UEF//eRepository
DSpace https://erepo.uef.fi
Rinnakkaistallenteet Luonnontieteiden ja metsätieteiden tiedekunta
2021
Primary hand motor representation areas in healthy children,
preadolescents, adolescents, and adults
Säisänen, Laura
Elsevier BV
Tieteelliset aikakauslehtiartikkelit
©2021 The Authors
CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/
http://dx.doi.org/10.1016/j.neuroimage.2020.117702
https://erepo.uef.fi/handle/123456789/24191
Downloaded from University of Eastern Finland's eRepository
ContentslistsavailableatScienceDirect
NeuroImage
journalhomepage:www.elsevier.com/locate/neuroimage
Primary hand motor representation areas in healthy children, preadolescents, adolescents, and adults ✩
Laura Säisänen
a,b,c,∗, Mervi Könönen
a,c,d, Eini Niskanen
c, Timo Lakka
e,f, Niina Lintu
e, Ritva Vanninen
b,d, Petro Julkunen
a,c, Sara Määttä
aaDepartment of Clinical Neurophysiology, Kuopio University Hospital, P.O. Box 100, 70029 KYS, Kuopio, Finland
bInstitute of Clinical Medicine, University of Eastern Finland, Finland
cDepartment of Applied Physics, University of Eastern Finland, Kuopio, Finland
dDepartment of Clinical Radiology, Kuopio University Hospital, Kuopio, Finland
eInstitute of Biomedicine, Faculty of Health Sciences, University of Eastern Finland, Finland
fDepartment of Clinical Physiology and Nuclear Medicine, Kuopio University Hospital, Kuopio, Finland
a r t i c le i n f o
Keywords:
Transcranial magnetic stimulation Neuronavigation
Human physiologic maturation Development
Motor mapping Overlap
a b s t r a ct
Thedevelopmentoftheorganizationofthemotorrepresentationareasinchildrenandadolescentsisnotwell- known.Thiscross-sectionalstudyaimedtoprovideanunderstandingforthedevelopmentofthefunctional motorareasoftheupperextremitymusclesbystudyinghealthyright-handedchildren(6–9years,n=10), preadolescents(10–12years,n=13),adolescents(15–17years,n=12),andadults(22–34years,n=12).
Theoptimalrepresentationsiteandrestingmotorthreshold(rMT)fortheabductorpollicisbrevis(APB)were assessedinbothhemispheresusingnavigatedtranscranialmagneticstimulation(nTMS).Motormappingwas performedat110%oftherMTwhilerecordingtheEMGofsixupperlimbmusclesinthehandandforearm.
Theassociationbetweenthemotormapandmanualdexterity(boxandblocktest,BBT)wasexamined.The mappingwaswell-toleratedandfeasibleinallbuttheyoungestparticipantwhoserMTexceededthemaximum stimulatoroutput.Thecenters-of-gravity(CoG)forindividualmuscleswerescatteredtothegreatestextentinthe groupofpreadolescentsandcenteredandbecamemorefocusedwithage.Inpreadolescents,theCoGsintheleft hemispherewerelocatedmorelaterally,andtheyshiftedmediallywithage.Theproportionofhandcomparedto armrepresentationincreasedwithage(p=0.001);intherighthemisphere,thiswasassociatedwithgreaterfine motorability.Similarly,therewaslessoverlapbetweenhandandforearmmusclesrepresentationsinchildren comparedtoadults(p<0.001).Therewasaposterior-anteriorshiftintheAPBhotspotcoordinatewithage, andtheAPBcoordinateinthelefthemisphereexhibitedalateraltomedialshiftwithagefromadolescenceto adulthood(p=0.006).Ourresultscontributetotheelucidationofthedevelopmentalcourseintheorganization ofthemotorcortexanditsassociationswithfinemotorskills.ItwasshownthatnTMSmotormappinginrelaxed musclesisfeasibleindevelopmentalstudiesinchildrenolderthansevenyearsofage.
Abbreviations
APB abductorpollicisbrevis ADM adductordigitiminimi BB bicepsbrachii BBT Boxandblocktest CoG Centerofgravity CST corticospinaltract ECR extensorcarpiradialis FCR flexorcarpiradialis FDI firstdorsalinterosseus
fMRI functionalmagneticresonanceimaging
✩ Someoftheseresultshavebeenpresentedinabstractformpreviously.
∗Correspondingauthorat:DepartmentofClinicalNeurophysiology,KuopioUniversityHospital,P.O.Box100,70029KYS,Kuopio,Finland.
E-mailaddress:laura.saisanen@kuh.fi(L.Säisänen).
MEG magnetoencephalography MEP motorevokedpotential MRI magneticresonanceimaging M1 primarymotorarea MSO maximumstimulatoroutput nTMS navigatedTMS
rMT restingmotorthreshold SCD scalptocortexdistance SMA supplementarymotorarea S1 primarysensoryarea
https://doi.org/10.1016/j.neuroimage.2020.117702
Received27August2020;Receivedinrevisedform16December2020;Accepted19December2020 Availableonline30December2020
1053-8119/© 2021TheAuthors.PublishedbyElsevierInc.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/4.0/)
1. Introduction
Thecorticalmotorareas includetheprimary motorcortex(M1), theregionresponsiblefortheexecutionofmovementanditsconcrete aspects,aswellaspremotorareasthatprovidecognitive,sensory,or motivationalinputsformotorbehavior(comprehensivelyreviewedby PicardandStrick,2001).Thesensorycortexistightlyconnectedtothe motorcortex(Eidelberg1969;Nudoetal.,1995;Teraoetal.,1995), andtogetherwithpremotor cortex,they areconsideredasthemain executiveloci forsimple voluntarymovements (Gerloff et al., 1998; Wittetal.,2008).Thereisbothelectrophysiologicalandfunctionalcon- nectivitybetweentheM1andpremotorareasaswellassupplementary motorareas(SMA),parietalcortexandcerebellum(Wesseletal.,1997; Akkaletal.,2007;Narayanaetal.,2012;Genonetal.,2017).Connectiv- itystudieshaverevealedthatdorsalpremotorareasconstituteamosaic withseveralfunctionssuchasmotorlearning,imageryandplanningof motortasks,andactionformulation,whichisrelatedtohandprefer- ence(Legaetal.,2020).Thisinterconnectivityisthoughttoprovidethe flexibilitynecessarytomodifytheexistingnetworktoaccommodatea behavioralchange(SanesandDonoghue2000).
Theprimarysomatosensoryandmotorcorticesaresomatotopically organized,suchthatspecificbodypartsarerepresentedseparatelyand adjacenttootherbodyparts,resultinginabodymap(PenfieldandBol- drey1937;Schott1993;Plowetal.,2010;CardandGharbawie2020).
Large-scalesomatomotororganizationwithtopographicscaffoldingisa fundamentalprincipleofearlydevelopment.Thisphenomenonises- tablishedprenatally andprovidestheprotoarchitecture of theentire brain. Itboth directsand constrainsexperience-drivenmodifications withchangesinconnectivity(Dall’Orsoetal.,2018;Arcaroetal.,2019).
Therelativeposition,distance,andoverlapbetweenbody-partrepresen- tationsconstituteasomatotopiclayoutthatdiffersinthedifferentter- minalsofthesensorimotornetwork(Wassermannetal.,1992).Somato- topyisalsofoundinthecerebellarcorticesandputamen(Hahamyand Makin 2019). Functional somatotopy is a balance between discrete peaksofindividualmuscles,theirdistributionaswellasawithin-limb overlapofrepresentations (Schabrunetal., 2015).However,theex- actsomatotopy is also questionedandchallenged with actionmaps (Graziano 2016).There is a considerableoverlap between thehand andforearmterritories in healthyindividuals (Marconiet al., 2007; Plowetal.,2010),potentiallyprovidingopportunitiesforcoordinated movementsandefficientsynergies,whilethesomatotopicdistinctive- nessofcentersofwithin-limbrepresentationsmaybeinvolvedinen- suringfinecontrol.Thepresenceofmultipleareasofhighexcitability (peaksinmaps)orsomatotopicallydiscretecentersisthoughttoreflect thepotentialforsynergistic,intermuscularcoordinationandcomplex movementstrategies,andconversely,itisimportantforfinelyindivid- uatedmovement(Teetal.,2017).
Handedness is one behavioral trait that affects somatotopy (Nudo et al., 1992; Nicolini et al., 2019) providing clues to the asymmetrical organization of the human brain (Toga and Thomp- son2003; Kong etal., 2018). Theneuralbasis andtimingof hemi- sphericlateralizationduringdevelopmentarefarfrombeingfullyun- derstood(Wilsonetal.,2010;DennisandThompson2013;Cochet2016; Konget al., 2018).Inadditiontothe primaryandsecondary motor andsensoryareas(Cochet2016),astudyusingresting-statefunctional magneticresonanceimaginganddiffusiontractographycombinedwith connectivity-basedparcellationdetectedasymmetryintheorganization, functionsandconnectivitybetweenhemispheresinthedorsalpremotor cortices(Genonetal.,2017;Genonetal.,2018).Aneurodevelopmen- talTMSstudyhasrevealedadecreaseinasymmetrywithage,favoring earliermaturationofthedominanthemisphere(Garveyetal.,2003).In adults,trendstowardsalargerrepresentation inthedominanthemi- spherehave beendescribed in most publishedstudies(Triggset al., 1999;Coppietal.,2014;Chieffoetal.,2016),possiblyrelatedtobet- terdexterity,whereasinbothpreadolescentsandadolescents,atrend
towardareducedmotormapwasobservedfortheleft,dominanthemi- sphere(Grabetal.,2018).
Neuromotorfunctionplaysanessentialroleinnormalcognitivede- velopmentandisfrequentlyabnormalinchildrenwithdevelopmental disabilities.Noticeablegainsinmotorfunctionsaremadethroughout theearlyschoolyears;thefine-tuningandcontinuedimprovementof motorskills,reflectedasbetterqualityandspeedofmotormovements, occurupuntiltheageof30years(Fietzeketal.,2000).Themostdy- namicperiodofmotorperformancedevelopmentendsat10to12years ofage(Fietzeketal.,2000).Instructuralterms,corticalthicknessinM1 attainsitspeakattheageofaboutnineyears,followedbytheSMA(~10 years)andmostofthefrontalpole(Shawetal.,2008).Thereafter,the corticalthicknessstartstotaper-off,andthisprocessasymptotesaround 14yearsofage(Vandekaretal.,2015).However,thecorticospinaltract (CST)reachesitsfullymaturestateearlierthanotherwhitemattertracts (DennisandThompson 2013).TheCSTexhibitsa leftwardasymme- try;itsmaturationrevealsdifferencesbetweenthesexes,buttheredoes notseemtobeanyrelationshiptoage-relatedchangesinmanualskills (Herveetal.,2009).
Themotormapscanbeassumedtobefundamentallythesamein childrenandadults,butlittleisknownregardinghowlocalnetwork propertieswithinM1evolveduringmaturationandhowtheseareas- sociatedwithconcomitantchangesinthemotorrepertoire.Thereare somekeyfindingsemergingfromanimalmodelsontheuse-dependent mapplasticityrelatedtothelearningofmotorskills(Nudoetal.,1996; Kleimetal.,1998).Ithasbeensuggestedthattheemergenceoffinemo- torcontrolisassociatedwitharelativebroadeningofconnectivitybe- tweenfunctionallydiversecorticalmotorneuronsandchangesinsynap- ticpropertiesthatcouldenabletheemergenceofsmallerindependent networks(Bianeetal.,2015;Arcaroetal.,2019).However,thispro- posalisatoddswithotherevidencefromanimal experimentswhich indicatethattheverybroadmotorconnectivityanddistributionsinthe fetusandnewbornappeartobetightlyrefinedduringearlydevelop- ment(probablyduringthefirst2–3years(Martin2005).Neuroplastic changesduetophysicalpracticeormentalrehearsalmayleadtoeithera reductionorexpansionofcorticalrepresentationsofactivelyusedmus- cles(Schieber2001;Kleimetal.,2004;Adkinsetal.2006;Vaaltoetal., 2013)ortosomekindofalteredoverlap(Tycetal.,2005).
Inhumans,thefunctionalmotormapandits developmentalplas- ticitycanbeassessedbyseveralnon-invasivebrainmappingmethods (Narayanaetal.,2015a).Neuronavigatedtranscranialmagneticstim- ulation (nTMS)isonesuchformofadirectmethod.TMShasanev- ident potential with both diagnostic andtherapeutic applications in pediatric neuropathologies (Frye et al.,2008; Hameedet al., 2017), suchasforassessingplasticityinprenatal,perinatal,orpediatricstroke (Waltheretal.,2009;Staudt2010;Kirtonetal.,2016).ThoughnTMS mapping hasprovedadvantageous intheclinicalexaminationof the presurgicalevaluationoftheeloquentareasinbothchildrenandadults (Säisänen et al., 2010; Vitikainen etal., 2013; Kaye andRotenberg 2017),themotormapshavenotbeen determinedatdifferent stages ofdevelopmentinthehealthybrain.Developmentalstudiesusingother methodshaveshownhigheractivationofthebilateralsensorimotorcor- tex, parietalareas,theSMAandthecerebelluminadultsincompar- isontochildrenusingfunctionalmagnetic resonanceimaging(fMRI) (Mall et al., 2005). On the other hand, a magnetoencephalography (MEG)studythat examinedthegeneratorsandareasfor motorcon- trolintypicallydevelopingchildrenandadolescents,foundevidence fortheinvolvementofSMAandcerebellarcorticesinadditiontoM1 (Wilsonetal.,2010).
Theaimofthisstudywastoassessthefunctionalcorticalrepresen- tationofupperextremitymusclesatrestinhealthyright-handedindi- vidualsatdifferentstagesofdevelopmentfromschoolagetoadulthood byusing nTMS,especiallyduringthecriticalperiodforthedevelop- mentof finemotorskillsi.e. preadolescence.Themotormapresults werecorrelatedwithmanualdexterity.Basedontheresultsfromprevi-
Table1
Genderandage,handedness,scalp-to-cortexdistanceandmanualdexterityscores(numberofblocksmovedinoneminute),restingmotorthresholdsas maximumstimulatoroutput(%).Mean±SD(range).Significanceindicatesdifferencesbetweenagegroups(ANOVA,post-hocSIDAK).Significantasymmetrical differenceswithinagegroupareindicatedwithboldfont(pairedt-test).
Children( n = 10) Preadolescents( n = 13) Adolescents( n = 12) Adults( n = 12) Significance
Gender ( male/female ) 5 / 5 7 / 6 6 / 6 6 /6 –
Age (years) 7.7 ± 0.4
(6.8–8.3)
10.9 ± 0.4 (10.2–11.8)
15.8 ± 0.8 (14.3–17.0)
28.0 ± 3.8 (22.3–33.7)
F = 232.07 p < 0.001 a,b,c,d,e
Edinburgh handedness 31.1 ± 8.5
(11–40)
32.7 ± 5.7 (26–36) [ n = 3 ]
28 ± 5.2 (23–40)
29.0 ± 7.1 (17–40)
F = 0.401 p = 0.753 Scalp-to-cortex distance
(mm)
Left hemisphere 7.89 ± 1.37 (5.3–9.7)
9.41 ± 0.82 (7.3 –10.3)
12.46 ± 1.85 (9.1 –14.6)
15.23 ± 2.50 (11 –19.5)
F = 34.47 p < 0.001 b,c,d,e Right hemisphere 7.44 ± 1.26
(5.1–8.8) 8.28 ± 1.08
(6.4 –10.0) 10.93 ± 1.66
(7.3 –13.4) 12.78 ± 2.57
(9.0 –16.7) F = 20.95 p < 0.001 b,c,d Box and block test (score) Left hand 51 ± 8
(43–66)
65 ± 6 (56–75)
73 ± 11 (53 –87)
83 ± 8 (71–105)
F = 27.67 p < 0.001 a,c,d,e Right hand 55 ± 6
(46–65)
65 ± 6 (55–75)
77 ± 14 (56 –94)
84 ± 7 (75–98)
F = 23.37 p < 0.001 a,b,c,d Resting motor threshold (%) Left hemisphere 67 ± 17
(42 – 96) 52 ± 12
(31 – 75) 42 ± 8
(30 – 54) 41 ± 7
(32 – 54) F = 12.42 p < 0.001 a,b,c Right hemisphere 68 ± 11
(55 – 94)
53 ± 11 (30 – 73)
40 ± 8 (25 – 56)
40 ± 6 (32 – 49)
F = 21.15 p < 0.001 a,b,c,d,e achildrenandpreadolescents,.
b childrenandadolescents,.
cchildrenandadults,.
d preadolescentsandadolescents,.
epreadolescentsandadults,fadolescentsandadults.
ousdevelopmentalandanimalstudiesonmotorlearningandplasticity, wehypothesizedthattopographicmapswouldbeessentiallythesame betweenchildrenandadultsthough therelativesizeofmotorrepre- sentationareasmightchange,andinadditionthattheoverlapbetween handandarmrepresentationmightincreasewithdevelopment.Further- more,themapmeasureswerecorrelatedwiththeimprovementsinfine motorabilities.
2. Materialsandmethods 2.1. Participants
ThisstudywascarriedoutinthepremisesofKuopioUniversityHos- pital,intheDepartmentsofClinicalRadiologyandClinicalNeurophysi- ology,andispartofanearlierstudy(Määttä etal.,2017;Säisänenetal., 2018).Thirty-fivehealthyright-handedparticipants(range7–17years), withthegendersuniformlydistributed,werestudied(demographicsin Table1)andcomparedwith12youngadults(22–33yearsold).The participantsintheyoungestagegroupswererecruitedfromageneral populationsampleofpredominantlynormal-weightchildrenfromthe cityofKuopio.Adolescentswererecruitedfrompupilsinthe8thgrade fromthenearestcomprehensiveschool.Oneambidextrousboywascon- sideredinthisstudyasright-handed.Theexclusioncriteriawereneuro- logicalorpsychiatricdisorders,previouscentralnervoussystem(CNS) infectionortrauma,medicationswithknownCNSeffects,oranycon- traindicationofTMS(Rossietal.,2009).Allparticipantswereinformed aboutthenatureofthestudy.Afterhavingreceivedadetaileddescrip- tionoftheprocedure,theparticipantsprovidedwritteninformedcon- sent.Consentwasalsoprovidedfromtheguardianinthecaseofapar- ticipantbeingunder15yearsofage.Thestudywasapprovedbythe ResearchEthicsCommitteeof theHospitalDistrictof NorthernSavo (48/2010).Allproceduresperformedwereinaccordancewiththeeth- icalstandardsoftheinstitutionaland/ornationalresearchcommittee (ethicalpermission48/2010).
2.2. Motortask
TheBoxandBlockTest(BBT)wasusedtoassessmotorspeedand skill(Mathiowetzetal.,1985).Thistaskrequirestheparticipanttomove
asmanyblocksaspossiblewithin60sfromonesideofaboxtothe other.Eachhandwasassessedseparately,beginningwiththedominant hand.
2.3. MRimaging
Subjectswerescannedwitha3.0TMRI-scanner(PhilipsAchieva TX;PhilipsHealthcare,Eindhoven,TheNetherlands).Structuralthree- dimensional T1-weightedMR-imageswereacquired(TR8.07ms,TE 3.7ms,flipangle8°,1×1×1mm3resolution)forTMSnavigation.An experiencedneuroradiologistscreenedallthestructuralMRIsforfocal changesbeforenTMSexamination.Scalp-to-cortexdistance(SCD)was assessedinmmtothedepthofgraymattersurfaceusingthenavigation software(Määttä etal.,2017).
2.4. NavigatedTMS
nTMS was performed with an eXimia stimulator and a biphasic figure-of-eight coilcombinedwitha navigationsystem(3.2 research version,NexstimPlc.,Helsinki,Finland)inbothhemispheresinaran- domizedorder.Amorethoroughdescriptionofthestimulatorsetupis providedin ourpreviouspaper(Säisänen etal.,2018).TMS-induced motor evokedpotentials(MEPs) wererecordedusing disposableAg- AgCl surfaceelectrodes placedontheabductorpollicisbrevis(APB), abductordigitiminimi(ADM),firstdorsalinterosseus(FDI),extensor carpiradialis(ECR),flexorcarpiradialis(FCR),andbicepsbrachii(BB) using abelly-tendon montage.Throughoutthemeasurement, muscle activitywasmonitoredon-lineandrecordedbystimulus-lockedEMG (NexstimPlc.,Helsinki,Finland).First,theoptimalcorticalrepresenta- tionsite(“hotspot”)oftheAPBwasdetermined(Säisänenetal.,2008).
ThehotspotwasthestimulationsitewheretheMEPsofgreatestam- plitudewereelicitedrepeatedly.Atthatsite,byusingtheoptimalcoil orientation,theindividualrestingMT(rMT) wasdeterminedusinga thresholdhuntingparadigmTMSMotorThresholdAssessmentTool2.0 (Awiszus2003;AwiszusandBorckardt2012) asapercentageof the maximumstimulator output(%-MSO).Mappingofmotorrepresenta- tionareaswasperformedatthestimulationintensityof110%ofrMT oftheAPBwiththeaidofagrid(size5×5mmpersquare)thatwas
Fig.1. AExample ofthe mappingat 110%
of rMT using the grid targeted around the hotspot.BExampleoftheresultingmap:white dotsare locationselicitingMEP response in anyupperlimbmuscle;graydotsarenegative sites.Thearrowshowsthehotspotwithelectric fieldorientation.Orangedashedlinesindicate thecentralsulcus.CEMGwasrecorded;run- ningonline(leftpanel)andTMS-triggeredre- sponsesindividuallyfortheAPB(uppermost), FDI,ADM,ECR,FCR,BB,andFDI(thelow- est).Stimulusisgivenatthemomentofthe reddashedline.APB=abductorpollicisbre- vis;ADM=adductordigitiminimi;ECR=ex- tensorcarpiradialis;FCR=flexorcarpiradi- alis;BB=bicepsbrachii;FDI=firstdorsalin- terosseus.
individuallycenteredtothehotspot.Onestimuluswasappliedperspot, extendeduntiltherewasarim ofstimulationsiteselicitingno more MEPs(Fig.1)(Säisänenetal.,2015).Thecoilorientationwasapproxi- mately45° tomidline,perpendiculartothenearestsulcus,thesoftware ensuringoptimaltilting.Theinterstimulusintervalwas3to5s.Thedu- rationofthewholemeasurementsession(includingtheexplanationof theprocedure,attachingtheelectrodes,performingtheco-registration etc.)wasapproximatelytwohours.Thecoarsemappingatthebegin- ningof thesessionandthefollowingMTdeterminationlastedabout twentyminutesperhemisphere.Thereafter,thedurationsofbilateral motormappingsessionsrangedfrom9to33min,themeanwas18min dependingonthesizeofthemapandthesubjects’abilitytoremain relaxed.Atolerabilityquestionnairewasadministeredimmediatelyfol- lowingthesession(SupplementaryTable1).
2.5. Dataanalysis
TheMEPswithamplitudesof>50μVinrelaxedmuscleswereac- ceptedasresponses(Rossinietal.,2015;Groppaetal.,2012).Therep- resentationareasforhand(APB,FDI,ADM)andarm(ECR,FCR,BB) werecalculatedforeachsubjectusingthesplineinterpolationmethod (Julkunen2014),andtheirratiowasstudied.Therelativeoverlap(%) wasdeterminedastheareawherestimulationelicitedaresponseinboth handandarmmusclesdividedbytheunionofthehandandarmmus- clerepresentationarea(i.e.thetotalarea,atleasteitherhandorarm producedaMEP),theso-calledJaccardindex.
TheCenters-of-Gravity(CoGs)(Wassermannetal.,1992)formotor representationofeachhandmuscleweredeterminedusingtheMRIco- ordinatespaceof theeXimiasoftwarethat utilizestheLAS(i.e.left, anterior,superior)coordinatesystem,whichhasanoriginintheright- posterior-inferiorcorneroftheimage.TheindividualCoGcoordinates
andMRIswerespatiallynormalizedtostandardspace(origininanterior commissure,xcoordinatepositivevaluesinright,ycoordinateforward positive,zcoordinateuppositive)usingtheSPM8-softwarerunningon Matlab7.4(TheMathworks,Natick,USA).Twotemplateswereused dependingonage.Thetemplateusedforchildrenandpreadolescents wascreatedwiththeTOMtoolboxonSPMusingtheNIHreferencedata (Wilkeetal.,2008).Foradolescentsandadults,thestandardMNItem- plateprovidedbySPMwasused.Todeterminethevariationbetweenthe CoGswithinthegroups,ellipsoidsofthe90%confidenceintervalwere fittedtotheclustersoftheindividualsitesbyestimatingthelengthsand directionsoftheellipsoidmainaxisbasedonchi-squaredistributionus- ingMatlab(Niskanenetal.,2010).
2.6. Statisticalanalysis
ThedifferencesinrMT,SCD,andBBTscoresacrosstheagegroups were analyzedwithANOVA andthe posthoc Sidaktest. Interhemi- sphericdifferencesinrMT,SCD,andBBTscoreswithintheagegroups wereevaluatedwiththepairedt-testsincethesevariableswerenormally distributed.Therepresentationextents,ratiosofareasandoverlap(not normallydistributed,Kolmogorov-Smirnov)weretestedwithunivariate generallinearmodelforeffectsof age,hemisphereandtheirinterac- tion.TheCoGsofthemuscle-specificcoordinateswereevaluatedusing thenonparametricKruskal-Wallis testandposthocpairwisecompar- isonswithBonferronicorrection,andinterhemisphericdifferenceswith Mann-WhitneyUtest.TheLevenetestwasusedtoevaluatedthevaria- tionwithinagegroupsinCoGoftheAPBanterior-posteriorycoordinate andtherelativeoverlapwithintheagegroups.Spearmancorrelation andpartialcorrelationsadjustedforageorBBTwereusedtotesttheas- sociationsbetweenthehand/armratios,overlap,motordexterity,and age.AssociationsanddifferenceswithP-values<0.05wereconsidered
Table2
HotspotlocationsoftheAPBineachagegroupshownaspercentage(%).
Children Preadolescents Adolescents Adults
Precentral gyrus 60 50 71 92
Central sulcus 40 38 29 0
Postcentral gyrus 0 12 0 0
Premotor area 0 0 0 8
statisticallysignificant.Thestatisticalanalyseswereperformedwiththe SPSSsoftware,Version22(IBMCorporation,Somers,NY,USA).
3. Results
3.1. Safetyandtolerability
TheTMSwaswelltoleratedbyallparticipantswithnosignificant sideeffects.Minorcomplaintswerereportedbysevenparticipantsi.e.
tiredness(n=4),irritationofthestimulatedsite(n=2),andexcess compressionfromthetrackerband(n=2).
3.2. Hotspotlocations
TheAPBhotspotswereusuallyfoundintheprecentralgyrus(68%) orinthecentralsulcus(27%)(Table2).Therewasaposterior-anterior shiftwithage;thelocationsinthecentralsulcusorposteriortothat locationweremorecommonintheyoungeragegroups,whereasadults’
APBhotspotswerefoundintheprecentralgyrus(Table2).Themost commonlocationwasthehandknob,morespecifically,itslateralcorner (representativeexamplesfromeachagegroupareshowninFig.1).SCD increasedwithage(F=20.9ontherighthemisphere,20.8ontheleft, p<0.001)(Table1).SCDexhibitedastatisticallysignificantasymmetry (longerdistance ontheleft,dominanthemisphere)in allagegroups otherthaninthechildren(Table1).
3.3. Motorthresholdsandfeasibility
TherMTsoftheAPBareshowninTable1,andtheindividualrMTs andfeasibilitytoperform motormappingintheyoungest agegroup aredisplayedinTable3.TherMTwastoohightobeassessedinthe youngestparticipantaged6.8years.Despiteusingthemaximumstim- ulatoroutput in one participant(G4), 106% and 108% of rMT was reachedandusedformappinginsteadoftheintendedstimulusintensity, andforthissubject,onlytheresultsforCoGsareincludedinthemotor mapgroupanalysis.ThisparticipantalsohadMEPsofhighamplitude intheECRduringmapping.Basedonthisfinding,weperformedanad- ditionalrMTestimationusingamethodof50μV-levelrMTfortheECR muscle(Julkunenetal.,2011)inchildrentoensurethattherMTsinthe ECRwerenotlowerthanintheAPB(p=0.582forright,p=0.969for lefthemisphere,pairedt-test)(Table3).Theinput-outputcurveshave beenassessedearlier(Säisänenetal.,2018).ThemeanrMTfortheAPB didnotdifferbetweenthehemispheres,butsomesubjects(equallydis- tributedinallagegroups)exhibitedasubstantialinterhemisphericdif- ferenceinrMT(lefthemisphere– righthemisphere)rangingfrom−18 to19%-MSO.
3.4. Handandarmrepresentationextents,theirratio,andoverlap
Themotorrepresentationswerelocated mainlyontheprecentral gyrusspreadingtoregionsbothanteriorandposteriortothatlocation, butnotextendingtotheSMA(Figs.2and3).Motormapresultsatin- dividuallevelareshowninFig.3.Apreadolescentsubjectistakenas anexample,andhermuscle-specificmapsshowninFig.3B.Fig.3A showsthemappedareasbilaterally,andinthelefthemispheretheloca- tionswereMEPswereelicitedinallsixrecordedmuscles.Thesamekind ofmaponmulti-jointresponsesisshownforanother(adult)subjectin
Fig.3C.ThemappingresultsatgrouplevelareshowninTable4.In children,thehand/armratiowasbelowoneandincreasedwithage,i.e.
childrendifferedsignificantlyfromadolescentsandadults(F=6.157, p=0.001)(Table4).Therelativeoverlapofthehandandarmrepre- sentationswaslessinchildrenthanintheotheragegroups(F=7.864, p<0.001)(Fig.4).
3.5. CoG
Themuscle-specificnormalizedCoG-coordinatevolumesareshown inFig.5.Whenvisuallyevaluated,the90%confidentialellipsoidwas anatomicallynarrowandorientedintheantero-posteriordirectionin children.Inpreadolescents,theCoGsusedforfittingweremorescat- teredresultinginalargerellipsoid.Thereaftertheellipsoiddecreased insizeandconcurrently,itsshapechangedfromacircleinadolescents tobeingoval-shapedinadultsasaresultoftheuniformlyclusteredCoG locations.Inoneadult,alloftheCoGswereinpremotorareas.
TheCoGcoordinatefortheAPB,themainmuscleofinterest,was separatelyexamined(Table5).Themedio-lateralx-coordinateexhib- itedanage-relatedeffect:thisCoGmovedinamedialdirectioninthe lefthemisphere(p=0.006,posthocpairwisetestwithBonferronicor- rection betweenpreadolescentsandadults).Intherighthemisphere, thex-coordinatemovedslightlyinthelateraldirection,butthischange was not statisticallysignificant (Table 5).All othermuscles showed asimilareffectofageinthelatero-medialdirectionin thelefthemi- sphere,withtheposthocevaluationhighlightingrevealingthediffer- encebetweenpreadolescentsandadults(SupplementaryTable2).No shiftwasfoundintherighthemisphere.Byusingtheabsolutevalues ofthecoordinates,wedetectedinterhemisphericdifferencesinadoles- centsandadults(Table5).TheCoGfortheAPBwasmoremedialinthe lefthemispherethanintherighthemisphereinadolescentsandadults, andmoreposteriorinthelefthemispherecomparedtotherightinado- lescents.
Theantero-posteriory-coordinateshowedatrendtowardsanante- riorshiftwithageintheAPBintheleft(p=0.076,Table5),anda statisticallysignificantdifferencefortheECRandBB(Supplementary Table2).Intherighthemisphere,onlytheBBrevealedthiskindofan- teriorshiftassociatedwithage(p=0.004).Itisnoteworthythatthere isextensivevariationintheanterior-posteriororientation,specifically inpreadolescentsandchildren(F(3,42)=5.803,p=0.002ontheright, non-significantonthelefthemisphere).Adolescentsexhibitedasymme- try,withthey-coordinateontheleftbeinglocatedinamoreposterior location(p=0.045).
3.6. Correlationsbetweenmanualdexterity,ageandthemotormap parameters
Manualdexterityimprovedwithage(p<0.001,Table1).Thedexter- ityoftherighthandwasstatisticallysignificantlybetterthanthatofthe leftinchildrenandadolescents(t=2.483,p=0.017).Thehand/arm ratiointherighthemispherecorrelatedpositivelywiththeBBTscore ofthecontralateralhand(rho=0.522,p<0.001);atrendwasfoundfor thelefthemisphere(Fig.6).Thiscorrelationintherighthemispherere- mainedwhenadjustedforage(r=0.346,p=0.020),butagenolonger correlatedwiththehand/armratiowhenadjustedfortheBBTscore.
TherelativeoverlapoftherighthemispherecorrelatedwiththeBBT scoreforthecontralateralhand(rho=0.500,p<0.001),butintheleft hemispherethiswasevidentonlyasatrend(Fig.6).Thecorrelation disappearedwhenadjustedforagebutremainedforoverlapandage whenadjustedfortheBBTscore(r=0.306,p=0.041).
4. Discussion
Thiscross-sectionalstudydescribesthecorticalmaturationandde- velopmentofhandmotorrepresentationareasandtheirintrinsicorgani- zationalprinciplesusingnTMSmappinginanagerangefromchildhood
Table3
Individualrestingmotorthresholds(rMT)ofabductorpollicisbrevis(APB)inthechildren.TherMTstoohighformappingat110%ofrMTareindicatedwith emboldenedtext.MappingwasperformedatthemaximumstimulatoroutputinG4,correspondingto106%and108%ofrMT.Mappingwasnotperformed forB5andonlyontherighthemisphereforG2.
Subject Age (years) Mapping rMTfor APB(%) Estimated rMT for ECR (%)
L
hemisphere
R hemisphere
L hemisphere R hemisphere
G1 7.7 Both hemispheres 42 55 42 NA
G2 7.9 R hemisphere 96 77 NA NA
G3 8.3 Both hemispheres 65 66 60 61
G4 ∗ 7.8 Both hemispheres
at 100% MSO
92 94 NA NA
G5 7.4 Both hemispheres 64 60 60 54
B1 8.1 Both hemispheres 62 66 60 63
B2 7.3 Both hemispheres 59 62 60 59
B3 7.8 Both hemispheres 64 66 65 66
B4 7.7 Both hemispheres 61 70 56 NA
B5 6.8 no > 100 > 100 NA NA
B=boy,G=girl,ECR=extensorcarpiradialis,L=left,R=right,rMT=restingmotorthreshold.NA=notassessed.
∗Duringmapping,largeMEPswereelicitedinECRmusclesindicatingthattherMTofarmmuscleswasprobablylowerthanthatofAPB.
Fig. 2. Upper panel shows representative, qualitativeexamplesofCoGforeachmuscle.
Inchildren,theCoGsarescattered.Inanadult brain,theCoGsareclosetoeachotherinthe precentralgyrus,nearthelateralcornerofthe omega-shapedhandknob.Lowerpanelshows rawdataonindividualmapsineachagegroup.
RedincidatesMEPsinhand,blueinarmand purpleinboth.Thesubjectsinupperandlower panelarenotthesame,butrandomlychosen.
Fig.3.Themappingdataononepreadoles- cent. A All stimulated locations bilaterally.
Inthelefthemisphere thesiteswhereMEPs wereelicitedinallsixmusclesareindicated withbluesquares.ThesiteswhereMEPswere elicitedinanyofthemusclesareshownwith heatmap,white>1mV,yellow500–1000μV andred50–500μV.BMuscle-specificmaps.
CAsimilarmapshowingthemulti-jointre- sponses(MEPselicitedin allmuscles)foran adultsubject.
Table4
Motormappingresultsreportedasmean(standarddeviation)forhandandarmextentsandtheirratioandoverlapasabsoluteareaandaspercentage.
Significanceindicatesdifferencebetweenagegroups(generallinearmodel).Nointeractionwasfoundbetweenageandhemisphere.
Children Preadolescents Adolescents Adults Significance
Extent hand (cm 2) 6.83 (3.03) 6.90 (3.04) 8.26 (3.38) 8.62 (3.34) p = 0.158
Extent arm (cm 2) 10.18 (3.48) 7.68 (4.17) 7.75 (4.72) 7.90 (3.56) p = 0.195
Ratio hand/arm (-)
0.684 (0.24) 1.035 (0.43) 1.283 (0.59) 1.224 (0.48) F = 6.157
p = 0.001 b,c Overlap area
(cm 2) 5.86 (2.70) 6.32 (3.26) 6.61 (3.85) 6.74 (2.87) p = 0.857
Overlap (%) 0.47 (0.13) 0.63 (0.13) 0.61 (0.16) 0.69 (0.12) F = 7.864
p < 0.001 a,b,c achildrenandpreadolescents,
b childrenandadolescents,
cchildrenandadults.
Fig.4. AHand/armratioandBoverlapindifferentagegroups.
Fig. 5. Ellipsoids showing the locations of centers-of-gravity(CoGs)ofhandandarmmus- cleswith90%confidenceintervalinnormal- izedstandardbrainfromtwodifferentorien- tations.Thechildren’stemplatewasusedfor childrenandpreadolescents;theadulttemplate wasusedforadolescentsandadults.Inthetwo youngestagegroups,thevolumewasspread alsoposteriorlytothecentralsulcus.Thedots outsidethe adult ellipsoid (in the premotor area)werethoseofonemalesubject.Thesur- faceisthatintheborderbetweenthegrayand whitematter.Blackdotsandredellipsoidin- dicatesthehand,bluedotsandgreenellipsoid indicatesarmmuscles.
Table5
Normalizedcoordinates(inmm)forCoGoftheAPBpresentedasmean(standarddeviation).Originisinanteriorcommissure,xisthemedial-lateralorientation (leftbeingnegative)andyisanterior-posterior(negativeincreasingposteriorly).Significanceindicatesdifferencebetweenagegroups(Kruskal-Wallis,post- hocpairedtestwithBonferronicorrection).InterhemisphericdifferenceswereexaminedwithMann-WhitneyU,significantdifferenceswerefoundforxin adolescents(p=0.020),yinadolescents(p=0.045)andxinadults(p<0.001),showninboldfont.
Children Preadolescents Adolescents Adults Significance
x coordinate Left hemisphere − 36.3 (2.4) − 38.5 (5.4)
− 36.1 a (4.6)
− 32.7 c(2.8) p = 0.006 ∗ Right hemisphere 38.5
(2.4) 39.8
(4.9) 40.6
(4.2) 40.6 (4.2) p = 0.414
y coordinate Left hemisphere − 12.6 (7.0) − 17.1 (10.3)
− 14.9 b (3.9)
− 10.6 (4.2) p = 0.076 Right hemisphere − 10.2 (8.4) − 16.5
(8.3)
− 11.7 (3.2)
− 10.5 (3.4) p = 0.371 Thesignificance(∗)isbetweenpreadolescentsandadults.
Rho=0.522, p <0.001 Rho=0.500, p <0.001
Rho=0.266, p =0.077 Rho= -0.233, p =0.128
Hand/arm le hemisphere Hand/arm right hemisphere Overlap right hemisphereOverlap le hemisphere
3
2
1
3
2
1
1.0
0.4 0.8
0.2 0.6
1.0
0.4 0.8
0.2 0.6
40 60 80 100
40 60 80 100
Children Preadolescents Adolescents Adults
BBT le hand
BBT right hand
BBT right hand BBT le hand
Fig.6. Significantcorrelationsbetweenhand/armrepresentationareasandoverlapwithmanualdexterityofthecontralateralhandwerefoundintherighthemi- sphere.Nosuchcorrelationswerefoundintheleft,dominanthemisphere.Colorsindicatedifferentagegroups.
toyoungadulthood.Bothanexpansionandtheextensivevariationin theCoGsoftheupperlimbmusclerepresentationswerefoundinchil- drenaged10—12years,which isthetimewindowwhenfinemotor abilitiesimprove.Thereafter,theCoGsofthemotorrepresentationsbe- camemoreconcentratedwithage.Inthelefthemisphere,theCoGswere locatedmorelaterallyinpreadolescentsandthenshiftedmediallywith age.Thehandandforearmmusclerepresentationratioincreasedwith development;intherighthemisphere,thisisassociatedwithgreaterfine motorability.Thehandandforearmmusclerepresentationsoverlapped lessinchildrencomparedtootheragegroups.
4.1. Hotspotlocation
Inadults,theAPBhotspot wasfoundin theexpectedlocationin theprecentralgyrusoranteriortothatlocation,usuallyinthelateral cornerofthehandknob(Ahdabetal.,2016;Reijonenetal.,2020a).
Our observationof thepremotor hotspot inone adult is notunique (Ahdab etal.,2016).Inchildren,thehotspot wasmostoftenlocated inthecentralsulcus,andtherewasaposteriortoanteriorshiftwith age.Wealsoobservedanunexpectedposteriortoanteriortrendwith age,similartothatoccurringin thelocationofthehotspot.Previous developmentalworkhasindicatedthatboththeCoGcalculatedbased onthemotormapsaswellasthehotspotarevaluableparameters,pro- vidingarobusttoolforestimatingthetargetedsite(Grabetal.,2018).
Intheirseminalmodelingwork,Foxandcolleaguesappliedacolumn- basedmodelandfoundthehotspotlocationinadultstobedeepinthe sulcus(Foxetal.,2004).Lateron,thislocationwasspecifiedtotheante- riorwallofthecentralsulcustowardthegyrallip,thoughsubsequently itwasclaimedthatseveralissuessuchasaxonbendingneededtobe considered(Gomez-Tamesetal.,2020).Therewasanexpectedeffectof
ageontheSCD,andtheusedsimplifiedsphericalEFmodel,insteadofa realisticheadmodel,mayhaveintroduceduncertaintiesintothemap- pingresult(Beauchampetal.,2011;Danneretal.,2012;Julkunenetal., 2012).
4.2. Anatomicalmaplocation
Anatomically, thehand motor representationswere foundin M1 aroundthehandknob,extendingslightlyfrontallytothepremotorar- eas.However,thisismainlyduetothespreadoftheelectricfieldand notaseparateconnectionfromtheremotepremotorarea(Teittietal., 2008).Inourstudy,acoilorientationofapproximately45° wasused in allsubjects. TheEF modelingstudyrevealedthat thepreferential coil orientationsareeitherperpendiculartothegyrusortoward the CoGorthehotspotonthetopofthegyriandfurthermore,coilorienta- tioncanbecrucialfortheaccuracyofmotormapping(Reijonenetal., 2020b).Thefunctionalmotormapswerenotextendedtonon-primary motorcorticessuchastheSMA,whichhasanimportantroleincom- plex movement (Shibasaki 2012). In practice, short bursts of high- frequency TMSstimulationareneededtotransientlydisturbthemo- tor functionwhenassessingtheeffectsontheSMA(Schrammetal., 2019).Activityin theSMA(andcerebellum)in additiontothecon- tralateralprecentralgyrushasbeenobservedinbrainsundergoingmat- urationamongchildrenaged8—15yearsusingasimplemotortaskin anMEG(Wilsonetal.,2010).AnfMRIstudyusingwriststimulation detectedlarge activationareasincludingtheSMAinpreterminfants (Dall’Orsoetal.,2018).Instead,anotherfMRIstudyrevealedlargerac- tivatedareasincludingtheSMAandcerebelluminadultsincomparison tochildren(Malletal.,2005).Notonlythetype(precisionvs.power grip)oftask(Ehrssonetal.,2000),butthemodalities(fMRIvs.TMSvs.
MEG)andfunctionalconnectivitynetworksdifferandtheresultsare notdirectlycomparabletootherpublishedreports(Wangetal.,2020; WeissLucasetal.2020).
4.3. Topographicspecificity,actionmapsandmotorprimitives
Whenexaminingtheindividualmapsineachagegroup,children’s mapsweremoresporadicordiverse,thoseofpreadolescentswerelo- catedpostcentrally,inadolescents,themapswerelarger,whereasin adults,theyweremuchmorefocusedandclusteredneartothehand knob. However, the individual variation in motor maps was exten- sive,andcaution isnecessarywheninterpreting ourresults.Our re- sultsofdiversemapsmaybe inlinewithahighlyinterestingbranch ofresearch thatquestionsthetopographicorganizationofthemotor cortex,andinsteadsuggestsfunctionallydistinctareas (actionmaps), that are based on coordinated effects on animal’s behavior, for re- view see (Graziano 2016). Complex motor primitives exhibit inter- speciesdifferences,butareconnectedtoethodologicallyrelevantbe- haviors(Desmurgetetal.,2014).Theresearchersweresurprisedtofind anonuniformrepresentationofsynergiesbyusingintracorticalmicros- timulation,wheremovementstendedtoconvergetowardparticularpos- tures(Overduinetal.,2012).These actionmaps,onein theparietal cortex,oneinthemotorcortexalsoshowconnectivitybetweenthem.
Theyarepartlyshapedbyexperience,andexhibitreorganizationand capabilityofdevelopingnewzones.However,itisstillunclearwhether theyarelargelyfixedafterdevelopmentordoesitcontinuouslychange withlearning.
4.4. Handandarmrepresentationratios
Regardingthehandandforearmmusclerepresentationratios,the childrenshowedlargerarmthanhandareas,andtheproportionofthe handincreasedwithageaccordingtoourhypothesis.Inaseminalmap- pingstudyofdistalandproximalmusclesinadults,itwasobservedthat theAPBmapwaslargerthantheFCR(Wassermannetal.,1992)whereas insquirrelmonkeys,thearmareawaslargerthanthehand(Cardand Gharbawie2020).Inourstudy,adolescentsexhibitedextensivevaria- tioninthismeasure,whichmaysuggestthatadolescenceisaperiodof dynamicplasticityinthemotorareas.Ithasbeenclaimedthatcortical thinning,whichisguidedbyagenetictimetableandbyexperience,pre- dominatesinadolescenceandincreaseswellbeyondadolescenceinto middleageandisassociatedwithwhitematterdevelopmentandmyeli- nation(Caseyetal.,2005;Gieddetal.,2009;Paus2010;Vandekaretal., 2015).Thoughagedependencyinthehand/armratiowasobserved, thisdependencydisappearedwhencontrolledfortheBBT,suggesting thatthehand/armratiocorrelateswithdexterityratherthanwithage.
Earlier,wedetectedsimilarpronouncedarmrepresentationintheleft, dominanthemisphereinsubjectswithAspergersyndromeascompared totypicallydevelopingpreadolescentboys,andthiswasrelatedtotheir poorerdexterity(Säisänenetal.,2019).Thefiner-scaledifferentiation oftheindividualfingersincreases(i.e.,topographicrepresentationsare refined)overthefirsttwoyears(duringearlydevelopment),coinciding withthematurationoffinemotorskills(Arcaroetal.,2019).Inastudy conductedwithyoungandoldmice,ashortdurationofreachtraining increasedtheareaofproximalforelimbmovementrepresentationsat theexpenseofdistalrepresentationsintheyounganimals,thisreorga- nizationmayhavebeenpartiallymediatedbyalong-termpotentiation (LTP)-likemechanismandsynapticpruning(Tennantetal.,2012).
4.5. Overlapbetweenhandandarmmuscles
Theoverlapisamapmetricsrelatedtothehand/armratio.Inour study,theoverlapbetweendistalandproximalmusclesvariedbetween 40and80%,inlinewithpreviouspublications(Melgarietal.,2008; Chieffoetal.,2016).Childrendisplayedlessoverlapthanthoseinthe olderagegroups.Thematurationoftheoverlapappearstooccurearlier
(inpreadolescence)thanthehand/armratio(atadolescence).Weob- servedapositivecorrelationbetweentheoverlapandmotorskills,but thiscorrelationdisappearedwhencontrolledforage.Ontheotherhand, theoverlapcorrelatedwithagewhencontrolledfordexterity,suggest- ingthattheoverlapintherighthemispherewasdirectlyrelatedwith ageratherthanindirectlyduetodexterity.Astructuralplasticitystudy examiningmotorlearninginratshasrevealedthatdistalandproximal forelimbprojectingneuronsareintermingledinthemotorcortexwith anoverlappingdistribution,whichdoesnotchangeasafunctionofskill acquisition(Wangetal.,2011).Theoverlapseemstobehighlyimpor- tantformovementcoordination,whilethesomatotopicdistinctiveness ofcentersofwithin-limbrepresentationscouldensurefinelyindividu- atedcontrol.Ithasbeensuggestedthatinpianoplayers,thelessex- tensiveoverlapinthedominanthemisphereandthereducedmaparea arereflectionsofthelong-termplasticityrelatedtomotorlearningof askill(Chieffoetal.,2016).Althoughseveralaspectsoftheoverlap— thetask-specificity(Masse-Alarieetal.,2017),training-relatedplasticity (Tycetal.,2005;Vaaltoetal.,2013),aswellasitspresenceinpatholog- icalconditionssuchaschronicpain(Schabrunetal.,2009;Tsaoetal., 2011;Schabrunetal.,2015)— havebeenstudied,theirinterpretation stillneedsfurtherclarification.
4.6. Changeinthedistributionofmuscle-specificCoGswithage
Thescatterofmuscle-specificCoGswas enlarged(mainlyposteri- orly)inpreadolescentindividualsascomparedtochildrenorthosein theolderagegroups.Thisreorganizationcoincideswithadynamicpe- riodinfinemotorbehavior.Ourpreviousinvestigationsintoexcitabil- ityandmaturationrevealedthattheexcitabilitycurvedidnotreach itsplateauinpreadolescents(Säisänenetal.,2018),whichmaypointto thepotentialforplasticity.Preadolescence,i.e.theso-calledcriticalwin- dowformotorfunctions(Shawetal.,2008),occursinparallelwiththe rapiddevelopmentofexecutiveskillsofplanning,workingmemory,and cognitiveflexibility,whereasmature“cognition” ischaracterizedbythe abilitytofilterandsuppressirrelevantinformationandactionsinfavor ofrelevantcues.Allaspectsofthemovementssuchasmovementiniti- ation,gripforceadaptation,andmaximumspeedofmovementarenot reachedatexactlythesametime(Forssberg1999;Fietzeketal.,2000).
OurfindingofscatteredCoGsinpreadolescence correspondswithan imagingstudythatobservedadditionalnodeswhichhavebeenpostu- lated toreflecttheneuraldynamicsunique tothematuratingmotor system,probablytoservethesomewhatelementarymovementspresent duringthepreadolescentyears(DennisandThompson2013).Astruc- turaldevelopmentalstudyonthecorticospinaltracthasshownthatat theageof11,theCSToriginwaslocatedinbothpre-andpostcentral gyrus,whereasanexclusivepre-orpostcentraloriginwasmorecom- moninyoungerchildren(Kumaretal.,2009).Ontheotherhand,an- otherstudyrevealedthatthematurationaldirectionintheCSTorigin wasfromanteriortoposteriorandincreasedcomplexitywithage,con- trastingourresults(Kwonetal.,2016).
4.7. Maturationofmapmetricsandmotorlearning
WeobservedtheCoGstobecomemoreconcentricwithage,andthe CoG ofourmainmuscleofinterestshiftedmediallywithageon the lefthemisphere.Areorganizationofthemotormapwasnecessaryfor theacquisitionofaskill(motorlearning),butthemaintenanceofthe reorganizedstatewas notnecessaryforthemaintenanceof thatskill (motorperformance)(Tennantetal.,2012).Lowermovementthresh- oldsmaybeassociatedwithincreasedsynapticefficacy,andsynaptic potentiationmaybemaintainedafterthemaphasreturnedtocontrol levels(Tennantetal.,2012).Duringearlyadulthood(20to30years ofage),atimeofoptimalbrainhealthandbehavior,spontaneousbeta activityofthemotorcortexasstudiedbyanMEGwasatitsminimum, whichwassuggestedastobeindicativeofgreaterneuronalefficiency
(Heinrichs-Grahametal.,2018).Adultshaveshownbothgreaterinte- grationwithinnetworksandgreatersegregationbetweennetworksthan isthecasein children(DennisandThompson2013).Regionswhose brainactivitycorrelateswithtaskperformancebecomemorefocalor fine-tuned,whereasotherregionsnotinvolvedintheaspectsofthebe- haviorthatundergospecificrefinementremainunchangedbytheexpe- rienceorevendecreaseinactivitywithage(Caseyetal.,2005).There isadynamicinterplayofsimultaneouslyoccurringprogressiveandre- gressiveevents;thegradualeliminationofexcessiveconnections,with theestablishmentandstabilizingofnascentsynapsesandstrengthening ofrelevantconnectionswithdevelopmentandexperience(Caseyetal., 2005).Thematurationofsuperficialwhitemattercontinuesuntilthe ageof18years(Wuetal.,2014).Thereducedinterconnectivitymay alsogenerateagreaternumberofindependentnetworks,supportingen- hancedfractionationandresolutionofmovement(Bianeetal.,2015).
Whenataskhasbeenextensivelypracticed,fewerneuronswithinthe motorcortexneedtobeactivetoproducethesamemovementwhile thenumberofsynapsesisincreasedonlyinthereorganizedareasofthe motorcortex.Ourpreviouslarge-scalestudyin youngandoldadults exploitingthesamemethod,revealedthelocationsoftheAPBmuscle hotspotsaroundbothM1andS1(Niskanenetal.,2010).Bothpre-and postcentralgyricontaincorticospinalcells,andthemapsdonotstrictly adhere toarchitectonic borders (Groppa et al., 2012; Arcaro et al., 2019).IthasbeenspeculatedthatbilateralS1-M1mightplayanim- portantroleinthe preparationandexecution ofcomplex movement (Shibasaki2012).
4.8. Mapcorrelationswithmotortasks
Asabehavioralcorrelateoffinemotorskillperformance,weuseda rathersimpletaskoftheBBTthatinvolvestheactivationofhandand forearmmuscles,andobservedanexpectedeffectofage.Thedexter- ityofthelefthand continuedtoimprovefromadolescencetoadult- hood,whereasthedominantrightonedidnotimprovefurther.Pread- olescentsandadultswereequallygoodwithbothhands;childrenand adolescentswerebetterwiththeirdominanthands.Thereareseveral alternativefinemotortestssuchasthenineholepegtest(NHPT),the fingertappingtest,orpinchgriptests(Bashiretal.,2014;Chieffoetal., 2016;Masse-Alarieetal.,2017;Grabetal.,2018;Sirkkaetal.,2020).
Thesetestsmayevaluatedifferentaspectsofthemotorfunctionandalso revealdifferencesbetweenthesexesduringmaturation(Herveetal., 2009).Overall,inter-individualvariationisamajorfeaturewithdiffer- entmotorproficiencytaskstypicallyencounteredindevelopingyoung childrenagedfromthreeto18yearsofage(Kakebeekeetal.,2018).No correlationswerefoundrelatedtotheself-reporteddegreeofhandpref- erence,butitisnoteworthythatallofourtestedsubjectswereclearly right-handed.
4.9. Interhemisphericdifferencesandasymmetry
Wedidnot observeasymmetry in rMT.Theincreasingsymmetry inactiveMT(aMT)withagesuggeststhatthematurationofthenon- dominantcortexiscompletebyearlyadulthood(Garveyetal.,2003).
Ourhypothesiswasthattherelativeproportionoftheleft,dominant hemispheremotormapwouldincreasewithage,butmapasymmetry wasnotobservedinthetotalstudypopulationorinanyspecificage group.Aroboticmappingstudyinparticipantsagedeightto18years, butnotspecificallyexaminingtheeffectofage,showedatrendtoward reducedmotormapareaandvolumeforthedominantlefthemisphere comparedtotheright(Grabetal.,2018).Asmallerareainthedom- inanthemispherehasbeeninterpretedasreflectingthemoreefficient organizationofrefinedmusclerepresentationsascomparedtothenon- dominantside.OurpreviousstudyinpreadolescentswithAspergersyn- dromerevealedasymmetryreflectedasalargerrepresentationintheleft hemispherethantherightone,ascomparedtocontrolsubjectswhose motormapsweresymmetric(Säisänenetal.,2019).Thepreadolescents
withAspergersyndromealsohadslightlylargeroverlapinthelefthemi- spherecomparedtotherighthemisphere(Säisänenetal.,2019).This topicisevidentlysomewhatcontroversial.Theoretically,mapasymme- trycouldbeexpectedsincethereisanatomicalasymmetryintheleft dorsalpremotorareawhichhasahardlydistinguishableisolateddor- salsubregion(Genonetal.,2018).Inadults,theincreasedoverlapand increasedareainthelefthemisphereascomparedtotherightwascon- sideredtoreflectdominance,whereasinpianists,thelessandthemore symmetricoverlapswerepostulatedreflectingthelong-termplasticity formotorlearning(Chieffoetal.,2016).Ahigheroverlapintheleft hemisphereover therightwasconsideredtoreflect thehighertrain- ingoverthelifetime,includinggrasping,lifting,andjointstabilization (Melgarietal.,2008).
Significantcorrelations between themapmetrics andmotorper- formancewerefoundonlyintheright,non-dominanthemisphere,al- thoughthelefthemisphereshowedasimilartrend.Acorticalplastic- ity studyin which 1Hz rTMSwas applied to therighthemisphere detectedsignificantdifferencesonlyintheright,non-dominanthemi- sphere(Bashiretal.,2014).Theauthorssuggestedthatthisfindingwas inagreementwiththerighthemi-agingmodeli.e.,therighthemisphere ismoresensitivetoagingeffectsthanitsleftcounterpart.Ourprevious resultsonthesamedatadetectedstrongerinhibitionmeasuredwitha silentperioddurationintherighthemisphereinchildren,butnotin otheragegroups(Säisänenetal.,2018).Structurally,thecortexwas locateddeeperinthelefthemispherecomparedthanintherightinall agegroupsotherthaninchildren.Thisagreeswithonepublishedstudy inwhichtheleftprecentralgyruswasfoundtobelocateddeeperthan therightone(Davis2020),andanearlierstudythatfoundtheleftcen- tral sulcustobe deeper,andtheintrasulcalsurfaceoftheprecentral gyrusincreasedcomparedtotheright-sidedstructures(Amuntsetal., 1996).Subsequently,thisfindingwasalsoassociatedwithhandpref- erencebutwaslimitedtomales(Amuntsetal.,2000).Itneedstobe mentioned thatoppositefindingshave alsobeenreported(Togaand Thompson2003).
4.10. Clinicalimplications
The motor mapping was safe and well-tolerated in all subjects, though occasional pain or discomfort related to TMS has been re- ported(in children,adolescentsandadults) in accordance withpre- viousstudiesinyounger subjects(Garveyetal.,2001;Gilbertetal., 2004;Coarkinetal.,2011;Narayanaetal.,2015a;Grabetal.,2018).
Theriskforadverseeffectsinchildrenandtoddlersissimilartothat inadults(Krishnanetal.,2015;Narayanaetal.,2015b).Considering applicability,theperformancelevelwassuggestedtobemoreimpor- tantinpredictingasuccessfulmappingoutcomethanthechronological age(Narayanaetal.,2015a).Inpediatricpopulations,highstimulation intensitiesarerequired(onaverage93%ofthemaximalstimulatorout- put(Coburgeretal.,2012)insteadofrelatingstimulationintensityto rMT,buteven100%maynot alwayssufficienttoelicitMEPswitha focalcoildespitethesupportofnavigation(Ciechanskietal.,2017).A previousstudyprovedthefeasibilityofroboticmappinganditssuit- abilityforinvestigatingdevelopmentalplasticity,thoughintwochil- dren(eight yearsof age),therMT wasexcessivelyhigh(Grabetal., 2018).Muscle activationcanbe used tolower theMT,butweonly studiedrelaxedmuscles,whichiscriticalifonewishestoacquireaccu- rateandreliableresultsinpresurgicalmappings(Lucenteetal.,2018).
Thestimulusintensityof110%ofrMTproducedquitelargemaps.How- ever,weconsideritappropriatesinceitisclearlysuprathreshold,but doesnotproducedtoostrongE-fieldandthusresultinamaptoolarge (KallioniemiandJulkunen2016).Alowerstimulusintensity,suchas 105%ofrMT,couldresultinsmallermapsinpresurgicalmapping;in factthisisalreadyoftendoneinclinicalpractice(Jungetal.,2019).The outcomeofmotormappinginchildrencanbemorerobust— whether ornotaninvoluntaryMEPiselicited.TheAPBseemedtobeanoptimal muscleofchoiceformotormapping,asistheFDI,whichisoftenused
inTMSstudies(Groppaetal.,2012).Theexcitabilitythresholdcanbe thelowestinanyhandorforearmmuscle,asseenforexampleinthe left-handedboyaged6.8years,whohadtherMTsfortheAPBof91and 100%,butthosefortheADMwere62and78%.Thereisevidencein adultsthatatthegrouplevel,theMTsoftheAPBandFCRaresimilar (Wassermannetal.,1992),whichweverifiedinthegroupofchildren, thoughtheAPBhotspotmaynotbeoptimalfortheECR.Themotor mappingwasbasedontheMToftheAPB,andwedidnotdetermine itforeachmuscleseparatelywhichwouldberecommendediffeasibly possible.Anelegantstudyonplasticityinpianistsalsousedanapproach assessingtherMTinahotspotinwhichtheMEPswereelicitedineither theAPBorADM(Chieffoetal.,2016).
4.11. Strengthsandlimitations
Comparedtoearliernon-navigatedstudies,electricfieldonlinenav- igationallowedustogainaccuratehigh-resolutionanatomicalinforma- tioninrelationtothehandknob,resultinginmorestableMEPswith significantlyhigheramplitudesandshorterlatencies(Julkunenetal., 2009).Strengthinourstudywas thatbothhemisphereswereexam- ined,inarandomizedorder,andseveraldistalandproximalmuscles wererecordedsimultaneously.However,itneedstobementioned,that secondarymusclesmayhavedifferentrepresentationsandoverlapsas theusedmappingintensitycouldbedifferentfromtherMTofAPB.
Whengatheringdataforcomputingthemotormapsize,weapplied onlyasinglestimuluspergridpoint.Byusingadensegridandthesoft- warewithanarrowbrightnessindicatorthataidedinholdingthetilting optimal,wecouldreliablydetectevensmallchanges.Previously,ithas beenshownthattherepeatingofstimulusateachgridpointincreases theaccuracyofthemapmeasures(Cavalerietal.,2017).However,the gridweused(0.5cmx0.5cm)wasdensercomparedtothat(1cmx 1cm)usedinCavalerietal.(2017),andifweweretoplaceourstimuli withinthatgrid,itwouldcorrespondfourofourstimuliwithinonegrid element(1cm2).Inourstudy,thestimuluslocationswereconsidered inconjunctionwiththeresponseamplitude,and2-dimensionalspline interpolationwasused,whichreducesthevariabilityofindividualre- sponseswhen50μVstreamlinewastakentorepresenttheedgeofthe representationarea.Thepotentiallimitationsofusingonlyonestimulus pergridpointariseattheedgesofthemotormap,whereprobabilityof inductionofaresponseabovethethresholdamplitudearecloseto50%
andtherefore,thislimitationmayhaveaddedunbiasednoisetoourmap measures,buttheuseofdensegriddefusestheabove-mentionedphe- nomenon.Therestrictednumberofstimuliwasconsideredajustified compromisebykeepingthedurationofmeasurementsessiontolerable forthesubjects.Theinterstimulusintervalwaslongenoughtoprevent anyhabituationeffect.
Thenumberofparticipantswasnotlargeenoughtopermitreliable examinationof differences in motor map metricsbetween the sexes that areknowntoinfluence braindevelopment (Giedd et al., 2012; DennisandThompson2013;Akilanetal.,2020).Inasmallgroupsuch asours,thedifferenceneedstobegreaterinordertoresultinasmall p-value.However,despiteofthesmallgroupsizes,wefoundstatisti- callysignificantdifferencesinmotormapmetrics.Thestudywouldhave benefitedifwehadconductedseveralmotortasks.Mostofthemea- surementswereperformedatthesametimeoftheday,andtomaintain theattentionlevel,theparticipantswatchedaDVDduringtheTMS, buttheremaybevariabilityintermsofbothattentionandfatigue.The leisure-timeactivitiesrelatedtomotorskillssuchassportsorplayingan instrumentwerenotcontrolled.Cross-talkfromadjacentmuscleswhen usingsurfaceelectrodesmayhavecompromisedtheinterpretationof theresults(e.g.mapoverlap)(Masse-Alarieetal.,2017).Itshouldalso benotedthatthereisextensivevariabilityinbrainstructureamongin- dividuals,especiallyduringdevelopment(Caseyetal.,2005).
5. Conclusions
Weevaluated theeloquentupperlimbmotormapsin relationto theanatomyindifferentagegroupsaccuratelyusingelectricfieldnav- igatedTMSandmusclesatrest.Topographicmapswerefoundtobe ratheranalogousinallagegroupswithoutamajorcontributionfrom higherorder motorareasaroundM1.Themuscle-specificCoGswere scatteredinlargeareasincludingpostcentralareasinpreadolescence, whichisadynamicphaseinmotorfunctionimprovement.Associations between themapmetricsandhand dexteritywerefoundonlyin the righthemisphere.Clinically,theseresultsmayprovideareferencefor outliningfunctionalareasasapartofamultimodalpresurgicalevalua- tioninthepediatricpopulation.Weencouragerecordingofseveralup- perlimbmusclesduringmotormapping,astheexcitabilitycanbelower insomeotherhandorforearmmuscle,especiallyinchildrenwhohavea proportionallylargerarmrepresentationareaascomparedtothehand.
nTMSwasfoundtobewell-suitedforstudyingadevelopmentalcourse intheorganizationofthemotorcortex,andmotormappingmaybeuse- fulinfuturestudiesasabiomarkeroftreatment-relatedimprovement indevelopmentaldisabilitiessuchasperinatalstroke.Thedevelopment ofthemotormapbeforeschoolagemeritsfurtherinvestigation,andin thefuture,itwouldbeextremelyinterestingtostudythelongitudinal changeinthemotormapsaswellasotherbrainareas(thalamus,cere- bellum,spinalcordandcognitiveareas)inadditiontothecortex,allof whicharecandidatelocationsformotorskilllearning-inducedplasticity (Tennantetal.,2012).
DeclarationofCompetingInterest
LSandPJhavereceivedtravelbursariesunrelatedtothisstudyfrom NexstimPlc.PJhasasharedpatentwithNexstimPlc.Therestofthe authorsdeclarenoconflictofinterest.
Acknowledgments
VirpiLindi(deceased)PhDandAino-MaijaElorantaPhDfromthe UniversityofEasternFinlandareacknowledgedfortheirhelpwithre- cruitment.EwenMacDonaldisacknowledgedforlanguageediting.The authorswishtothankall subjectsfor participatingin thestudy. The fundingsourceshadnoinvolvementinthestudydesign;inthecollec- tion,analysis,andinterpretationofdata;inthewritingofthereport;
andinthedecisiontosubmitthearticleforpublication.
Funding
ThestudywasfundedbytheStateResearchFunding(grantnumber 5041730)andtheJuhoVainioFoundation.Laura Säisänenwassup- portedbytheArvoandLeaYlppö Foundation,andtheAcademyofFin- land(grantnumber322423).
Creditauthorstatement
LauraSäisänen;involvedinconceptualizationoftheresearch,anal- ysisofthedata;preparationofthemanuscript
MerviKönönen;developmentanddesignofmethodology,analysis ofthedata;preparationofthemanuscript;datavisualization
EiniNiskanen;analysisofthedata;preparationofthemanuscript TimoLakka;involvedinconceptualizationoftheresearch;acquisi- tionofthefinancialsupportfortheprojectleadingtothispublication.
NiinaLintu;involvedinconceptualizationoftheresearch;recruiting theparticipants
RitvaVanninen;involvedinconceptualizationoftheresearch;acqui- sitionofthefinancialsupportfortheprojectleadingtothispublication.
Petro Julkunen;developmentanddesignof methodology,in pro- gramming,implementationofcodeandsupportingalgorithms,prepa- rationofthemanuscript
