Compensatory networks to counteract the effects of ageing on language

(1)

ContentslistsavailableatSciVerseScienceDirect

Behavioural

Brain

Research

jo u r n al hom e p ag e :w w w . e l s e v i e r . c o m / l o c a t e / b b r

Research

report

Compensatory

networks

to

counteract

the

effects

of

ageing

on

language

Rosa

Manenti

a,∗

_,

_Michela

_Brambilla

a,b

_,

_Michela

_Petesi

a

_,

_Carlo

_Miniussi

a,c

_,

_Maria

_Cotelli

a

a_IRCCS_Centro_San_Giovanni_di_Dio_{Fatebenefratelli,}_Brescia,_Italy

b_Center_for_Cognitive_Science,_Department_of_Psychology,_University_of_Turin,_Turin,_Italy

c_Dept_of_Clinical_and_Experimental_Sciences,_National_Institute_of_Neuroscience_University_of_Brescia,_Brescia,_Italy

h

i

g

h

l

i

g

h

t

s

•rTMSwasusedtoinﬂuencethefunctionoftheprefrontalcortex.

•TMSeffectsonnamingabilitydifferedaccordingtothesubject’sperformance. •Lessprefrontalasymmetryisastrategytocounteractage-relatednamingdecline.

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received21February2013 Receivedinrevisedform8April2013 Accepted11April2013

Available online 17 April 2013 Keywords: TMS Compensation Naming DLPFC Elderly

a

b

s

t

r

a

c

t

Background:Word-retrievaldifficultiesareacommonconsequenceofhealthyageingandareassociated withareductioninasymmetricalrecruitmentofthedorsolateralprefrontalcortex(DLPFC),althoughthe significanceofthisreductionhasnotyetbeenclarified.Usingrepetitivetranscranialmagneticstimulation (rTMS)ithasbeendemonstratedthatanasymmetricalinvolvementoftheDLPFCduringactionnamingin youngsubjects,whereasbilateralinvolvementwasshowninelderlyparticipants.ByusingrTMSduring anamingtaskinagroupofelderlysubjects,theaimofthepresentworkwastoinvestigatewhether themagnitudeofDLPFCasymmetry(left–rightrTMSeffect)duringactionnamingcorrelateswithtask performance,provingthepresenceofacompensationstrategyinsomebutnotallelderlyparticipants. Methods:WeaimedtotestiftherewasacorrelationbetweenDLPFCasymmetry(left–rightrTMSeffect) andnamingperformanceinagroupofelderlysubjects.

Results:TheresultsshowthatrTMSaffectsactionnamingdifferentlyaccordingtoindividualnaming ability.Inparticular,thepredominanceofaleftvs.rightDLPFCeffectwasobservedonlyinthe low-performingolderadults,whileanasymmetricreductionwasselectivelyshowninthehigh-performing group.Interestingly,high-performingolderadultsalsodisplayedbetterperformancesonaphonemic ﬂuencytest.

Conclusion:Thepresentdatasuggestthatsuccessfulageingislinkedtolessprefrontalasymmetry,an efﬁcientstrategyforcounteractingage-relateddeclinesincognitivefunction.

1. Introduction

Ample evidence has illustrated the relationship between age-relatedchangesinthebrainandalterationsincognitive per-formance[1].Studiesattheneuronallevelhavedemonstratedthat duringphysiologicalageing,dopaminergicdeclineaswellasgrey andwhitematteratrophyarebothcorrelatedwithspeciﬁc cog-nitivechanges[2,3].Arecentreview[4]concludedthatthebrain shrinksinvolumeandthattheventricularsystemexpandsinthe ageingbrain(withhighheterogeneity).Thevolumetricreductions

∗ Correspondingauthorat:IRCCSCentroSanGiovannidiDioFatebenefratelli,Via Pilastroni4,25125Brescia,Italy.Tel.:+390303501593;fax:+390303533513.

E-mailaddress:rmanenti@fatebenefratelli.it(R.Manenti).

arelikely related,toa minorextent,toneurallossbecausethe shrinkageofneurons,reductionofsynapticspinesandlower num-bersofsynapsesprobablyaccountforthereductioningreymatter [4].However,reductionsinprocessingspeed,memory,executive functionsandlanguageproductionhavebeenshownduring physi-ologicalageing.Suchreductionsaremediatedbyneuroanatomical changes,withincreasedageoftenassociatedwithlower perfor-manceonawide varietyofcognitivetasks[4,5].Accordingly,it isclearthatageingisinﬂuencedbyalargenumberoffactorsthat varyfromindividualtoindividual,and,todate,itisunclearhow dif-ferencesinbrainactivityrelatetocognitiveperformanceinelderly participants[5,6].Becausethemeanageofthepopulationincreases thenumberofpeoplethatwillexperienceanage-relatedreduction incognitiveabilitieswillincrease;despitethis,althoughsome indi-vidualswillshowpronouncedcognitivedeﬁcits, otherswillnot. 0166-4328/$–seefrontmatter © 2013 Elsevier B.V. All rights reserved.

(2)

Understandingthebasisofminorvs.majorage-relatedcognitive declineisofgreatinterest,andsomeinvestigationshaveassessed thepredictivevalueofseveralfactorsforsuccessfulageing[7,8], andnon-pharmacologicalinterventionsforcognitivedifﬁcultiesin healthyolderadultshavegainedmuchattentioninrecentyears [9–11].Relevant studieshave suggested thatcerebral plasticity andcognitivereserveplayimportantrolesinphysiologicalageing [9,10,12,13].

One method used to investigate age-related modifications in cognition is neuroimaging, and neuroimaging data from somecognitivemodelsonageinghavebeenacquired.A recent review highlightsthat theapplication offunctional neuroimag-ingtohealthyageingparticipantssuggestsfunctionalage-related changesduringtheexecutionofseveralcognitivetasks,despite the many discrepancies in this literature [14]. Alterations in hemisphericasymmetryhavebeenreportedinstudiesof physi-ologicalandpathologicalageing[15–17].Frontalasymmetrywas significantlyreducedduringacognitivetaskinhealthyelderly par-ticipantsascomparedtoyoungparticipants[16].Asoneofthe majorcognitivedeclinesinbothhealthyandpathologicalageing isareductioninmemoryperformance,mostoftheneuroimaging studieson healthy ageingparticipants have investigated mem-oryprocessesandsuggestthepresenceoffunctionalage-related reductionsintheleftprefrontalcortex(PFC)andtemporo-occipital regionsandincreasesinactivityintheinsularregions.Generally, thesestudieshave shown a reduction in functionalasymmetry intheprefrontalcortex[18]andhaveledtodevelopmentofthe hemisphericasymmetryreductioninolderadults(HAROLD)model [16], subsequentlyconfirmed byrTMS [16,19–24]. Additionally, over-activationofthePFCinolderadultsrepresentsthestarting pointoftheposterior–anteriorshiftinageing(PASA)model,which implicatesage-relatedincreasesinfrontalactivityanddecreasesin occipitallobeactivity[25].

Languagerepresentsanotherexampleofacognitiveabilitythat dramaticallychangesduringageing.Evidencefrombothlesionand imagingstudiessuggestsacentralrolefortheleftprefrontal, tem-poralandparietalareasduringnaming,althoughsomedifferences existfor object(noun)vs.action(verb)naming[26–28].A pro-gressivereductioninlexicalretrievalhasbeenwell-documented in older adults [29–31]. Unlike word-meaning retrieval, which appearstobepreservedorevenenhancedwithageing[32,33],the abilitytoretrievethesoundorphonologyofwordsseemstodecline inolderadults.Moreover,thefrequencyoftip-of-the-tongue expe-riencessigniﬁcantlyincreaseswithageing,andolderadultsreport an inabilityto produce well-known wordsas one of themost annoyingcognitive“symptoms”thattheyexperience[34,35].Afew imagingstudieshavereportedonaneuralage-relatedchangethat correlateswiththeincreaseofthisdisturbanceinhealthyageing. Theresultsofthesestudiessupportaneuralcauseofwordretrieval difﬁculties(tip-of-tongueeffect)inphysiologicalageing, highlight-ingtheeffectsofageingontheactivityoftheinsula[36]andin thewhitematterintegrityoftheposterioraspectofthesuperior longitudinalfasciculus[37].

Evenifnamingisanabilitythatshowsadaptationduringageing, onlyafewstudieshaveinvestigatedhowage-relatedchangesin activityrelatetoperformanceduringnamingtasksinolderadults. Differenthypotheseshavebeenconsideredtoexplainthe sig-nificance of these age-related changes. First, thecompensation hypothesis[25,38–41] proposesthat increasedfunctional hemi-sphericsymmetryinolderadultsiscounteractedbyage-related neurocognitivedeficits. Second,thededifferentiationhypothesis explainsthereducedasymmetryasadifficultyinrecruiting spe-cialisedneuralnetworks[42,43].Last,thedefaultnetworktheory postulatesthatthemajorcomponentsofthedefaultmodenetwork remainstableinhealthy,olderindividuals, whereastheactivity ina number ofdiscrete corticalareaslocatedin theprefrontal,

temporal,andoccipitalregionschangesovertime[44].Thereisa largedegreeofoverlapamongthesehypotheses,butallofthemat leastpartiallyunderlinetheabilityofcerebraltissuetochangeits structureandfunctioncontinuouslyinresponseto environmen-taldemands.Tocomparethesehypotheses,severalstudieshave investigatedwhetherPFCfunctionalsymmetryislinkedtoa reduc-tionorenhancementofmemoryperformance,buttheresultsare stillinconclusive.Regardingnamingtasks,onlyafew neuroimag-ingstudieshaveshownage-inducedincreasesincorticalactivation inthebrainareastypicallyassociatedwithlanguageprocessing (Broca’sandWernicke’sareas)[45,46].Interestingly,onlyonefMRI studyhasexaminedtheroleoffunctionalage-relatedchangesin namingabilitiesandfoundapositivecorrelationbetweenBOLD responses,namingaccuracyandresponselatenciesinthefrontal areasofolderadults[46],verifyingthepresenceofcompensatory mechanismsinthefrontalcortexthatareinvokedtomaintain per-formanceinhealthyageing.Arecentreviewofthefunctionalbrain imaging correlatesof successful cognitiveageing found frontal cortex responsesinolder,as compared toyounger,individuals, independentoftask[41].

However,functionalneuroimagingdatacannotprovethe neces-sityofanareaduringacognitivetaskbecauseanactivatedarea maysimplybecorrelatedwith, ratherthanresponsiblefor,task performance.Incontrast,repetitivetranscranialmagnetic stimula-tion(rTMS)caninduceatemporaryimpairmentinperformance onlyifthestimulatedareaiscausallyengagedin thetask[47]. Basedonthisassumption,TMShasbeenusedinmanydifferent cognitivedomainstoestablishcausalityinbrain–behaviour rela-tionshipsand,speciﬁcally,toverifyasymmetricalinvolvementof thePFCduringselectedtasks.Hemisphericspecialisationsofthe dorsolateralPFC(DLPFC)duringepisodicmemoryhavepreviously beendemonstratedusingrTMSinyoungsubjects[19,21–23,48–52] andinelderlyadults[20,53].Foranextensivereviewofthis lit-erature,pleaserefertoManentietal.[24].Regardingnaming,an rTMSstudyofyoungparticipantsshowedaleftpredominanceof theDLPFCduringactionnaming[54].Auniquestudythat inves-tigatedtheeffects ofphysiological ageingonDLPFCasymmetry duringactionandobjectnamingshowedaselectivebilateral facil-itatory effectduringaction naming,namelya decreaseinvocal reactiontime,inagroupofolderadults[55].

Thepurposeofthisstudywastoexamineifthemagnitudeof DLPFClateralisation(left–rightrTMSeffect)duringactionnaming wascorrelatedwithtaskperformance,provingthepresenceofa compensationstrategyinsomebutnotallelderlyparticipants.In thepresentreport,were-analyseddataacquiredinourprevious study[55]tothoroughlyinvestigatethisissue.Basedonprevious studies[20,46,56],wepredictedthatDLPFChemispherical symme-tryduringactionnamingwouldonlybepresentinhigh-performing olderadults.If identiﬁed,this resultwould provethata reduc-tioninDLPFCasymmetryinsomebutnotallelderlyparticipants representsacompensationstrategyusedtocounteractage-related changes,resultinginbettertaskperformance.

2. Materialsandmethods 2.1. Participants

Priortobeingenrolledintheexperiment,participantswereadministereda standardhealthhistoryquestionnaireandcompletedaMiniMentalState Exam-ination(MMSE)[57].Potentialparticipantswereexcludediftheyreportedahistory ofneurologicaldisease,cardiovasculardisease,psychiatricdisordersoralcoholor othersubstanceabuse.Individualswhoreportedsubjectivememorycomplaintsor scoredbelow27outof30ontheMMSEwerealsoexcluded.Inaddition,a neuro-psychologicalbatterywasapplied,andapathologicalscoreinatleastoneofthe testswasafurtherexclusioncriterion.Thirteen(4male,9female)healthyolder adults(age:65–78years,mean=70.2years,education:mean=13.8years) partici-patedintherTMSexperiment.AllparticipantswerenativeItalianspeakersandhad normalorcorrected-to-normalvision.Allparticipantswereright-handed[58]and hadnocontraindicationsforrTMS[59].Thisstudywasapprovedbythelocalethics

(3)

committee,andparticipantswereinformedaboutthepossiblerisksofrTMSbefore informedconsentwasobtained.

Theneuropsychologicaltestbatteryincludedmeasuresusedtoassess non-verbalreasoning(Raven’sColouredProgressiveMatrices),languagecomprehension (TokenTest),verbalﬂuency(phonemicandsemantic),memory(StoryRecall, Rey-OsterriethComplexFigureRecall,DigitSpan,SpatialSpan),visuo-spatialabilities (Rey-OsterriethComplexFigure,Copy),attentionandexecutivefunctions (Trail-MakingTestAandB).Allthetestswereadministeredandscoredaccordingto standardprocedures[60].Inaddition,somesubtestsoftheBatteryforAnalysisof AphasicDeﬁcits(BADA)wereapplied[61].Theresultsofthecognitiveassessments arepresentedinTable1.

2.2. Stimuli

Thestimuliusedintheactionandobjectpicture-namingtaskswereprovided bytheCentreforResearchinLanguage-InternationalPictureNamingProject (CRL-IPNP)[62].Weselected84items(42actionsand42objects)inapreviousexperiment [55],andnoneoftheactionstimuliincludedinthistaskwereassociatedwith theselectedobjects.Thenounsandverbscorrespondingtothesetsofobjectsand actionswerematchedfortargetwordfrequencyandlength.

Theitemsweredividedintothreeblocksdesignedforthethreestimulation conditions(leftDLPFC,rightDLPFCandsham).Thefrequenciesandlengthsofthe targetwordswerecounterbalancedineachexperimentalblock.Thevisual com-plexityandimageabilityofthepictureswerealsomatchedbetweenblocks.Ten additionalobjectsandactionswereusedasapracticeblock(5actionsand5objects). Thepresentationorderwascounterbalancedacrosssubjects.

2.3. Procedure 2.3.1. Namingtask

Subjectssatinfrontofa17-in.monitorcontrolledbyapersonalcomputer run-ningthePresentationsoftware(www.neurobs.com).Afteraframethatindicated thecategoryofthestimulustothesubject(“ACTION”or“OBJECT”)wasdisplayed, a50mssoundwaspresentedupontheonsetofacentrallylocatedfixationcross. Thefixationcrosswaspresentfor1000ms.Afterthefixationcrossdisappeared,the stimuluswaspresentedandremainedonthescreenfor1000ms.Ablankscreen followedforanywherefrom4000to5000ms.Thesubject’staskwastoaccurately

namethestimuliappearingonthecomputerscreenasquicklyaspossible.Verbal responseswererecordedanddigitisedat44.1kHzusingtheGoldWaveprogram (V.5.12,www.goldwave.com).Theresponseswerethenanalysedoff-linetoobtain accuracyandvocalreactiontimes(vRTs).Foreachstimulus,wecalculatedthemean vRTandthemeanresponseaccuracypercentage.

2.3.2. Transcranialmagneticstimulation

ForleftandrightDLPFCstimulation,theTalairachcoordinatesofthecortical sitesunderlyingthecoilwereestimatedforeachsubjectusingaSofTaxicEvolution Navigatorsystem(V.2.0;www.emsmedical.net)andbyidentifyingthestimulation siteonthescalpaboveBrodmannarea8(TalairachcoordinatesX=±35,Y=24,Z=48, middlefrontal).Weuseda70mmﬁgure-eightcoilplacedabovethetargetpoint todeliverrTMSfor500msfromtheonsetofthevisualstimulusatafrequency of20Hz.Thestimulationintensityusedduringtheexperimentwassetat90%of eachsubject’srestingmotorthreshold.Fortheshamcontrolcondition,a 3-cm-thickpieceofplywoodwasappliedtothecoilsothatnomagneticﬁeldsreached thecortexandthejunctionofthetwocoilwingswasplacedabovethevertex(CZ intheEEG10/20internationalsystem).Theseparametersareinlinewiththesafety recommendationsforrTMS[59],andnoneofthesubjectsshowedsideeffectsof stimulation.

2.4. Analysisofthelinkbetweenhemisphericasymmetryandactionnaming performance

Weonlyanalysedactionnamingtaskperformance,followingtheresultsofour previousstudythatshowedsigniﬁcantrTMS-inducedeffectsexclusivelyduring actionnaming[55].BoththevRTsandaccuracywereanalysed.Oneindexof inter-hemisphericasymmetrywascomputedatthesingle-subjectlevelbyexaminingthe differencebetweenthevRTsobtainedintherightandleftstimulationconditions.

Statisticalanalyses wereperformedusing Statistica(version 10) software (www.statsoft.com).Totestthecompensationhypothesis,wecalculatedthe z-pointsofthevRTsduringactionnamingintheshamconditionbydividingour participantsintotwogroups:high-performing(N=7)andlow-performing(N=6). Performanceachievedintheshamconditionwasusedtodividetheparticipantsinto twogroupsbecauseitrepresentstheparticipants’baselineabilities.Thedistribution ofz-pointsischaracterisedbyaminimumof−1.1andamaximumof+2.0andthe meanofthisdistribution(i.e.,0)isusedascut-offinordertodistinguishbetween

Table1

Demographiccharacteristics,neuropsychologicalandexperimentalassessmentoftheolderadultgroupasawholeandofthehigh-andlow-performinggroupsseparately. p:p-value;ns:notsigniﬁcant.

Demographiccharacteristics Allgroup(N=13) High-performing(N=7) Low-performing(n=6) p

Age(years) 70.2±4.0 70.6±4.3 69.8±3.5 ns

Education(years) 13.8±4.9 13.9±4.8 13.8±4.9 ns

EdinburghHandednessInventory +99.6±1.4 +100±0.0 +99.2±2.0 ns

Gender(female/male) 9/4 4/3 5/1 ns

Neuropsychologicalassessment Allgroup(N=13) High-performing(N=7) Low-performing(N=6) Cut-off p Screeningfordementia

MMSE 29.5±0.8 29.6±0.8 29.5±0.8 ≥24 ns

Non-verbalreasoning

Raven-ColouredProgressiveMatrices 30.9±2.3 31.7±2.8 30.0±1.4 >17.5 ns Memory

StoryRecall 15.2±2.0 14.6±2.0 15.8±1.8 >7.5 ns

Rey-OsterriethComplexFigure,Recall 14.6±4.4 15.9±4.5 13.2±4.1 >9.46 ns

DigitSpan 5.7±0.9 5.4±1.0 6.0±0.6 >3.5 ns

SpatialSpan 4.9±0.5 4.7±0.5 5.2±0.4 >3.5 ns

Praxia

Rey-OsterriethComplexFigure,Copy 33.1±1.7 33.1±2.0 33.1±1.5 >28.87 ns Executivefunction Trail-MakingTestA(s) 40.2±9.1 43.3±8.2 36.5±9.3 <93 ns Trail-MakingTestB(s) 117.1±41.6 128.7±44.1 103.5±37.6 <282 ns Language TokenTest 33.6±1.3 33.7±1.5 33.5±1.0 >26.25 ns Fluency,Phonemic 39.6±8.2 43.6±7.3 35.0±7.0 >16 0.05 Fluency,Semantic 45.3±7.6 45.0±10.0 45.7±4.1 >24 ns

ObjectOralComprehension(BADA) 99.6±1.4 100.0±0.0 99.2±2.0 ns ActionOralComprehension(BADA) 99.6±1.4 100.0±0.0 99.2±2.0 ns ObjectWrittenComprehension(BADA) 99.8±0.7 100.0±0.0 99.6±1.0 ns ActionWrittenComprehension(BADA) 99.6±1.4 99.3±1.9 100.0±0.0 ns

OralObjectNaming(BADA) 97.2±2.3 97.6±2.5 96.7±2.1 ns

OralActionNaming(BADA) 96.4±2.9 95.9±2.5 97.0±3.5 ns

ExperimentalassessmentduringShamcondition Actionnaming

Correctness,% 87.4±13.7 88.4±10.0 85.7±18.1 ns

Vocalreactiontimes,ms 1077.5±129.3 978.9±35.1 1192.4±95.9 0.0001 Rawscoresarereported.Cut-offscoresreferredtoItaliannormativedata.Boldscoreshighlightsigniﬁcantdifferencesbetweenhigh-andlow-performinggroups.

(4)

thetwogroups.Weusedt-testscomparisonsinordertocompareperformances achievedinthetwoelderlygroups.Accordingly,wedeterminedifanystandard neuropsychologicaltestwasabletodifferentiatebetweenparticipantswith bet-terperformanceandparticipantswithreducedaccuracy(seeTable1).Importantly, weinvestigateddifferencesinthetwogroupsusingourindexofinterhemispheric asymmetry.

Consideringthetwogroups,wecomparedtheindexofinterhemispheric asym-metryofthetwoolderparticipantgroupswiththemeanleft–rightasymmetry (mean=−120ms)ofyoungindividuals[54].Asimilaranalysiswasconductedto comparethevRTsduringactionnaminginthehigh-andlow-performinggroupsof olderadults(shamconditionasbaseline)withthemeanvRTsduringactionnaming inyoungparticipants(mean=911ms).

Finally,theindexofinterhemisphericasymmetrywasregressedagainstthevRTs andaccuracyintheshamcondition(i.e.,individualabilities).Alinearalgorithmwas appliedtoinvestigatewhethergreaterasymmetrycorrespondedtobetterorworse performanceontheactionnamingtask.

Thelevelofsigniﬁcancewassetatp<0.05.

3. Results

3.1. High-andlow-performingelderlygroups

Thehigh-performing(N=7)andlow-performing(N=6)groups didnotdifferinageoreducation(p>0.05).Thetwogroups did notdifferinactionnamingaccuracy,althoughtheaccuracywas slightly lower in thelow-performing group. Independent sam-plet-testsontheresultsoftheneuropsychologicaltestshighlight betterscoresinphonemicverbalfluencyinthehigh-performing group(t=2.17,p=0.05).Theindexofinterhemisphericasymmetry wasdifferentbetweenthetwogroups,withasignificantlylower score(higherasymmetry)forthelow-performinggroup(t=5.04, p=0.002).Fig.1Ashowsagraphicalrepresentationofmeanscoreat phonemicverbalfluencytest(y-axis)plottedseparatelyforthetwo groups(meanindexofinterhemisphericasymmetryonthex-axis). Noneoftheothertestsshowedsignificance.

3.2. High-andlow-performingelderlygroupsvs.younggroup Acomparisonofthetwooldergroupswiththeyounggroup showedsimilarDLPFCasymmetryin thelow-performinggroup (p>0.05)anda signiﬁcantdifferenceinasymmetryinthe high-performinggroup(t=5.04,p=0.002) ascompared totheyoung group[54].Furthermore,thisanalysisshowedfastervRTsinthe younggroupascomparedtoboththehigh-(t=5.1,p=0.002)and low-(t=7.2,p=0.0008)performingolderadults,eventhoughthe high-performingoldergroupshowedshortenedvRTsascompared to the low-performing older group (t=−5.5, p=0.0001). These resultssuggestthatthehigh-performingolderadultssigniﬁcantly reducedDLPFCfunctionalasymmetrytoobtainbetterperformance. Fig.1BshowsagraphicalrepresentationofmeanvRTsduringthe experimentalnamingtask(y-axis)plottedseparatelyforthetwo elderlygroupsandfortheyounggroup[54](meanindexof inter-hemisphericasymmetryonthex-axis).

3.3. Correlations

A signiﬁcant negative correlation (R=0.68, t=3.07, p=0.01) betweentheindexofinterhemisphericasymmetryandthevRTs intheshamconditionsuggeststhattheslowerparticipantswere characterised by greater involvement of the left hemisphere. Accordingly, a signiﬁcant positive correlation (R=0.54, t=2.17, p=0.05)betweentheindexofinterhemisphericasymmetryand accuracyintheshamconditionindicatesthattheparticipantsthat achievedbetterperformanceshowedreducedinvolvementofthe lefthemisphere.Fig.2presentsadetailedrepresentationofthese results(vRTsandaccuracyony-axisandtheindexof interhemi-sphericasymmetryonx-axis).

Fig.1. (A)Meanphonemicverbalfluencyscoreinolderparticipantswithhigh (N=7)orlow(N=6)actionnamingperformance.Interestingly,high-performing elderlyparticipantsshowedlesslefthemisphereinvolvementandbetter phone-micverbalfluencyscores,whereastheresultsfromthelow-performingelderly participantsshowedtheoppositepattern.(B)Meanvocalreactiontimes(vRTs)in young([54];N=9)andolderparticipantswithhigh(N=7)orlow(N=6)action namingperformance.Higherinvolvementofthelefthemisphere(i.e.,negative dif-ferencesbetweenvRTsintheleftandrightrTMSconditions)isevidentinyoung participantsandinolderparticipantswithlowperformances,whereasolder par-ticipantswithhighperformancesshowrightorbilateralinvolvementoftheDLPFC. Interestingly,thevRTsofthehigh-performingelderlyparticipantsaremuchmore similartothoseofyoungparticipants,buttheasymmetryinthehigh-performing elderlyparticipantsissignificantlydifferentthanthatofyoungindividuals.Inorder toobtainbetterperformance,functionalreorganisationisneeded.Low-performing olderadultsshowthesameDLPFCasymmetryasyoungindividualsbutwith signif-icantlyhighervRTs.

4. Discussion

Theprimarygoalofthepresentstudywastoexamineifthe asymmetricreductionshowninolderadults,ascomparedtoyoung individuals,couldbeacompensationstrategyusedtocounteract ageing.Accordingly,weinvestigatedtheinﬂuenceofleftandright DLPFCrTMSonperformanceduringactionnaming.

WeappliedrTMSduringactionnaminginhealthy,high-or low-performing,olderparticipantsandfoundthatgreaterdifferences between left and right DLPFCinvolvement during action nam-ingresultedinhigheraccuracyandfasterresponsesinthistask. Theseresultssuggestselectiveasymmetricalrecruitmentof the DLPFC(left>right)duringactionnaminginlow-performingolder adults,aspreviouslydescribedinyoungindividuals,whereas high-performingolderadultsdidnotshowpredominance(left=right). Low-performingolderadultsrecruitedasimilarnetworkofbrain regions as theyoung adults but used them inefﬁciently.These ﬁndingshave tobediscussedinterms offunctional changesin

(5)

Fig.2. Correlationsbetweentheindexofinterhemisphericasymmetry(verbal reac-tiontimesduringleftDLPFCstimulationminusverbalreactiontimesduringright DLPFCstimulation)inactionnamingandtheverbalreactiontimes(A)oraccuracy (B)duringshamconditions.

successfulageing[41].Inarecentreviewofthefunctionalbrain imagingcorrelatesofsuccessfulcognitiveageing,Eyleretal.[41] concluded that augmented brain responses, particularly in the frontalcortex,servea compensatoryrole inolderadults,but in somecases,extraneousactivationmight bedetrimental(i.e.,in theposteriorregions).OurresultsareinlinewiththoseofEyler etal.[41],asfrontalcortexfunctionalrecruitmentwasselectively modiﬁedduringsuccessfulageing(i.e.,inhigh-performingolder adults).

Our results seem to be in agreement with the compensa-tion hypothesis, which has been investigated in the memory domain[63],becauseDLPFCrecruitmentwasmodifiedonlyinthe high-performingparticipants,ascomparedtoyoungparticipants. However,thisresultdoesnotfitthededifferentiationhypothesis [42,43].Ifreducedlateralisationismerelyanotherexampleofthe deleteriouseffectsofageingonthebrain,thenitshouldhavebeen foundinthelow-performingelderlygroupofadultsbecausethey displayedmorepronouncedage-relatedcognitivedeficits; how-ever,thatwasnotthecase.Onthecontrary,reducedlateralisation wasfoundinthegroupofhigh-performingparticipants,suggesting thatthisfeaturemightbeconsideredasabeneficialchange. More-over,thehigh-performingparticipantsachievedhigherscoresin phonemicfluency,underlingbetterwordfindingabilitiesinthese participants.

Ourgroupfoundsimilarresultsinapreviousstudythatalsoused brainstimulationduringanepisodicmemorytask[20].We veri-fiedthatduringtheencodingofwordpairs,onlyhigh-performing olderadultsshowedbilateralDLPFCinvolvement,whereasaleft predominance,asobservedinyoungindividuals,waspresentin thelow-performing participants.Asreported fornaminginthe present work, functional modification of the DLPFC was posi-tivelycorrelatedwithbettermemoryperformance,suggestingthat thesechangesrepresentacompensatorystrategyinsuccessful age-ing.Importantly,agreementintheresultsregardingthedifferent cognitiveabilitiesstrengthensthecompensationhypothesis.We acknowledgethattherelativelysmallnumberofelderly partici-pantsincludedinourstudyrepresentalimitation.However,the findingofthisstudycouldpavethewaytofurtherstudieswitha largersample.

Althoughfurtherstudiesareneededtoconﬁrmpresentresults and clarify the basis of these effects, the possibility to better deﬁnethepresenceofageingstrategiesresponsibleforimpairedor maintainedcognitiveabilityitisalsoimportantfordesigning reha-bilitationprotocolsinapatientwithlanguagedisabilities[64–69]. Methods to reduce or magnify inter-individual differences in performance, and theintegration of thesemechanisms with functionalage-relatedplasticity,mayhavepracticalandsocietal implications.A greater understandingof howthese patternsof compensatorystrategiesdevelopandhelptomaintaincognitive functioninageingisanimportantgoalforfutureresearch.

References

[1]SalthouseTA.Effectsofageontime-dependentcognitivechange.Psychological Science2011;22:682–8.

[2]KramerJH,MungasD, ReedBR,WetzelME,BurnettMM,MillerBL,etal. LongitudinalMRIandcognitivechangeinhealthyelderly.Neuropsychology 2007;21:412–8.

[3]Sullivan EV, Rohlﬁng T, Pfefferbaum A. Longitudinal study of callosal microstructureinthenormaladultagingbrainusingquantitativeDTIﬁber tracking.DevelopmentalNeuropsychology2010;35:233–56.

[4]FjellAM,WalhovdKB.Structuralbrainchangesinaging:courses,causesand cognitiveconsequences.ReviewsintheNeurosciences2010;21:187–221. [5]Reuter-LorenzPA,ParkDC.Humanneuroscienceandtheagingmind:anew

lookatoldproblems.JournalsofGerontologySeriesB,PsychologicalSciences andSocialSciences2010;65:405–15.

[6]GradyC.Thecognitiveneuroscienceofageing.NatureReviewsNeuroscience 2012;13:491–505.

[7]FormigaF,FerrerA,AlburquerqueJ,Fernandez-QuevedoM,RoyoC,PujolR, etal.Thechallengeofmaintainingsuccessfulagingat87yearsold:theoctabaix studytwo-yearfollow-up.RejuvenationResearch2012;15:584–9.

[8]HowiesonDB,CamicioliR,QuinnJ,SilbertLC,CareB,MooreMM,etal.Natural historyofcognitivedeclineintheoldold.Neurology2003;60:1489–94. [9]BallK,BerchDB,HelmersKF,JobeJB,LeveckMD,MarsiskeM,etal.Effects

ofcognitivetraininginterventionswitholderadults:arandomizedcontrolled trial.JournalofAmericanMedicalAssociation2002;288:2271–81.

[10]Acevedo A, LoewensteinDA. Nonpharmacological cognitive interventions in aging and dementia. Journal of Geriatric Psychiatry and Neurology 2007;20:239–49.

[11]WillisSL,TennstedtSL,MarsiskeM,BallK,EliasJ,KoepkeKM,etal.Long-term effectsofcognitivetrainingoneverydayfunctionaloutcomesinolderadults. JournalofAmericanMedicalAssociation2006;296:2805–14.

[12]SternY.CognitivereserveandAlzheimerdisease.AlzheimerDiseaseand Asso-ciatedDisorders2006;20:S69–74.

[13]ValenzuelaMJ,SachdevP.Assessmentofcomplexmentalactivityacrossthe lifespan:developmentoftheLifetimeofExperiencesQuestionnaire(LEQ). Psy-chologicalMedicine2007;37:1015–25.

[14]Guidotti Breting LM,Tuminello ER, Duke Han S.Functional neuroimag-ing studies innormal aging. CurrentTopics in BehavioralNeurosciences 2012;10:91–111.

[15]CherbuinN,Reglade-MeslinC,KumarR,SachdevP,AnsteyKJ.Mild cogni-tivedisordersareassociatedwithdifferentpatternsofbrainasymmetrythan normalaging:thePATHthroughlifestudy.FrontPsychiatry2010;1:11. [16]CabezaR.Hemisphericasymmetryreductioninolderadults:theHAROLD

model.PsychologyandAging2002;17:85–100.

[17]Long X, Zhang L, Liao W, Jiang C, Qiu B. The Alzheimer’s Disease Neuroimaging. I. Distinct laterality alterations distinguish mild cognitive impairmentand Alzheimer’sdisease fromhealthyaging: statistical para-metric mappingwith high resolution MRI.Human BrainMapping 2012, http://dx.doi.org/10.1002/hbm.22157,inpress.

(6)

[18]CabezaR,GradyCL,NybergL,McIntoshAR,TulvingE,KapurS,etal.Age-related differencesinneuralactivityduringmemoryencodingandretrieval:apositron emissiontomographystudy.JournalofNeuroscience1997;17:391–400. [19]FloelA,PoeppelD,BuffaloEA,BraunA,WuCW,SeoHJ,etal.Prefrontalcortex

asymmetryformemoryencodingofwordsandabstractshapes.CerebralCortex 2004;14:404–9.

[20]Manenti R, Cotelli M, Miniussi C. Successful physiological aging and episodic memory: a brainstimulation study.BehaviouralBrain Research 2011;216:153–8.

[21]RamiL,GironellA,KulisevskyJ,Garcia-SanchezC,BerthierM,Estevez-Gonzalez A.Effectsofrepetitivetranscranialmagneticstimulationonmemorysubtypes: acontrolledstudy.Neuropsychologia2003;41:1877–83.

[22]RossiS,CappaSF,BabiloniC,PasqualettiP,MiniussiC,CarducciF,etal. Pre-frontalcortexinlong-termmemory:aninterferenceapproachusingmagnetic stimulation.NatureNeuroscience2001;4:948–52.

[23]SandriniM,CappaSF,RossiS,RossiniPM,MiniussiC.Theroleofprefrontal cortexinverbalepisodicmemory:rTMSevidence.JournalofCognitive Neuro-science2003;15:855–61.

[24]ManentiR,CotelliM,RobertsonIH,MiniussiC.Transcranialbrainstimulation studiesofepisodicmemoryinyoungadults,elderlyadultsandindividualswith memorydysfunction:areview.BrainStimulation2012;5:103–9.

[25]Davis SW,DennisNA,Daselaar SM,FleckMS,Cabeza R.Que PASA?The posterior–anteriorshiftinaging.CerebralCortex2008;18:1201–9.

[26]DanieleA,GiustolisiL,SilveriMC,ColosimoC,GainottiG.Evidenceforapossible neuroanatomicalbasisforlexicalprocessingofnounsandverbs. Neuropsy-chologia1994;32:1325–41.

[27]PeraniD,CappaSF,SchnurT,TettamantiM,CollinaS,RosaMM,etal.The neuralcorrelatesofverbandnounprocessing.APETstudy.Brain1999;122(Pt 12):2337–44.

[28]PriceCJ,DevlinJT,MooreCJ,MortonC,LairdAR.Meta-analysesofobject nam-ing:effectofbaseline.HumanBrainMapping2005;25:70–82.

[29]BarresiBA,NicholasM,ConnorLT,OblerL,AlbertML.Semanticdegradation andlexicalaccessinage-relatednamingfailures.Aging,Neuropsychologyand Cognition2000;7:169–78.

[30]ConnorLT,SpiroA,3rd,OblerLK,AlbertML.Changeinobjectnamingability duringadulthood.JournalsofGerontologySeriesB,PsychologicalSciencesand SocialSciences2004;59:P203–9.

[31]MorrisonCM,HirshKW,DugganGB.Ageofacquisition,ageing,andverb pro-duction:normativeandexperimentaldata.QuarterlyJournalofExperimental PsychologyA,HumanExperimentalPsychology2003;56:705–30.

[32]GoralM,Spiro3rdA,AlbertM,OblerLK,ConnorLT.Changeinlexical–retrieval skillsinadulthood.TheMentalLexicon2007:215–38.

[33]VerhaeghenP.Agingandvocabularyscores:ameta-analysis.Psychologyand Aging2003;18:332–9.

[34]BurkeDM,ShaftoMA.Agingandlanguageproduction.CurrentDirectionsin PsychologicalScience2004;13:21–4.

[35]ShaftoMA,BurkeDM,StamatakisEA,TamPP,TylerLK.Onthe tip-of-the-tongue:neuralcorrelatesofincreasedword-ﬁndingfailuresinnormalaging. JournalofCognitiveNeuroscience2007;19:2060–70.

[36]ShaftoMA,StamatakisEA,TamPP,TylerLK.Wordretrievalfailuresinoldage: therelationshipbetweenstructureandfunction.JournalofCognitive Neuro-science2010;22:1530–40.

[37]StamatakisEA,ShaftoMA,WilliamsG,TamP,TylerLK.Whitematterchanges andwordﬁndingfailureswithincreasingage.PLoSONE2011;6:e14496. [38]BerlingeriM,BottiniG,DanelliL,FerriF,TraﬁcanteD,SacheliL,etal.With

timeonourside?Task-dependentcompensatoryprocessesingracefulaging. ExperimentalBrainResearch2010;205:307–24.

[39]BerlingeriM,SacheliL,DanelliL,FerriF,TraﬁcanteD,BasilicoS,etal. Neuro-functionalandneuromorphologicalevidenceofthelackofcompensationin pathologicalaging.BehaviouralNeurology2010;23:185–7.

[40]Cabeza R,LocantoreJK,AndersonND.Lateralizationofprefrontalactivity duringepisodicmemoryretrieval:evidencefortheproduction-monitoring hypothesis.JournalofCognitiveNeuroscience2003;15:249–59.

[41]EylerLT,SherzaiA,KaupAR,JesteDV.Areviewoffunctionalbrainimaging correlatesofsuccessfulcognitiveaging.BiologicalPsychiatry2011;70:115–22. [42]CheeMW,GohJO,VenkatramanV,TanJC,GutchessA,SuttonB,etal. Age-relatedchangesinobjectprocessingandcontextualbindingrevealedusing fMRadaptation.JournalofCognitiveNeuroscience2006;18:495–507. [43]VossMW,EricksonKI,ChaddockL,PrakashRS,ColcombeSJ,MorrisKS,etal.

Dedifferentiationinthevisualcortex:anfMRIinvestigationofindividual dif-ferencesinolderadults.BrainResearch2008;1244:121–31.

[44]Beason-HeldLL,KrautMA,ResnickSM.Stabilityofdefault-modenetwork activ-ityintheagingbrain.BrainImagingandBehaviour2009;3:123–31. [45]FridrikssonJ,MorrowKL,MoserD,BaylisGC.Age-relatedvariabilityincortical

activityduringlanguageprocessing.JournalofSpeech,Language,andHearing Research2006;49:690–7.

[46]WierengaCE,BenjaminM,GopinathK,PerlsteinWM,LeonardCM,RothiLJ, etal.Age-relatedchangesinwordretrieval:roleofbilateralfrontaland sub-corticalnetworks.NeurobiologyofAging2008;29:436–51.

[47]MiniussiC,RuzzoliM,WalshV.Themechanismoftranscranialmagnetic stim-ulationincognition.Cortex2010;46:128–30.

[48]EpsteinCM,SekinoM,YamaguchiK,KamiyaS,UenoS.Asymmetriesof pre-frontalcortexinhumanepisodicmemory:effectsoftranscranialmagnetic stimulationonlearningabstractpatterns.NeuroscienceLetters2002;320:5–8. [49]KohlerS,PausT,BucknerRL,MilnerB.Effectsofleftinferiorprefrontal stim-ulationonepisodicmemoryformation:atwo-stagefMRI-rTMSstudy.Journal ofCognitiveNeuroscience2004;16:178–88.

[50]ManentiR,CotelliM,CalabriaM,MaioliC,MiniussiC.Theroleofthe dor-solateralprefrontalcortexinretrievalfromlong-termmemorydependson strategies:arepetitivetranscranialmagneticstimulationstudy.Neuroscience 2010;166:501–7.

[51]ManentiR,TettamantiM,CotelliM,MiniussiC,CappaSF.Theneuralbasesof wordencodingandretrieval:afMRI-guidedtranscranialmagneticstimulation study.BrainTopography2010;22:318–32.

[52]RossiS,PasqualettiP,ZitoG,VecchioF,CappaSF,MiniussiC,etal.Prefrontal andparietalcortexinhumanepisodicmemory:aninterferencestudyby repetitivetranscranialmagneticstimulation.EuropeanJournalofNeuroscience 2006;23:793–800.

[53]RossiS,MiniussiC,PasqualettiP,BabiloniC, RossiniPM,Cappa SF. Age-related functionalchanges of prefrontal cortexin long-term memory: a repetitivetranscranialmagneticstimulationstudy.JournalofNeuroscience 2004;24:7939–44.

[54]CappaSF,SandriniM,RossiniPM,SostaK,MiniussiC.Theroleoftheleftfrontal lobeinactionnaming:rTMSevidence.Neurology2002;59:720–3.

[55]CotelliM,ManentiR,RosiniS,CalabriaM,BrambillaM,BisiacchiPS,etal.Action, andobjectnaminginphysiologicalaging:anrTMSstudy.FrontiersinAging Neuroscience2010;2:151.

[56]Cabeza R, Anderson ND, Locantore JK, McIntosh AR. Aging gracefully: compensatorybrainactivityinhigh-performing olderadults.Neuroimage 2002;17:1394–402.

[57]FolsteinMF,FolsteinSE,McHughPR.Mini-mentalstate.Apracticalmethodfor gradingthecognitivestateofpatientsfortheclinician.JournalofPsychiatric Research1975;12:189–98.

[58]OldﬁeldRC.Theassessmentandanalysisofhandedness:theEdinburgh inven-tory.Neuropsychologia1971;9:97–113.

[59]RossiS,HallettM,RossiniPM,Pascual-LeoneA,GroupSoTC.Safety,ethical considerationsandapplicationguidelinesfortheuseoftranscranical mag-neticstimulationinclinicalpracticeandresearch.Aconsensusstatementfrom theInternationalWorkshoponPresentandFutureofTMS:safetyandethical guidelines.ClinicalNeurophysiology2009;120(12):2008–39.

[60]LezakM,HowiesonD,LoringDW.Neuropsychologicalassessment.4thed. Oxford:UniversityPress;2004.

[61]MiceliG,LaudannaA,BuraniC,PapassoR.Batteriaperl’AnalisideiDeﬁcit Afa-sici.BADA(BatteryforAnalysisofAphasicDeﬁcits).Milano:CEPSAG,Università CattolicadelSacroCuore;1994.

[62]BatesE,AndonovaE,D’AmicoS,JacobsenT,KohnertK,LuCC,etal. Introduc-ingtheCRLInternationalPicture-NamingProject(CRL-IPNP).In:Centerfor ResearchinLanguageNewsletter.LaJolla:UniversityofCaliforniaSanDiego; 200012.

[63]CabezaR,DennisNA.Frontallobesandaging:deteriorationandcompensation. In:StussDT,KnightsRT,editors.Principlesoffrontallobefunction.NewYork: OxfordUniversityPress;2012.

[64]CotelliM,ManentiR,CappaSF,GeroldiC,ZanettiO,RossiniPM,etal.Effectof transcranialmagneticstimulationonactionnaminginpatientswithAlzheimer disease.ArchivesofNeurology2006;63:1602–4.

[65]CotelliM,ManentiR,CappaSF,ZanettiO,MiniussiC.Transcranialmagnetic stimulation improves naming in Alzheimer disease patients at differ-entstages of cognitivedecline. European Journalof Neurology2008;15: 1286–92.

[66]CotelliM, Calabria M,Manenti R, Rosini S, Zanetti O, Cappa SF, et al. Improved language performance in Alzheimer disease following brain stimulation.Journal ofNeurology, Neurosurgeryand Psychiatry2011;82: 794–7.

[67]CotelliM,ManentiR,AlbericiA,BrambillaM,CossedduM,ZanettiO,etal. Pre-frontalcortexrTMSenhancesactionnaminginprogressivenon-ﬂuentaphasia. EuropeanJournalofNeurology2012;19(11):1404–12.

[68]CotelliM,FertonaniA,MiozzoA,RosiniS,ManentiR,PadovaniA,etal.Anomia trainingandbrainstimulationinchronicaphasia.Neuropsychological Rehabil-itation2011;21:717–41.

[69]SzaﬂarskiJP,VannestJ,WuSW,DiFrancescoMW,BanksC,GilbertDL.Excitatory repetitivetranscranialmagneticstimulationinducesimprovementsinchronic post-strokeaphasia.MedicalScienceMonitor2011;17:132–9.