove,1997). Accordingtorecentresearch,wordswithanangryprosody, forexample,elicitedmorenegativeERPresponses(latency 450ms) over fronto-central sites compared to words with happy or neutral prosody in 7-month-old infants (Grossmannetal.,2005).Furtherstudiesshowedincreased brainactivation patternsin theposterior temporal cor-textoangrycomparedtoneutralandhappyprosodyin 7-month-oldinfants(Grossmannet al.,2010).Similarly, functionalMagneticResonanceImaging(fMRI)studiesin adultshaveshownthatangryrelativetoneutralprosody elicitedenhancedresponsesinmiddlesuperiortemporal sulcus(Grandjeanetal.,2005;Sanderetal.,2005).In con-trast,happyvoices(andnotangryorneutralvoices)have beenfoundtoevokeanincreasedresponseinright infe-riorfrontalcortexin7-month-oldinfants(Grossmannetal., 2010)andadults(Johnstoneetal.,2006).

Severalresearchstudieshaveshownthatchildrencan reliablyrecogniseangeronthebasisofvocaltone(Baum and Nowicki, 1998; Nowicki and Mitchell, 1998; Tonks etal.,2007),howeverlittleisknownabouthowthis pro-cesschanges withdevelopment.TheN100isanegative componentoccurringatabout100mspoststimulusand isthoughttoreflectsounddetectionandearlyattentional orienting(Bruneauetal.,1997;NäätänenandPicton,1987). The N100is sensitiveto differences in sound intensity orfrequencyofacousticstimuli (Ceponieneetal.,2002; NäätänenandPicton,1987).In4–8-year-olds, theN100 peaksaround170msattemporalsites,forexample,andis thoughttoreflectauditorybehaviouralorienting(Bruneau et al., 1997).A recent study showed noeffect of vocal emotion(angry,happy)ontheamplitudeandlatencyof anearly(i.e.,100–200ms)negativecomponentduringa passivelisteningtask in13 healthy 9–11-year-old boys (Korpilahtietal.,2007).However,previousresearchhas typicallyemployedasmallsamplesizeandutilisedspeech stimuli(i.e., ‘giveit tome’; seeKorpilahti etal., 2007); makingitdifficulttodisentanglesensitivitytoemotional prosodyfromsensitivitytothelanguagecontent.In the currentstudyweisolatedtheN100suggestedtoreflect earlyauditoryattentionorienting(Bruneauetal.,1997).

Inaddition, weexplored theN400;a centro/parietal negativecomponentoccurringatabout400msafter stim-ulusonset.TheN400istypicallylargerforwordspresented inincongruousthancongruoussentencecontexts(i.e., cof-feewithsocks/sugar)(KutasandHillyard,1980).Schirmer etal.(2005)presentedadultlistenersprimesentences(i.e., ‘shehadherexam’)withhappyorsadprosodyfollowed byvisualtargetwords.Halfthetargetwordsmatchedthe prosodyoftheprecedingsentence(i.e.,‘success’ follow-inghappyprosody)andhalfdidnot(i.e.,‘failure’following happyprosody).Participantsindicated whetherthe tar-getwordmatchedtheprosodyoftheprecedingsentence. Resultsshowedthattargetwords(i.e.,‘success’)elicited a larger N400 when preceded by a prime with incon-gruous(i.e.,sad)ascomparedtocongruous(i.e.,happy) emotionalprosody (Schirmeret al.,2005).This demon-stratessensitivityoftheN400componenttoanemotional prosodic context. In a similar study, Bostanov and Kotchoubey(2004)presenteddifferentword-exclamation pairsusingspokenemotionwordsasprimes(i.e.,‘joy’)and

corresponding emotional exclamations as targets (i.e., grief)withconsistent(i.e.,positive–positive)and inconsis-tent(i.e.,positive–negative)combinations.Resultsshowed that N400 amplitude was significantly larger for the inconsistentcondition.Insummary,theN400isthought to reflect cognitive processes related to context viola-tionduring emotionrecognitionandcanprovea useful electrophysiologicalmarkerofemotionalprosody compre-hension.

Insummary,previousresearchhasfocusedontheN400 tovocalemotioninadultsandincongruityeffectsfollowing aprime,however,thetimecourse(ERPs)ofvocalanger processinginmiddlechildhoodremainsunexplored. 1.1. Thecurrentstudy

The goal of the current study was to isolate the neural signature of anger prosody processing in chil-dren. To this end, we focused on N100, a putative markerofearlyauditoryprocessing,andtheN400,alate emotion-specificprocessingcomponent.Following previ-ousresearch(BostanovandKotchoubey,2004;Schirmer etal.,2005;ToivonenandRämä,2009)wepredictedthat theN400componentwouldbemodulatedbyemotional prosody(i.e.,angervs.neutralstimuli).Second,following previousfindings(i.e.,Korpilahtietal.,2007)wepredicted thattheN100wouldbeinsensitivetotheemotional con-tentofthesoundsandsobeinsensitivetoangerrelated stimuliin children.WefocusedontheN100toruleout thatpredictedN400effectswerenotdrivenbypervasive problemsinauditoryorienting.

2. Methods 2.1. Participants

Of the105children initially approachedvia primary schools,80wererecruitedtothestudy.Childrenandtheir mothersgaveinformedwrittenconsentforparticipation. Ofthose,pilotdatafrom5children(meanage=7.30years, SD=.73,agerange6.08–8 years,2 boys)wereexcluded fromanalysesduetoincompletedataand ERPartifacts. Oneboy(5.42years)withahearingthresholdinthe atypi-calrange(46dB;seebelow)intherightearwasexcluded.In theend,completeERPandbehaviouraldatawereavailable from55children(meanage=8.90years,SD=1.60 years, agerange6.08–11.83,39boys).Thestudywasapproved bytheUniversityofSouthampton,SchoolofPsychology EthicsCommittee.

2.2. Vocalexpressionstimuli

Three types of vocal stimuli expressed emotional prosody(interjection‘ah’).Onestimulusexpressedanger, one happiness, both at a highintensity.There wasalso a neutralexpression (Maurage etal., 2007).1 _All_stimuli derivedfromanadultfemalemodelandwerestandardised

1_Further_information_about_the_vocal_stimuli_is_provided_in

(4)

Table1

Duration,fundamentalfrequency−f0(inHz)andintensity(indB)ofthevocalstimuli.

Emotion Meanf0 Minf0 Maxf0 MeandB MindB MaxdB

Angry 294.85 79.07 355.74 76.85 63.43 81.92

Happy 350.31 221.59 525.85 76.50 68.36 83.42

Neutral 191.30 181.21 194.54 76.34 70.30 78.31

foracousticparametersincludingmeanintensity(76dB)

leading to a correspondent Sound Pressure Level (SPL)

of 0.13Pa, duration (700ms), recording frequency

(16,000Hz),andriseandfallramptimes(20ms).Acoustic

analyseswereconductedusingPraatsound-analysis

soft-ware(BoersmaandWeenink,2009).Acousticproperties ofthevocalstimuliarepresentedinTable1.Vocalstimuli werepresentedbinaurallyviaheadphones.Thesestimuli are wellvalidated in adults(Maurage et al.,2007).The battery hashigh internal consistency for each emotion setandhighlevelsofspecificity(independencebetween theratingsinthedifferentemotionsets)(Maurageetal., 2007).

We conducted a behavioural validation studyof the stimuliinaseparatecommunitysampleof65 6–11-year-oldchildren (meanage=8.31years,SD=1.55, agerange 6.00–10.75years,31boys).Thesechildrengaveinformed written consenttoparticipate outof a total of 97 chil-dreninitiallyapproachedvialocalprimaryschools.After listeningtoonevocalstimulusata time(Angry, Happy and Neutral, 12 trialsper emotion type),children were askedtoidentifytheemotioninthevoiceandpressoneof thethreekeyboardbuttonswiththelabels‘angry’,‘happy’ or ‘okay’ toindicate theirresponse. Themean percent-age of trials classified correctly was as follows: Angry: M=82.43,SD=28.34,Happy: M=71.28,SD=33.17, Neu-tral:40.14,SD=21.99.Accuracywasabove-chanceforall emotion types, withchance definedas 33.3% given the threeresponseoptions(dataavailablefromtheauthors). Therewerenosignificantassociations betweenageand recognitionaccuracyforhappy(r=.19,p>.05)andneutral (r=−.09,p>.05)voicesandamarginalsignificant associ-ationbetweenageandaccuracyforangryvoices(r=.26, p<.05).

2.3. Experimentalparadigmandprocedure

At the beginning of the experimental session, pure toneaudiometrictestingwasconductedwithastandard clinicalaudiometertoestablishwhetherparticipants’ hear-ing threshold was withinthe average range defined as 25dB (see British Society of Audiology Recommended Procedures, 2004). Childrenwere instructedto indicate witha buttonpress when a test tonewas present and whenitceasedtobepresent.Thisprocedurewasrepeated for each ear (Right/Left) and for three frequency levels (1000,1500,500Hz)separately.Eachparticipant’s hear-ingthresholdwasdefinedasthelowestlevelofsoundthey couldhear.Anaverageofthethresholdsfromthethree frequencies wasderived in each ear. Anaverageof the thresholdsfrombothearswasfurthercreated.

Subsequently,childrenparticipatedinanemotion iden-tificationtaskwiththreeresponseoptions(angry,happy

andneutral/‘ok’).Afterlisteningtoonevocalstimulusat atime,childrenwereinstructedtoidentifytheemotion inthevoiceandpressoneofthethreekeyboardbuttons withthelabels‘angry’,‘happy’or‘okay’toindicatetheir response.Therewere180experimentaltrials(60trialsper emotiontype)presentedintwoblocksof90trialseach. Therewasa5-minrestbreakinbetweenthetwoblocks. Childrenparticipatedin12practicetrials(four presenta-tionsofeachemotion)atthebeginningofthetask.Button pressresponseswereloggedonthecomputervia Presen-tationsoftware.Eachtrialbeganwiththepresentationof acentralfixationcross(500ms)followedbythe presen-tationofthestimulus,followedbyablankscreenuntilthe participantsgavearesponseanda1000msinter-trial inter-val(ITI).Emotionstimuluspresentationwasrandomised across participants. The overall testing time, including audiometricassessment,capfittingandexperimental test-ing,wasoneandahalfhours.

2.4. Electrophysiologicalrecordingandprocessing

EEGdatawererecordedfromanelectrodecap (Easy-cap, Herrsching, Germany) containing 66 equidistant silver/silverchloride(Ag/AgCl)electrodesusingNeuroscan Synamps2₇₀_channel_EEG_system._Cap_electrodes_were ref-erencedtothenose.TheEEGdataweresampledat250Hz withabandpassfilterat0.1–70HzusinganACprocedure andrecordedfrom19sites(seeFig.1).Analysesfocused on15sitesatcentral(sites1,2,4,6,10,16),parietal(sites 12,13,14,24,26)andoccipital(sites37,38,39,40)areas. Selectionofthesesiteswasliteratureinformed(Schirmer andKotz,2006)andaimedatmaximisingthenumberof artifact-freeepochs. Agroundelectrode wasfitted mid-waybetweentheelectrodeatthevertexandfrontalsite32. Verticalelectro-oculogram(vEOG)wasrecordedfromfour electrodes: two bipolarelectrodes were placed directly beneaththeleftandrighteyesandaffixedwithtape,while thetwoelectrodesplacedabovetherightandlefteyewere includedwithintheelectrodecap.ImpedancesforvEOG, referenceandcapelectrodeswerekeptbelow5k.The ERPepochwasdefinedas100mspre-stimulusto1000ms post-stimulus.Eachepochhadabaselineof100msof pre-stimulusactivityand wasfilteredwitha low-passfilter down48dBat32Hz.Anocularartifactreductionprocedure (Semlitschetal.,1986)basedonvEOGactivitywasusedto removetheinfluenceofblinkandothereyemovement; epochswererejectedifamplitudesexceeded±150␮Vin

anyEOGorscalpsiteincludedinanalysesorifparticipants respondedincorrectly.AverageERPswerecalculatedfor eachemotiontype(Angry,Happy,Neutral).Aminimumof 20artifactfreeepochsoutofatotalof60epochsforeach emotiontypewereusedforcalculatingERPaverages.

(5)

Fig.1.GrandmeanERPstoangry,happyandneutralvoicesperscalpregion.Amplitude(␮V)andtime(ms)aremarkedatallregionswithapre-stimulus baselineof−100ms.Scaleis−3to+12␮V.Ontheleft,below,themontagewiththesitesusedinEEGrecording(inbluethesitesusedinEEGrecording butnotinanalyses).

Themeanand SD ofthenumber ofepochs included for each emotion condition in theoverall sample were asfollows:Angry:M=45.74,SD=9.32,Happy:M=45.40, SD=9.50, Neutral: M=45.99,SD=8.91. Given theyoung ageofthechildren,weexaminedpotentialeffectsofage onartifactandtrialrejectionbydividingoursampleinto younger (6.08–8.75years) and older (8.83–11.83years) childrenusingamediansplit.Therewasasignificanteffect of ageonthe number ofcorrect and artifact free trials includedinanalysesforhappy(F(1,54)=6.84,p<.05)and neutral(F(1,54)=6.98,p<.05)voicesbutnotforangry(F (1,54)=3.41,p>.05),withfewertrialsforyoungerchildren (Angry:M=43.35,SD=9.40, Happy:M=42.04,SD=9.60, Neutral:M=42.80,SD=8.90)comparedtoolderchildren (Angry:M=47.90,SD=8.87, Happy:M=48.41,SD=8.47, Neutral:M=48.82,SD=8.00).Ageeffectsinourstudyare similar to age effects reported in other developmental

ERPstudieswithchildrenofsimilarages(Vlamingsetal., 2010).

A baseline-to-peakmeanamplitudemethodwas fol-lowedfortheN100(90–180ms)andN400(380–500ms). Theabovetimewindowswereselectedbecausetheybest capturedeachERPcomponentidentifiedbyvisual inspec-tion. Mean amplitude was initially calculated for each individualsiteandsubsequentlythemeanamplitudefor eachERPcomponentwascalculatedasacombinedscore foranumber ofdefinedgroupsofelectrode sites(‘scalp regions’-seeFig.1)toincreasethereliabilityof measure-ment (Dien and Santuzzi, 2005). Selection of electrode groups was basedon thestrongstatistical similarity of the grandaverageERPs for each electrode. Correlations between ERP waveforms within each region were sta-tisticallystronger (Pearson’sr=from.63 to.95,p<.001) comparedtocorrelationsbetweenamplitudesofelectrode

(6)

Table2

Mean percentage (SD) of trials classified correctly (in bold) and misattributions.

Vocalexpression Childresponse

Angry Happy Neutral

Angry 92.30(9.87) 2.24(3.72) 3.21(3.65) Happy 2.72(4.16) 91.33(9.98) 3.42(6.72) Neutral 3.12(5.68) 5.81(7.92) 91.12(10.37)

sitesbelongingtodifferentregions(Pearson’sr=from.25

to.64,p<.01).

3. Dataanalysis

3.1. Performancedata

Rawdataweretransformedintomeasuresofaccuracy

accordingtothetwo-high-thresholdmodel.

Discrimina-tionaccuracy(Pr)wascomputedforeachtargetemotion

using ‘hits’ (i.e., number of happy, angry or neutral

expressions classified correctly) (Corwin, 1994). Values

ofdiscriminationaccuracy(Pr)wereenteredinrepeated measuresnon-parametricFriedman’sANOVAwith emo-tion(Angry,HappyandNeutral)aswithin-subjectfactor andpairedWilcoxonfollowuptests.

3.2. ERPdata

Preliminary correlational analyses examined associ-ations between child age and N100 and N400 mean amplitudesatthethreescalpregions.Also, independent-samplest-testsexamineddifferencesinN100andN400 meanamplitudevaluesbetweenmalesandfemales,given gendereffectsonemotionalprosodyperceptionreported intheliterature(Schirmeretal.,2002).Themainanalyses examinedthemaineffectofemotiontypeonamplitude ofN100andN400componentineachscalpregion.N100 and N400ERP datawere entered intoa repeated mea-suresAnalysisofVariance(ANOVA)withemotiontypeas thewithin-subjectfactor.Becausethefocusofthepresent studywasontheERPcorrelatesofvocalanger process-ing,simpleplannedcontrastscompared theangryvoice conditionwiththeneutralandhappyvoicecondition. 4. Results

4.1. Performance

Inordertooptimisethechancesofemotional modula-tionofERPcomponentsthevocalstimuliweredesigned tobeeasytoclassify.Aspredicted,therefore,mean accu-racyforallthreeemotionswashigh(M=91.60%,SD=8.68%, see Table 2). Therewas a significant effect of emotion type onaccuracy (x2 ₍₂₎₌_12.91,_p_<_.001). _Accuracy_was significantlyhigherforangrycomparedtoneutralvoices (T=364.50,p<.001,r=−.51).Non-parametricSpearman’s correlationsshowedthatchildren’smeanhearing thresh-oldwasnotsignificantlyassociatedwithaccuracyforangry or happyvoices(p>.05)and itwassignificantly associ-atedwithaccuracyforneutralvoices(rs=−.28,p=.036).

Childs’agewasnotsignificantlyassociatedwithaccuracy forangry,happyorneutralvoices(ps>.05).

4.2. ERPdata

Fig.1plotsthegrandmeanaveragesforthewaveforms toeachvocalemotionstimuliforthethreeregionsof inter-est.Fig.2 plots thegrandmean averages toeach vocal emotionperindividualsite(seeAppendixB).DistinctN100 andN400componentscouldbeseenatthecentral, pari-etalandoccipitalregions.Onthewhole,neitheragenor genderwasassociatedwithERPsamplitudeatanyregion foreachcondition(intherangeofPearson’srfrom.02to −.22,p>.05).Theexceptionwasforamplitudestoneutral voiceswhichwerepositivelyassociatedwithchildagefor theoccipitalN400(r=.28, p=.034). Thus,analyseswere repeatedforthisregionandcomponentincludingchilds’ ageinthestatisticalmodelasacovariate.

Themainanalysesshowedlittleevidenceofmodulation ofN100byemotionaltoneofstimuli.Resultsarereported withaBonferronicorrectionformultiplecomparisonswith anacceptedalphaof.05/12=.004.Therewasnoeffectof emotiontypeonN100amplitudesfortheparietalandthe occipitalregions(F(2,54)<4.04,p>05).Therewasaneffect ofemotiononcentralN100amplitude(F(2,108)=5.41, p<01)withangryvoiceshavingsmalleramplitudesthan happy (F (1, 54)=9.60, p<01, Angry: M=1.60, SE=.44, Happy: M=−.27, SE=.62) and neutral (F (1, 54)=9.97, p<01,Neutral:M=−.26,SE=.55)voices.Thiseffectdidnot survivecorrectionformultipletesting.Nootherdifference reachedsignificance(p>.05).

In contrast,there wasa consistentpatternof effects ofemotionalprosodyontheN400acrossdomainswhich wasmostobviousinoccipital(F(2,108)=10.62,p<001) and parietal (F (2, 108)=11.83, p<001) regions (see

Fig.1), althoughstill significantin thecentralregion(F (2,108)=6.70,p<01).Inparticular,theN400was atten-uated to angry compared to happy and neutral voices inallregions(p<.001),althoughtheangry-neutral con-trastwasmarginallysignificant(p=.040)intheoccipital region.Afterincludingchildageinthestatisticalmodel asacovariate,thedifferenceinoccipitalN400amplitude betweenangryand happywasrendered non-significant while the difference betweenangry and neutral voices becamemarginallysignificant(p=.055)anddidnot sur-viveBonferroni correction.Allotheremotion effects on theN400reportedabovesurvivedBonferronicorrection formultiplecomparisons.Wealsoexaminedtheeffectof ageonERPs(N100andN400)bydividingoursampleof 6–11-yearoldsintoyounger(6.08–8.75years)andolder (8.83–11.83years) children usinga median split.When running the analyses separately for older and younger children,theeffectsofemotionontheN400remained sig-nificantinbothagegroups(p<.005)andresultsregarding theN100didnotchange.

5. Discussion

Theaimofthepresentstudywastoisolatethe neu-ralcorrelatesofvocalangerprocessinginchildren.Based onpreviousresearchinadults(BostanovandKotchoubey,

(7)

G. Chronaki et al. / Developmental Cognitive Neuroscience 2 (2012) 268– 276 273 site 1 central time (ms) 700 600 500 400 300 200 100 0 -100 Amplitude -10 -5 0 5 10

site 2 central site 4 central

site 10 central site 12 parietal site 13 parietal site 14 parietal

site 16 central site 24 parietal site 26 parietal site 37 occipital

site 38 occipital site 39 occipital site 40 occipital

(8)

2004;Schirmeretal.,2005;ToivonenandRämä,2009)we identifiedtheN400,asimplicatedinvocalemotion pro-cessing.Inlinewithpreviousresearch,theresultsshowed thattheN400componentwaseasilydistinguishedonthe basis ofgrandmeanaverages,especially intheoccipital andparietalregions(Schirmerand Kotz,2006). Further-more,this componentwasmodulated bytheemotional toneoftheprosody;itwasattenuatedforangrycompared to happyand neutral voicesacross regions. In contrast totheN400,theN100didnotshowastrongpatternof emotioneffects.TheN100isthoughttoreflectearly atten-tionalorientingtoauditorystimuli(Bruneauetal.,1997; NäätänenandPicton,1987),ratherthanbeingan ‘emotion-specific’component.Findingsofthepresentstudysuggest thatlattercomponents(N400)ratherthanearliersensory processingmechanisms(N100)underlieemotionalvoice processingin6–11-yearolds.Findingsofthisstudy regard-ingtheN400arecompatiblewithadultresearchshowing reducedN400amplitudestonegativecompared to neu-tralemotionalstimuli(KanskeandKotz,2007).Thefinding ofreducedN400amplitudestonegativestimulihasbeen replicatedinanumberofrecentstudiesinhealthyadults (DePascalisetal.,2009;Gootjesetal.,2011;Stewartetal., 2010).ReducedN400amplitudetonegativestimulihas beensuggestedtoreflectfacilitatedprocessingofnegative comparedtoneutralwordsinhealthyadults(Kanskeand Kotz,2007;Stewartetal.,2010).

Thecurrentstudyhadanumberofstrengthscompared withpreviousstudies.First,thepresentstudyusedalarger sample(N=55)incontrasttopreviousstudieswithsmall samplesandrange.Second,incontrasttoprevious inves-tigationswithwordstimuli,thisstudyadoptednon-word stimuliandthesestimuliallowedustoexaminetheneural correlates of prosody processing independently of lan-guageprocessing.Third,thisstudyusedanexplicitemotion identification taskwhile previousstudies withchildren employedsimplepassivelistening(Korpilahtietal.,2007) or complex multimodal tasks (Shackman et al., 2007). Fourth,itincludedneutralvoiceasacontrolconditionfor comparison withemotionalstimuli. Finally,wetargeted bothearly(N100)andlate(N400)componentsofvoice pro-cessing.Therefore,theangermodulationoftheN400inthe presentagegroupisaninterestingandnovelfindingina vocalemotionrecognitiontask.Thesefindingsextendthe literatureontheN400asanelectrophysiologicalmarkerof emotionalprosodycomprehensionbydemonstrating emo-tionmodulationoftheN400.PreviousresearchontheN400 hasmainlyfocusedonadults(BostanovandKotchoubey, 2004;Kotchoubeyetal.,2009;Schirmeretal.,2005)and infants(Grossmannetal.,2005).Thisisthefirststudyto isolatetheN400asaneuralmarkerofvocalanger process-ingin6–11-year-oldchildren.Incontrast,theN100wasnot sensitivetotheemotionalcontentofangerrelatedvocal stimuliinchildren(Korpilahtietal.,2007).

Thestudyalsohadsomelimitations. First,thisstudy adopted a baseline-to-peak amplitude method, recom-mended when the component under analysis does not haveadefinitepeakaswasthecaseinthisstudy(Fabiani etal.,2007).Adisadvantageofthismethod,however,is that it may be sensitiveto noise or nonlinear fluctua-tionsinthebaselinetimewindow,relativetoother(i.e.,

peak-to-trough)measures(Pictonetal.,2000).Inaddition, oneshouldacknowledgethelimitations oftheERP sig-nalincasesinwhichthecomponentofinterestappears tobesuperimposedonaslowdriftorslowfrequency(i.e., delta)oscillation(seeparietalN400,Fig.1).Recentresearch suggeststhatemotionprocessingmayberelatedtodelta frequencyactivity.Forexample,unpleasantpictures pro-vokedgreaterdelta responsesthan pleasantpictures in healthy adults (Klados et al., 2009). The study didnot include‘control’soundcategoriesbeyondhumanvoicesto establishtheselectivityofneuralresponsestovocal sig-nals.Thisadditionalelementwouldmorerobustlyclarify whethertheeffectsobservedreflectedvocalanger process-ingortheacousticaldifferencesbetweensoundcategories. Thefindingscannotbegeneralizedtoallangerstimuli,such asforexamplehigherintensityangerstimuli.Such conclu-sionscanonlybedrawnwiththeinclusioninthestudy designofvaryingintensitylevelsofangerstimuli.

Futureresearchshouldreplicatethepresentfindingsin alargersampleoftypicallydevelopingchildren.In addi-tion,animportant avenuefor futureresearchwould be toexploreassociationsbetweenneuralmarkersofvocal angerprocessingandchildhoodbehaviourproblems. Chil-drenwho display externalisingsymptoms (i.e., conduct problems)arelessaccurateatrecognisingnegativevocal expressions (Stevens et al., 2001) and display a selec-tiveperceptualbiastovocalangerinparticular(Manassis et al., 2007). These children are especially susceptible toanger andfrustration (Silket al.,2003;Morris et al., 2002;Neumannetal.,2011)and tendtodisplayhostile behaviours following threat interpretations of ambigu-ousstimuli(Barrettetal.,1996).Neurobiologicalmodels viewimpairmentsinbrainsystemsthatfunctionto reg-ulateangerasakeymechanismforimpulsiveaggressive behaviour(Davidsonetal.,2000).Despiterecentevidence onatypicalbrainreactionstoaffectivespeechprosodyin childrenwithAspergerSyndrome(Korpilahtietal.,2007), researchonthetimecourse(ERPs)ofvocalanger process-inginchildrenwithexternalisingproblemsislimitedand wouldbeapromisingavenueforfutureresearch. Conflictofinterest

Theauthorsdeclarethefollowingconflictofinterest: ProfessorEdmundSonuga-Barke:Recentspeakerboard: Shire, UCB Pharma, Current & recent consultancy: UCB Pharma,Shire,Current&recentresearchsupport:Janssen Cilag,Shire,Qbtech,FlynnPharma,AdvisoryBoard:Shire, FlynnPharma,UCBPharma,AstraZeneca,Conference sup-port:Shire.Dr.MargaretThompson:Recentspeakerboard: JanssenCilagCurrent&recentconsultancy:Shire,Current &recentresearchsupport:JanssenCilag,Shire,Advisory Board:Shire.

Acknowledgements

Theauthorswishtothanktheparentsandchildrenwho tookpartinthisresearchandDrPierreMaurageforthe helpfuladviceonthevocalstimuli.Theresearchwas sup-portedbyPsychology,UniversityofSouthampton.

(9)

AppendixA. Supplementarydata

Supplementary data associated with this arti-cle can be found, in the online version, at

doi:10.1016/j.dcn.2011.11.007.

References

Banse, R., Scherer, K.R., 1996. Acoustic profiles in vocal emotion expression.Journal of Personality and SocialPsychology 70 (3), 614–636.

Barrett,P.M.,Rapee,R.M.,Dadds,M.R.,Sharon,M.,1996.Family enhance-mentofcognitivestyleinanxiousandaggressivechildren.Journalof AbnormalChildPsychology24(2),187–203.

Baum,K.,Nowicki,S.,1998.Perceptionofemotion:measuring decod-ingaccuracyofadultprosodiccuesvaryinginintensity.Journalof NonverbalBehavior22(2),89–107.

Belin,P.,Zatorre,R.J.,Lafaille,P.,Ahad,P.,Pike,B.,2000.Voice-selective areasinhumanauditorycortex.Nature403(6767),309–312. Boersma,P.,Weenink,D.,2009.Praat:DoingPhonetics byComputer

(Version5.1.04) [Computer Program], Retrieved December2009 http://www.praat.org/.

Bostanov,V.,Kotchoubey,B.,2004.Recognitionofaffectiveprosody: continuouswaveletmeasuresofevent-relatedbrainpotentialsto emotionalexclamations.Psychophysiology41(2),259–268. BritishSocietyofAudiologyRecommendedProcedure(2004).Puretone

air and bone conduction threshold audiometry with and with-outmaskinganddeterminationofuncomfortableloudnesslevels. BSA.

Bruneau,N.,Roux,S.,Guerin,P.,Barthelemy,C.,Lelord,G.,1997.Temporal prominenceofauditoryevokedpotentials(N1wave)in4–8-year-old children.Psychophysiology34(1),32–38.

Ceponiene,R.,Rinne,T.,Näätänen,R.,2002.Maturationofcorticalsound processingasindexedbyevent-relatedpotentials.Clinical Neuro-physiology113(6),870–882.

Corwin,J.,1994.Onmeasuringdiscriminationandresponsebias:unequal numbersoftargetsanddistractorsandtwoclassesofdistractors. Neu-ropsychology8(1),110–117.

Davidson,R.J.,Putnam,K.M.,Larson,C.L.,2000.Dysfunctioninthe neu-ralcircuitryofemotionregulation—apossiblepreludetoviolence. Science289(5479),591–594.

DePascalis,V.,Arwari,B.,D’Antuono,L.,Cacace,I.,2009.Impulsivityand semantic/emotionalprocessing:anexaminationoftheN400wave. ClinicalNeurophysiology120(1),85–92.

Dien,J.,Santuzzi,A.M.,2005.ApplicationofrepeatedmeasuresANOVAto high-densityERPDatabase:areviewandtutorial.In:Handy,T.C.(Ed.), Event-relatedPotentials.AMethodsHandbook.MITPress,Cambridge, MA.

Fabiani,M.,Gratton,G.,Federmeier,K.D.,2007.Event-related-potentials: methods,theoryandapplications.In:Cacciopo,J.J.,Tassinary,L.G., Berntson,G.G.(Eds.),HandbookofPsychophysiology.,3rded. Cam-bridgeUniversityPress,pp.85–119.

Frith,C.D.,Frith,U.,2007.Socialcognitioninhumans.CurrentBiology17, 724–732.

Goodfellow,S.,Nowicki,S.,2009.Socialadjustment,academicadjustment, andtheabilitytoidentifyemotioninfacialexpressionsof7-year-old children.JournalofGeneticPsychology170(3),234–243.

Gootjes,L.L.C.,Coppens,L.C.,Zwaan,R.A.,Franken,I.H.A.,VanStrien,J.W., 2011.Effectsofrecentwordexposureonemotion–wordStroop inter-ference:anERPstudy.InternationalJournalofPsychophysiology79 (3),356–363.

Grandjean,D.,Sander,D.,Pourtois,G.,Schwartz,S.,Seghier,M.L.,Scherer, K.R.,Vuilleumier,P.,2005.Thevoicesofwrath:brainresponsesto angryprosodyinmeaninglessspeech.NatureNeuroscience8(2), 145–146.

Grossmann,T.,Oberecker,R.,Koch,S.P.,Friederici,A.D.,2010.The devel-opmentaloriginsofvoiceprocessinginthehumanbrain.Neuron65 (6),852–858.

Grossmann,T.,Striano,T.,Friederici,A.D.,2005.Infants’electricbrain responsestoemotionalprosody.Neuroreport16(16),1825–1828. Hargrove,P.M.,1997.Prosodicaspectsoflanguageimpairmentin

chil-dren.TopicsinLanguageDisorders17(4),76–83.

Johnstone,T.,vanReekum,C.M.,Oakes,T.R.,Davidson,R.J.,2006.Thevoice ofemotion:anFMRIstudyofneuralresponsestoangryandhappy vocalexpressions.SocialCognitiveandAffectiveNeuroscience1(3), 242–249.

Kanske, P., Kotz, S.A.,2007. Concreteness in emotional words: ERP evidence from a hemifield study. Brain Research 1148, 138– 148.

Klados,M.A.,Frantzidis,C.,Vivas,A.B.,Papadelis,C.,Lithari,C.,Pappas, C.,Bamidis,P.D.,2009.Aframeworkcombiningdeltaevent-related oscillations(EROs)andsynchronisationeffects(ERD/ERS)tostudy emotionalprocessing.ComputationalIntelligenceandNeuroscience, 1–16.

Korpilahti,P.,Jansson-Verkasalo,E.,Mattila,M.-L.,Kuusikko,S., Suomi-nen,K.,Rytky,S.,Pauls,D.,Moilanen,I.,2007.Processingofaffective speechprosodyisimpairedinaspergersyndrome.JournalofAutism andDevelopmentalDisorders37(8),1539–1549.

Kotchoubey,B.,Kaiser,J.,Bostanov,V.,Lutzenberger,W.,Birbaumer,N., 2009.Recognitionofaffectiveprosodyinbrain-damagedpatientsand healthycontrols:aneurophysiologicalstudyusingEEGand whole-head MEG.Cognitive,Affective,&BehavioralNeuroscience9(2), 153–167.

Kutas, M., Hillyard, S.A., 1980. Event-related brain potentials to semanticallyinappropriateandsurprisinglylargewords.Biological Psychology11,99–116.

Manassis,K.,Tannock,R.,Young,A.,Francis-John,S.,2007.Cognition inanxiouschildrenwithattentiondeficithyperactivitydisorder:a comparisonwithclinicalandnormalchildren.BehavioralandBrain Functions3.

Maurage,P.,Joassin,F.,Philippot,P.,Campanella,S.,2007.Avalidated batteryofvocalemotionalexpressions.NeuropsychologicalTrends 2,63–74.

Morris, A.S.,Silk, J.S., Steinberg, L.,Sessa,F.M., Avenevoli, S.,Essex, M.J.,2002.Tempermentalvulnerabilityandnegativeparentingas interactingofchildadjustment.JournalofMarriageandFamily64, 461–471.

Näätänen,R.,Picton,T.,1987.TheN1waveofthehumanelectricand mag-neticresponsetosound:areviewandananalysisofthecomponent structure.Psychophysiology24(4),375–425.

Neumann,A.,vanLier,P.A.C.,Frijns,T.,Meeus,W.,Koot,H.M.,2011. Emotionaldynamicsinthedevelopmentofearlyadolescent psy-chopathology:aone-yearlongitudinalstudy.JournalofAbnormal ChildPsychology39(5),657–669.

Nowicki,S.,Mitchell, J.,1998.Accuracyinidentifyingaffectinchild andadultfacesandvoicesandsocialcompetenceinpreschool chil-dren.GeneticSocialandGeneralPsychologyMonographs124(1), 39–59.

Picton,T.W.,Bentin,S.,Berg,P.,Donchin,E.,Hillyard,S.A.,Johnson,R., Miller,G.A.,Ritter,W.,Ruchkin,D.S.,Rugg,M.D.,Taylor,M.J.,2000. Guidelinesforusinghumanevent-relatedpotentialstostudy cogni-tion:Recordingstandardsandpublicationcriteria.Psychophysiology 37(2),127–152.

Rogier,O.,Roux,S.,Belin,P.,Bonnet-Brilhault,F.,Bruneau,N.,2010. Anelectrophysiological correlateofvoice processingin4-to 5-year-oldchildren.InternationalJournalofPsychophysiology75(1), 44–47.

Rothman,A.D.,Nowicki,S.,2004.Ameasureoftheabilitytoidentify emo-tioninchildren’stoneofvoice.JournalofNonverbalBehavior28(2), 67–92.

Sander,D.,Grandjean,D.,Pourtois,G.,Schwartz,S.,Seghier,M.L.,Scherer, K.R.,Vuilleumier,P.,2005.Emotionandattentioninteractionsinsocial cognition:brainregionsinvolvedinprocessingangerprosody. Neu-roimage28(4),848–858.

Schirmer,A.,Kotz,S.A.,2006.Beyondtherighthemisphere:brain mech-anismsmediatingvocalemotionalprocessing.TrendsinCognitive Sciences10(1),24–30.

Schirmer,A.,Kotz,S.A.,Friederici,A.D.,2002.Sexdifferentiatestheroleof emotionalprosodyduringwordprocessing.CognitiveBrainResearch 14(2),228–233.

Schirmer,A.,Kotz,S.A.,Friederici,A.D.,2005.Ontheroleofattention fortheprocessingofemotionsinspeech:sexdifferencesrevisited. CognitiveBrainResearch24(3),442–452.

Semlitsch,H.V.,Anderer,P.,Schuster,P.,Presslich,O.,1986.Asolutionfor reliableandvalidreductionofocularartifacts,appliedtotheP300 ERP.Psychophysiology23(6),695–703.

Shackman,J.E.,Shackman,A.J.,Pollak,S.D.,2007.Physicalabuseamplifies attentiontothreatandincreasesanxietyinchildren.Emotion7(4), 838–852.

Silk,J.S.,Steinberg,L.,Morris,A.S.,2003.Adolescents’emotion regula-tionindailyfife:linkstodepressivesymptomsandproblembehavior. ChildDevelopment74(6),1869–1880.

Stevens,D.,Charman,T.,Blair,R.J.R.,2001.Recognitionofemotioninfacial expressionsandvocaltonesinchildrenwithpsychopathictendencies. JournalofGeneticPsychology162(2),201–211.

(10)

Stewart,J.L.,Silton,R.L.,Sass,S.M.,Fisher,J.E.,Edgar,J.C.,Heller,W., Miller,G.A.,2010.Attentionalbiastonegativeemotionasa func-tionofapproachandwithdrawalangerstyles:anERPinvestigation. InternationalJournalofPsychophysiology76(1),9–18.

Toivonen,M.,Rämä,P.,2009.N400duringrecognitionofvoiceidentity andvocalaffect.Neuroreport20(14),1245–1249.

Tonks,J.,Williams,W.H.,Frampton,I.,Yates,P.,Slater,A.,2007. Assess-ingemotionrecognitionin9–15-yearsolds:preliminaryanalysisof

abilitiesinreadingemotionfromfaces,voicesandeyes.BrainInjury 21(6),623–629.

Trentacosta,C.J.,Fine,S.E.,2010.Emotionknowledge,socialcompetence, andbehaviorproblemsinchildhoodandadolescence:ameta-analytic review.SocialDevelopment19(1),1–29.

Vlamings,P.H.J.M.,Jonkman,L.M.,Kemner,C.,2010.Aneyefordetail:an event-relatedpotentialstudyoftherapidprocessingoffearfulfacial expressionsinchildren.ChildDevelopment81(4),1304–1319.