• No results found

Longitudinal changes in white matter microstructure after heavy cannabis use

N/A
N/A
Protected

Academic year: 2021

Share "Longitudinal changes in white matter microstructure after heavy cannabis use"

Copied!
13
0
0

Loading.... (view fulltext now)

Full text

(1)

ContentslistsavailableatScienceDirect

Developmental

Cognitive

Neuroscience

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

Longitudinal

changes

in

white

matter

microstructure

after

heavy

cannabis

use

Mary

P.

Becker

a,b,∗

,

Paul

F.

Collins

a,b

,

Kelvin

O.

Lim

c

,

R.L.

Muetzel

a

,

M.

Luciana

a,b

aDepartmentofPsychology,UniversityofMinnesota,75EastRiverRoad,Minneapolis,MN55455,UnitedStates

bCenterforNeurobehavioralDevelopment,UniversityofMinnesota,717DelawareStreetSE,Ste.333,Minneapolis,MN55414,UnitedStates cDepartmentofPsychiatry,UniversityofMinnesota,2450RiversideAvenueSouth,Minneapolis,MN55454,UnitedStates

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received21May2015

Receivedinrevisedform3October2015 Accepted6October2015

Availableonline9October2015

Keywords: DTI Cannabis Longitudinal Youngadult Whitematter

a

b

s

t

r

a

c

t

Diffusion tensorimaging (DTI) studiesof cannabis users report alterations in brain whitematter microstructure,primarilybasedoncross-sectionalresearch,andetiologyofthealterationsremains unclear.WereportfindingsfromlongitudinalvoxelwiseanalysesofDTIdatacollectedatbaselineand ata2-yearfollow-upon23youngadult(18–20yearsoldatbaseline)regularcannabisusersand23 age-,sex-,andIQ-matchednon-usingcontrolswithlimitedsubstanceusehistories.Onsetofcannabis usewaspriortoage17.Cannabisusersdisplayedreducedlongitudinalgrowthinfractionalanisotropy inthecentralandparietalregionsoftherightandleftsuperiorlongitudinalfasciculus,inwhitematter adjacenttotheleftsuperiorfrontalgyrus,intheleftcorticospinaltract,andintherightanterior thala-micradiationlateraltothegenuofthecorpuscallosum,alongwithlesslongitudinalreductionofradial diffusionintherightcentral/posteriorsuperiorlongitudinalfasciculus,corticospinaltract,andposterior cingulum.GreateramountsofcannabisusewerecorrelatedwithreducedlongitudinalgrowthinFAas wasrelativelyimpairedperformanceonameasureofverballearning.Thesefindingssuggestthat contin-uedheavycannabisuseduringadolescenceandyoungadulthoodaltersongoingdevelopmentofwhite mattermicrostructure,contributingtofunctionalimpairment.

©2015TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Cannabisis experiencing itsmoment in thespotlight inthe UnitedStates. Currently,23 states andtheDistrict of Columbia havelegalizedcannabisformedicaluse,and4stateshave legal-izeditforbothmedicalandrecreationaluse.Inthecontextofits changinglegalstatus,cannabishasbeenthemostcommonlyused “illicit”substanceintheUnitedStates(SubstanceAbuseandMental

HealthServicesAdministration,2014),with14%of11thand12th

gradestudentsreportinglifetimeuse(Falcketal.,2012),and6.5% of12thgradestudentsreportingdailyuse(Johnstonetal.,2013). Adolescentusemaybeproblematic.

Adolescenceis a criticalperiod ofneuralmaturation, during whichthereisreorganizationofcorticalconnections,anincrease in the fidelity of corticolimbic interactions, and neurochemical changesthatpromoteadaptivebehavioralregulation(Colbyetal.,

2011;Giedd,2004; O’Hareand Sowell,2008; Wahlstrometal.,

2010).Graymatterdevelopsalonganonlineartrajectorythatvaries byregion,withpeakproliferationpriortopuberty,followedbya

∗ Correspondingauthorat:DepartmentofPsychology,UniversityofMinnesota, 75EastRiverRoad,Minneapolis,MN55455,UnitedStates.

E-mailaddress:petr0308@umn.edu(M.P.Becker).

gradualdeclineandpruningduringadolescenceandyoung adult-hood, extendinginto the30s (Giedd,2004; Giedd etal., 1999;

Gogtayetal.,2004;Pfefferbaumetal.,1994;Sowelletal.,2001,

2003).Incontrast,whitemattervolumeincreaseslinearlyinaless regionally-variantmannerfromchildhoodtoadolescence(Giedd

etal., 1999;Paus etal.,2001), following aposterior–inferior to

anterior–superior developmental trajectory (Colby et al., 2011;

Sowelletal.,1999).Whitemattervolumedoesnotreachitspeak

untiladulthood,betweenthemid-30sand40s(Bartzokisetal.,

2001;Sowelletal.,2003;Westlyeetal.,2010).White matteris

criticalforefficiencyofsignalconductionbetweenbrainregions, anditsdevelopmentisassociatedwithimprovedcognitivecontrol andexecutivefunctioningduringadolescence(Petersetal.,2014;

Treitetal.,2013).

Diffusion tensor imaging (DTI) is a sensitive neuroimag-ing technique that characterizes parameters of white matter microstructure,yieldingmeasuresofitsdirectionalorganization (fractionalanisotropy:FA),averagedwaterdiffusioninall direc-tions(meandiffusivity: MD),and waterdiffusionperpendicular totheprimaryfiberorientation(radialdiffusivity:RD).Although severaleigenvaluesofthediffusiontensorcontributetothe cal-culationofFAandareaveragedtoyieldMD,thereductioninRD http://dx.doi.org/10.1016/j.dcn.2015.10.004

(2)

hasbeenspecificallyassociatedwithincreasedmyelination and axonalpackinginanimalmodels(Buddeetal.,2007;Songetal.,

2003,2005),andage-relatedchangesinFAinhumansareoften

attributedtodecreasedRD(Giorgioetal.,2008;LebelandBeaulieu,

2011;Lebeletal.,2008).Ingeneral,increasesinFAanddecreases

inMDaswellasRDareobservedinmajorwhitematterfibertracts throughoutadolescenceandextendingintoadulthood.The biolog-icalmechanismsthatdriveMDchangesarelesscleargiventhat MDisanimprecisemeasure.

Estimatesofpeakdevelopmentofwhitemattermicrostructure varybetweenages20and42(Kochunovetal.,2012;Lebeletal.,

2012;SchmithorstandYuan,2010;Westlyeetal.,2010).Unlike

studiesofwhitemattervolume,thereisconsiderableregional vari-abilityinthematurationaltimingofwhitemattermicrostructure, withearliermaturationincommissuralandprojectionwhite mat-tertracts,andlaterdevelopmentduringadolescenceandyoung adulthoodintheassociationtracts(Ashtarietal.,2007;Kochunov

etal.,2012;LebelandBeaulieu,2011;Lebeletal.,2008).

Asisthecasewithotheractiveperiodsofbraindevelopment

(Fride, 2008; Huizink,2014; Linnet etal., 2003), substanceuse

duringadolescenceandyoungadulthoodmayinterferewith nor-mativedevelopment,alteringthestructuralintegrityandfunction oftheadultbrain.Agrowingliteraturesuggestsneurocognitive deficitsand neuroanatomical alterations among adolescent and youngadultcannabis users (CUs) relative tonon-users. Behav-iorally,cannabisuseinadolescenceandyoungadulthoodismost commonlyassociatedwithrelativedeficitsinverballearningand memory,decision-making,abstractreasoning,andcomplexspatial workingmemory(Beckeretal.,2014;Bollaetal.,2002;Dougherty

etal.,2013;Fontesetal.,2011;Gruberetal.,2012;Hansonetal.,

2010;LisdahlandPrice,2012;Medinaetal.,2007)withdeficits

inverballearningandmemoryrepresentingthemostreplicated finding.

Neuroimagingstudiesfindalterations inbrain structure and functionassociatedwithuse.Reducedhippocampal(Ashtarietal.,

2011;Demirakcaetal.,2011;Lorenzettietal.,2015;Schachtetal.,

2012;Yüceletal.,2008),parahippocampalgyrus(Battistellaetal.,

2014;Matochiketal.,2005),orbitalfrontal(Battistellaetal.,2014;

Churchwelletal.,2010;Filbeyetal.,2014),andincreased

cere-bellum(Battistellaetal.,2014;Cousijnetal.,2012;Medinaetal., 2010)graymattervolumesareamongthemoreconsistent find-ingswithinCUs.Alterationsinamygdalagraymatterarereported, butthe directionofalterations varies betweenreports (Gilman

etal., 2014; Lorenzetti et al.,2015; Schachtet al., 2012;Yücel

etal.,2008).Thedegreetowhichgroupsarematchedonpotential confoundsofsex,age,andothersubstanceusevariesbetween stud-ies.Onenotablerecentstudyfoundnogroupdifferencesbetween CUsandcontrols,whowerematchedonalcoholuse,onmeasures ofgraymattermorphologyinthenucleusaccumbens,amygdala, hippocampus, and cerebellum, suggesting alcohol use or other confoundsmaysignificantlycontributetoreportedgroup differ-ences(Weilandetal.,2015).Further,findingsacrossstudiessuggest cannabisuselikelyexertsonlyamodesteffectongraymatter vol-ume,indicatingthatlargesamplesizesmaybeneededtodetecta trueeffect(Weilandetal.,2015).

Amongmagneticresonanceimaging(MRI)andpositron emis-siontomography(PET)studiesoffunctionalrelationshipsamong brainregions,CUsdemonstratealteredpatternsofbrainactivity whencomparedtonon-usingcontrols,particularlywithinthe pre-frontalcortex(Batallaetal.,2013),orbitofrontalnetwork(Filbey etal.,2014),prefrontalandoccipitoparietalcortex(Hardingetal., 2012), and frontal-subcortical network (Filbey and Yezhuvath, 2013).Apotentialmechanismdrivingthesedifferencesmaybe dis-ruptionofwhitematterfibertractsthatsupportsignalingbetween neuronswithinandacrossneuralnodes.Thedirectionof alter-ationsinwhitemattervolumeinavarietyofcorticalandsubcortical

regionshasbeeninconsistentlyreportedinCUs,withonegroup findingincreaseddensityofwhitematterassociatedwiththeleft parahippocampalandfusiformgyriandlowerdensityinleft pari-etallobewhitematter(Matochiketal.,2005),andasecondgroup notingcerebellarwhitemattervolumereductioninCUs(Solowij

etal.,2011b).However,otherstudiesfindnoassociationbetween

cannabisuseandwhitematterextentorvolume(Blocketal.,2000;

Cousijnetal.,2012;Jageretal.,2007),makingitdifficulttodraw

generalconclusionsfromtheexistingstudies.Itappearsthat mea-suresofwhitemattervolumeorextentderivedfromT1weighted structuralMRI scansmaynot besufficientlysensitivetosubtle alterationsinwhitemattermicrostructurethatimpactinformation processing.

DTIhasthepotentialtoaddressthislimitationgiventhatwhite matterorganizationcanbeexaminedatthemicrostructurallevel. DTIstudiesindicatethatCUsdemonstratealteredwhitematter organization,althoughagainthereareinconsistenciesacross stud-iesinthedirectionofMRI-measuredchanges.Moststudiesindicate thatFAislowerinvariouswhitematterregionsinCUs.SomeDTI reportsincludefindingsofincreasedCURDandMD,also indica-tiveofreducedwhitematterorganizationinCUs.DTIfindingsfor CUshaveinvolvedwidelyscatteredbrainregions,includingthe superiorlongitudinalfasciculus(Ashtarietal.,2009;Bavaetal.,

2009;Thatcheretal.,2010;Yüceletal.,2010),arcuatefasciculus

(Ashtarietal.,2009),frontalwhitematteradjacenttotheanterior

cingulatecortex(Gruberetal.,2011)andhippocampus(Zalesky etal.,2012),internalcapsule(Gruberetal.,2014),and the

ros-trum(Arnoneetal.,2008),genu(Gruberetal.,2014),andsplenium

(Zaleskyet al.,2012)ofthecorpuscallosum.Cannabisusemay

disruptthedevelopmentaltrajectoryofwhitematterorganization sincelowerFAlevelshavebeenassociatedwithanearlierageof cannabisuseonset(Gruberetal.,2011,2014).

NotallDTIstudieshave foundevidencefor lowerFAand/or greaterRDandMDinCUs.GreaterFAandlowerRDhavebeen reportedintheforcepsminorofthecorpuscallosum,atractthat connectstheorbitofrontalcortices(Filbeyetal.,2014),although yearsofusewithintheCUgroupshowedacurvilinear(quadratic) associationwithdecliningFAlevelsandrisingRDlevels.Oneearly studyfoundincreasedFAinCUsinwhitematterassociatedwith medialfrontal,cingulate,precentral,andparietalcortex,aswellas decreasedMDincingulateandmedialfrontalwhitematter(Delisi

etal.,2006).

Though findings are compelling,this literature is comprised ofcross-sectionaldesigns,limitinginterpretationofgroup-based differences.Longitudinalresearchmayclarifysomeinterpretative complexities, by assessing neural changes over time in rela-tiontoongoingsubstanceuse.Recentlongitudinalstudieshave exploredwhitematterdevelopmentassociatedwithalcohol ini-tiation (Luciana et al., 2013), binge alcohol use withcomorbid cannabisuse(Jacobusetal.,2013a),andpolysubstanceuse(Bava etal.,2013).Nolongitudinalstudytodatehasexploredwhite mat-termicrostructurespecificallyrelated tosustainedcannabis use duringyoungadulthood.

Thisstudyexamineschangeinmeasuresofaxonalfiber orga-nizationacrosstimeasa functionofcannabisuse.Young adult, regularCUsandcontrol participantswereassessedattwo time points,withatwo-yeartimeintervalbetweenassessments.Itwas predictedthatongoingdevelopmentofaxonalfiberorganization, asmeasuredbyincreasingFAanddecreasingRD,wouldbe rela-tivelydiminishedamongCUsaftermultipleyearsofchronicuse. GiventhebroadlydistributedDTIfindingstodate,weexpected alterationsoffiberorganizationinfrontalwhitematteraswellas fibertractsconnectingfrontalandposteriorregions;hippocampal whitematter;thecorpuscallosum;andcortical–subcortical pro-jectionfibers suchasthecorticospinaltract.We predictedthat thesealterationsinfiberorganizationwouldbecorrelated with

(3)

amountofcannabisuseinthefollow-upintervalintheCUgroup. TofacilitateinterpretationoftheDTIfindingsintermsof behav-ioralfunctioning,weincludeacorrelationalanalysisofscoresona testofauditoryverballearningandmemory.CUshaveconsistently shownimpairmentsinthisdomain(e.g.,Beckeretal.,2014)andwe predictedthattestperformancewouldcorrelatepositivelywithFA changesintask-relevantbrainregions.

1. Methods 1.1. Sample

Thirty-sevenCUswere initiallyrecruited intothis longitudi-nalstudythroughuniversityadvertisements.CUswererecruited if they reported using cannabis at least 5 times per week for at least 1 year. Cannabis useonset was required to be before age17 sothatlengthofuseacrossstudyparticipantswouldbe relativelyuniformandtoindexadolescentuseonset.CUswere excludedif theyweredaily cigarette smokersor if alcohol use exceeded4 drinksfor females and 5drinks for malesonmore than2occasionsperweek.Participantswereinitiallytested on a comprehensivebehavioral andimaging battery (Beckeretal.,

2014;Muetzelet al.,2013).Ofthose whoinitially participated,

27CUsreturnedforfollow-upassessmentafteratwo-year inter-val. Of these, 2 CUs were excluded from the current analysis becausetheyreportedcannabisusepatternsmarkedlylowerthan themajorityoftheCUsample(≤5timesinpast12 months)at follow-up.Twenty-threeoftheremaining25subjects(16males, 7 females), aged 18–20 years (M=19.45, SD=0.66) generated artifact-free scanning data. Twenty-three controls(16 males, 7 females),aged15–23yearsold(M=19.19,SD=2.31),wereselected fromalargerlongitudinalstudyofadolescentbraindevelopment (describedinmoredetailinLucianaetal.,2013;Uroˇsevi ´cetal., 2012)tomatchtheCUs onageand sexand scanning parame-ters.

Generalinclusioncriteriaforallparticipantsincludedbeinga nativeEnglishspeaker,right-handed,andwith normal/corrected-to-normal vision and hearing. Exclusion criteria included any contraindicationstoMRIscanning, areportedhistory of neuro-logicalproblemsorsignificantheadinjury,intellectualdisability, or current pregnancy. Controlsubjects were excludedat study enrollmentforevidenceofAxisIDSM-IVpsychopathology.Adult participantsandparentsofminorsprovidedinformedconsent,and minors(thoseunderage18)assentedtoparticipateinthestudy. TheUniversityofMinnesota’sInstitutionalReviewBoardapproved the protocol. Inclusion criteria were verified during an initial visit bya demographic and healthinterview questionnaire,the EdinburghHandednessInventory(Oldfield,1971), theWechsler AbbreviatedScaleofIntelligence(WASI:Wechsler,1999),andby structureddiagnosticinterviews(theKiddieScheduleforAffective DisordersandSchizophrenia;Kaufmanetal.,1997)administeredto theparticipantaswellas(forthoseunderage18)aparent. Partici-pantswhometinclusioncriteriareturnedforasecondassessment, whichincludedMRIscansaswellasabatteryofneuropsychological tasks.Cannabisuserswereaskedtorefrainfromdruguseforatleast 24hbeforetestingsoasnottobeacutelyhighduringthe assess-ment.Longerperiodsofabstinencewerenotrequired,becausewe didnotwishtostudyindividualsinthemidstofdrugwithdrawal andbecauseagoalofthestudywastocapturefunctional capaci-tiesinthecontextofactiveuse(Beckeretal.,2014;Muetzeletal.,

2013).

Participants completed MRI scanning and the neuropsycho-logicalbattery atstudyenrollment(Time 1)and ata follow-up assessment(Time2),approximately2yearsafterstudyenrollment (M=2.23years,SD=0.52).

1.2. Verballearningandmemory

We selected the Rey Auditory Verbal Learning Test (RAVLT;

Lezaketal.,2004;Rey,1993)toassessverballearningand

mem-oryperformance amongthesample. ThistaskdistinguishedCU fromcontrolsinapriorcomprehensiveassessmentof neuropsy-chological performance atbaseline (Beckeret al., 2014).It is a robustmeasureofeffortfulperformance,andrepresentsacognitive domainthatiscommonlyreportedtobediminishedamongCUs

(Bollaetal.,2002;Gonzalezetal.,2012;Grantetal.,2003;Hanson

andLuciana,2010;Harveyetal.,2007;Solowijetal.,2011a).The

RAVLTmeasuresacquisition,storage,andretrievalofverbal infor-mation.Duringthelearningphase,participantsarefirstreadand askedtorecallalistof15words.Thisprocedureisrepeatedfour additionaltimestoyieldthenumberofitemscorrectlyrecalled. Thisfive-triallearningphaseassessestheparticipant’simmediate learningandtemporarystorageofverbalinformation.Following thelearningtrials,participantsarethenreadandaskedtorecall anewlistof15words(interference trial).Theinterferencetrial assessesimmediatelearningofnewinformation,andispresented onlyonce.Followingtheinterferencetrial,participantsarethen askedtofreelyrecallasmanywordsastheycanfromthefirstlist (immediaterecall)andagainfollowinga30-mindelay(delayed recall).Statisticalanalyseswereconductedusingthetotal num-berofwordsrecalledduringthelearningtrials,interferencetrial, immediaterecall,anddelayedrecalltrials.

1.3. Substanceuseassessment

Substanceusefrequencyandamountwereassessedwiththe K-SADSinterview,andsubstanceusefrequencywasfurtherassessed withtheself-reportedPersonalExperienceInventory(PEI;Henley

and Winters, 1989). Within the PEI, participants reported the

numberoftimestheyusedsubstances(alcohol,cannabis,and non-cannabisillicitsubstances)withinthelast12monthsona5-point scale(never,1–5times,6–20times,21–49times,50–99times,100+ times).TheadolescentversionofthePEIwasusedwithparticipants undertheageof18,andtheadultversionofthePEIwasusedwith participants18yearsandolder.Scoringwasharmonizedtocreate comparable metricsacrossthetwo versions.Anin-house inter-viewquestionnaire,basedonguidelinesprovidedbytheNational InstituteonAlcoholAbuseandAlcoholism,wasalsoimplemented toassessalcoholandcannabisusecharacteristicsinthesample. Participantsreportedtheirtypicalusepatterns,asindexedby fre-quencyandquantityofuse,aswellasheavierusepatterns,defined asthenumberofoccasionsCUsusedtheirself-definedmaximum numberofcannabishits.

The Achenbach’s Adult Self-Report (ASR; Achenbach and

Rescorla,2003)questionnaireassessedself-reporteddailytobacco

use during the previous 6 months for subjects who endorsed tobaccouse.

1.4. MRIdataacquisitionandprocessing

MRI data were collected on a Siemens 3T Tim Trio scan-ner (Siemens Medical Systems, Erlangen, Germany) using a 12-channelarrayheadcoil attheUniversity ofMinnesota Cen-ter for Magnetic Resonance Research. Diffusion weighted data were acquired in the axial plane using a dual spin echo, single-shot,pulsed-gradient,echo-planarimaging(EPI)sequence (TR=8500ms,TE=90ms,64slices,nogap,FOV=256mm,voxel size=2.0mm×2.0mm×2.0mm, b value=1000s/mm2, GRAPPA iPAT=2).Theacquisitionboxwaspositionedtocoverthe cere-brumandasmuchofthecerebellumaspossible.Thirty-sixvolumes wereacquiredinthediffusionscan:6non-diffusion-weighted vol-umes(b=0)and30diffusion-weightedvolumes(b=1000s/mm2),

(4)

using gradient vectors distributed uniformly in 3-dimensional spaceaccordingtoanelectrostatic repulsionalgorithm (i.e.,the “Jones30”;Joneset al.,1999).For off-lineEPIgeometric distor-tioncorrection,b0fieldmapswereconstructedfromgradient-echo images acquired using different echo times (TE=4.62ms and 7.08ms; TR=700ms, flip angle=90◦, 64 slices, no gap, voxel size=2.0mm×2.0mm×2.0mm,FOV=256mm).

DiffusionMRIdatawereprocessedusingtheFDTandTBSS pack-agesinFSL(FMRIBSoftwareLibraryv4.0.1,http://www.fmrib.ox.

ac.uk/fsl,Smithetal.,2004).Eachdiffusion-weightedvolumewas

correctedforheadmotionandeddycurrentdistortionsusingan affineregistrationtothefirstb0referencevolume.Thediffusion series was corrected for geometric distortion caused by mag-neticfieldinhomogeneityusingPRELUDE(PhaseRegionExpanding LabelerforUnwrappingDiscreteEstimates)andFUGUE(FMRIB’s UtilityforGeometricallyUnwarpingEPIs)inconjunctionwiththe b0fieldmaps.BraintissuewasextractedusingBET(Brain Extrac-tionTool).Thediffusiontensorwasfitateachvoxelbysubmitting thepreprocesseddatafromtheb0referencevolumeandthe30 diffusion-weightedvolumestoFSL’sDTIFIT.Diagonalizationyields theeigenvectors(V1,V2,V3)andcorrespondingeigenvalues(L1, L2,L3)ofthediffusiontensor,whichdescribethedirectionsand apparentmagnitudesofwaterdiffusionwithineachvoxel.Two scalarvariableswerecomputedfromthesetensor components: radialdiffusivity(RD),whichistheaverageofL2andL3andreflects themagnitudeofdiffusionperpendiculartowhitemattertracts; andfractionalanisotropy(FA),avariancemeasurethatreflectshow stronglywaterdiffusionisrestrictedtotheprincipleeigenvector andrangesfrom0(equivalentdiffusionalongV1,V2,andV3)to 1(nodiffusionalongV2andV3).Givenourfocusonthe devel-opmentoffiberorganization,wedidnotanalyzemeandiffusivity, whichislessinterpretablethanRDinregionswithcomplexfiber orientations(Vosetal.,2012),oraxialdiffusivity,whichismore informativeinthecontextofaxonalinjuryanddegeneration(e.g.,

Buddeetal.,2009).

Tract-BasedSpatialStatistics(TBSS;Smithetal.,2006)wasused toaligntheDTIscalarvolumessothatvoxelwisestatisticalanalysis couldbeperformedacrossparticipantsandbetweentimepoints. AllFAvolumeswerealignedtoacommonspacetemplateusing thenonlinearregistrationIRTK(Rueckertetal.,1999;www.doc.ic.

ac.uk/∼dr/software).ThetemplatewasanaverageFAvolume

con-structedfrom72participantsdrawnfromourfullresearchsample, alignedto2mmisotropicMNIvoxelspace(Lucianaetal.,2013;

Olsonetal.,2009).RDvolumeswerealignedtocommonspaceby

applyingthetransformationmatricespreviouslycomputedduring theFAvolumealignments.FAvolumesweremaskedata thresh-oldof FA≥0.15 toreduce partialvolumeeffects, and thesame voxelmaskswereappliedtoRDvolumes.Toassesschangesover timewhilecontrollingforbaselinedifferences,commonspaceDTI scalarvolumesatfollow-upwereregressedonbaselinevolumes, yielding follow-up vs. baseline regression residuals for FA and RD.Theseresidualizedvolumesweresmoothedat4mmFWHM priortostatisticalanalysis.Inlongitudinaldataanalysisusingthis regressorvariablemethod(Allison,1990),thevarianceintime2 performance,residualizedforpredictionbasedontime1variables, reflectschangeovertimethatwasnotpredictedbybaseline

lev-els(seeCohenandCohen,1983,pp.402–422;CronbachandFurby,

1970;Metalskyetal.,1987).Baselinedatawereseparately

exam-ined.

1.5. Statisticalapproach

Distributionsofvariableswereexamined,andvariablesthatdid notmeettheassumptionsforparametricanalysisweresquareroot transformed,includingcannabisusevariablesoftotalnumberof hitsforthepast12monthsandmaximumcannabishitsfrequency

inthelast12months.Demographicdataaswellasbehavioral char-acteristicsrelatedtosubstanceusewerecomparedbetweengroups usinganalysesofcovariance,Chi-squareanalysisofindependence, andtheMann–WhitneyUtestatbaselineandfollow-up(computed inSPSS,version22).

SPM8(www.fil.ion.ucl.ac.uk/spm/software/spm8)wasusedto

performvoxelwisemultipleregressionanalysisoftheFAandRD follow-upvs.baselineresiduals.Predictorsintheregression equa-tionsincludedgroup,sex,ageatbaseline,timeintervalbetween baselineandfollow-upassessments,andaveragealcoholuse(the meanofbaselineandfollow-upPEIratingsforuseinthepast12 months).Preliminaryanalyseswereconductedusingbothbaseline andfollow-upPEIratingsforalcoholuseinthepast12months,but nosupra-thresholdclusterswereproducedforthealcohol predic-tors.Accordingly,thefinalregressionmodelsweresimplifiedto includeasinglealcoholcovariaterepresentingaverageduselevels acrossbaselineandfollow-up.Toassessgroupdifferencesat base-line,multipleregressionswereperformedonbaselineFAandRD volumeswithgroup,sex,ageatbaseline,andalcoholuseat base-line(past12months)aspredictorsinthemodel.Theseanalyses yieldtherelationshipbetweenapredictorvariableandcriterion variablewhileregressingoutthevarianceaccountedforbyother predictorsinthemodel(i.e.,ageatbaseline,sex,intervalbetween assessments,averagealcoholuse).

Allregression analyseswere evaluatedaccording to cluster-levelstatisticalsignificanceafterfamily-wiseerrorcorrectionfor multiplecomparisonsbasedonrandomfieldtheory(Worsleyetal., 1992),usinganinputvoxelwiseheightthresholdofp<0.01.

MeanvalueswerecomputedwithinclustersfromtheFA-change regressionanalysesandthencorrelatedwithcannabisusevariables withintheCUgroup,whilecontrollingforvariablesofno inter-est.Specifically,follow-upcannabisusemeasureswereregressed onbaselinecannabisusemeasures,along withsex,ageat base-line,timeinterval betweenbaselineandfollow-upassessments, andaverage12-monthalcoholuse.Residualsfromthese regres-sionswereusedastheoutcomeusevariables.Measuresofcannabis useexaminedincluded:ageofregularcannabisuse,totalnumber ofhitsforthepast12months,andmaximumcannabishits fre-quencyinthepast12months.Forcomparison,correlationswere performedonfollow-upmeasuresofalcoholuse(i.e.,total num-berofdrinks,binge-drinkingoccasionsinthelast12months,and maximumdrinksfrequencyinthepast12months,controllingfor intervaltimeandsexasdescribedabove)andtobaccouse, con-trollingforsex,timeintervalbetweenbaselineandfollow-up,and average12-monthalcoholuse.

Additionally,partialcorrelationswerecomputedovertheentire sample(bothCUsandcontrols)betweenclustermeanvaluesand totalscoresontheRAVLT,controllingforsex,ageatbaseline,time intervalbetweenbaselineandfollow-upassessments,average 12-monthalcoholuse,andWASIIQscores.Thiscorrelationalanalysis wasconductedtoclarifythedirectionoffunctionalassociationsfor thelongitudinalFA-changeclusters,i.e.,todetermineforeach clus-terwhetherincreasingordecreasingFAoverthefollow-upperiod wasassociatedwithbettercognitiveperformance.Arepeated mea-suresanalysisofcovariance(ANCOVA)wasperformedtoassessthe temporalstabilityofRAVLTscoresoverthefollow-upintervalas wellasgroupdifferencesinperformanceatbothtimepoints.More comprehensiveassessmentoftheneuropsychologicalfunctioning ofthisCUsampleatbaselinecanbefoundinBeckeretal.(2014). 2. Results

2.1. Demographicandsubstanceusecharacteristics

Groups were matchedon sexdistribution, ethnic and racial background,yearsofeducation,andestimatedIQ(seeTable1).The studyparticipantswerelargelyCaucasianandhadaboveaverage

(5)

Table1

Demographicandsubstanceusecharacteristicsofcannabisusers(CUs)andhealthycontrols.Valuesrepresentmeansandstandarddeviationunits,unlessotherwisespecified. GroupcomparisonswereconductedusingChi-square,Mann–WhitneyU,andone-wayanalysisofvariance.

Variable Control CU F,U,2 p

n 23 23

Time1:Baselinecharacteristics

Age 19.19(2.31) 19.45(0.66) 0.27 0.61

#Male/#Female 16/7 16/7 2=0.00 1.00

#Caucasian/#OtherEthnicity 21/2 19/4 2=0.77 0.38

Yearsofeducation 13.09(2.07) 13.26(0.92) 0.14 0.71

EstimatedFullScaleIQ 115.65(9.46) 115.17(11.02) 0.03 0.88

VocabularyT-Score 62.09(5.43) 61.83(8.34) 0.02 0.90

MatrixreasoningT-Score 55.65(7.06) 55.35(5.58) 0.03 0.87 Alcoholuse,past12months(MedianPEIrating) 1 4 U=54.00 <0.00

0:Never(n) 9 1 – – 1:1–5times(n) 7 0 – – 2:6–20times(n) 3 3 – – 3:21–49times(n) 3 6 – – 4:50–99times(n) 1 8 – – 5:100+times(n) 0 5 – –

Cannabisuse,past12months 0 5 U=0.00 <0.00

0:Never(n) 21 0 – – 1:1–5times(n) 2 0 – – 2:6–20times(n) 0 0 – – 3:21–49times(n) 0 0 – – 4:50–99times(n) 0 0 – – 5:100+times(n) 0 23 – –

Ageofregularcannabisuseonset – 15.35(1.16) – –

Total#hits,past12months – 3032.55(2395.31) – –

Maximumcannabishitsfrequency,past12months – 11.21(13.77) Time2:Follow-upcharacteristics

Yearsbetweenassessments 2.12(0.66) 2.34(0.31) 2.18 0.15

Ageatfollow-up 21.31(2.43) 21.79(0.82) 0.82 0.37

Agerangeatfollow-up 17.2–26.0 20.6–23.3 – –

Alcoholuse,past12months(MedianPEIrating) 3 4 U=84.5 <0.00

0:Never(n) 2 1 – – 1:1–5times(n) 3 0 – – 2:6–20times(n) 6 1 – – 3:21–49times(n) 8 2 – – 4:50–99times(n) 3 11 – – 5:100+times(n) 1 8 – –

Cannabisuse,past12months 0 5 U=0.00 <0.00

0:Never(n) 14 0 – – 1:1–5times(n) 7 0 – – 2:6–20times(n) 2 0 – – 3:21–49times(n) 0 1 – – 4:50–99times(n) 0 6 – – 5:100+times(n) 0 16 – –

Total#hits,past12months – 2637.92(2203.77) – –

Maximumcannabishitsfrequency,past12months – 16.20(28.66) – –

IQestimates.Groupswerematchedintermsofmeanage,though alargeragerangecharacterizedthecontrolsampletobalancethe sexdistributionacrossbothgroups.Timetofollow-upassessment wasmatchedbetweengroups,withanoverallaverageof2.23years (SD=0.52).

AllCUsreportedconsistentcannabisusepriortoboththeTime 1andTime2assessments(seeTable1).Giventhehighprevalence ofcannabisuseinthegeneralpopulation,controlswhohad min-imallyexperimentedwithcannabisuse(thatis,usethatdidnot exceed5timesinthepastyear)wereincluded.Minimal experi-mentationwithcannabisusewasreportedby2healthycontrolsat theTime1assessment(bothreportedusingcannabis1–5timesin thepast12months);atTime2,n=7reportedusing1–5timesinthe past12months,n=2reportedhavingtransitionedintousing6–20 timesinthepast12 months.Additionally, CUsreportedgreater alcoholuse(Time1&Time2CUMdnuse=50–99timesinpast12 months)thanhealthycontrols(Time1Mdnuse=neverinthepast 12months;Time2Mdnuse=21–49timesinpast12months). Min-imalnon-cannabisdrugusewasreportedamongCUs(usefewer than20times),andcontrols(Time1:noreporteduse;Time2:fewer than5times,seeTable1).Asindicatedinourpriorpublications

(Beckeretal.,2014;Muetzeletal.,2013),CUslargelymetcriteriafor

DSM-IV cannabis dependence at baseline without evidence of comorbidpsychopathology.

2.2. FAdevelopment

Significant differences in two-year FA change between CUs andnon-usingcontrolswereidentifiedinfivevoxelclusters(see

Table2)withtrend-leveldifferencesin two additionalclusters.

Infiveofthesesevenclusters,theControlshadmorepositiveFA changethanCUsoverthetwo-yearfollow-upinterval(Fig.1),while theconverseheldintheremainingtwoclusters(Fig.2).Thelargest Controls>CU FA-change cluster(3864mm3)was locatedin the righthemisphereprimarilyalongthesuperiorlongitudinal fascicu-lus(SLF)adjacenttotheprecentralandpostcentralgyri.Itextended anteriorlyintothejunctionbetweentheSLFandthecorticospinal tract(CST).ThenextlargestFAcluster(1632mm3)haditspeak adjacenttotheparietaloperculuminthelefthemisphere.The clus-terfollowed theSLFadjacenttothesupramarginaland angular gyri,extendingintotheforcepsmajorofthecorpuscallosum(CC) neartheanteriorportionofthelateraloccipitalcortex;abovethe forcepsmajor,itoverlappedacaudodorsalportionoftheinferior longitudinalfasciculus(ILF).TheposteriorportionofthethirdFA

(6)

Table2

Analysisofgroupdifferencesin2-yearchangeinfractionalanisotropy(FA)andradialdiffusivity(RD).ResultsfromvoxelwisemultipleregressionanalysisofFAandRD2-year changevalues(i.e.,residualsfromregressionofeachparticipant’sfollow-upFAorRDvolumeonthecorrespondingbaselineFAorRDvolume).Multipleregressioncovariates weregroup,sex,ageatbaseline,timeintervalbetweenbaselineandfollow-upassessments,andaveragealcoholuse(meanofbaselineandfollow-upPEIratingsforusein past12months).Allregressionanalyseswereevaluatedaccordingtocluster-levelstatisticalsignificanceafterfamily-wiseerror(FWE)correctionformultiplecomparisons basedonrandomfieldtheory,usinganinputvoxelwiseheightthresholdofp<0.01;trend-levelresults(FWEp≤0.10)areincludedaswell.Abbreviations:ATR=anterior thalamicradiation;CC=corpuscallosum;CST=corticospinaltract;CWP=cluster-wiseFWE-correctedp-value;FOF=fronto-occipitalfasciculus;SLF=superiorlongitudinal fasciculus.

Measure Contrast PeakZ-stat Peak partialr

Clustersize (mm3)

CWP X Y Z Hemisphere Anatomical region

Fractionalanisotropy Controls>CU 5.80 0.76 3864 0.001 32 −32 40 Right SLF,extendingtojunction withCST

Fractionalanisotropy Controls>CU 4.12 0.59 1632 0.017 −38 −44 24 Left SLF,extendingtoCC forcepsmajor

Fractionalanisotropy Controls>CU 3.88 0.56 1464 0.030 −18 16 42 Left Whitematteradjacentto superiorfrontalgyrus Fractionalanisotropy Controls>CU 4.46 0.63 1232 0.065 −16 −28 50 Left CST,adjacenttopre-and

postcentralgyri Fractionalanisotropy Controls>CU 4.25 0.61 1208 0.071 26 16 12 Right ATR;superiorFOF;

adjacenttofrontal operculum Fractionalanisotropy Controls<CU 4.49 0.63 1808 0.010 −14 16 26 Left AnteriorCC

Fractionalanisotropy Controls<CU 3.75 0.55 1696 0.014 −10 −24 −6 Left Whitematteradjacentto posteriorthalamus Radialdiffusivity Controls<CU 4.54 0.64 6848 0.001 16 −36 38 Right CST;SLF;posterior

cingulum Radialdiffusivity Controls>CU 3.97 0.57 2368 0.006 −24 −18 4 Left CST

cluster(1464mm3)waslocatedinthewhitematterofthemiddle regionofthesuperiorfrontalgyrusinthelefthemisphere,while theanteriorportionextendedintothelateralanteriorcingulumand anteriorthalamicradiation(ATR)adjacenttotheanteriorsuperior frontalgyrus.MostofthefourthFAcluster(1232mm3;trendlevel atclusterwiseFWEp=0.065)waslocatedwithintheCSTadjacent totheprecentralandpostcentralgyriinthelefthemisphere,with thelateralaspectoftheclusterextendingintotheSLF.Thefifth

Controls>CUFA-change cluster(1208mm3;trend levelat clus-terwiseFWEp=0.071)wasadjacenttotheinferiorfrontalgyrus (specifically,thefrontaloperculum)intherighthemisphere,with themedialaspectoftheclusteroverlappingtheanteriorthalamic radiationandthelateralaspectextendingintothesuperior por-tionoftheFOF.Aswillbediscussed,thesetrendlevelfindingsare mentionedbecausetheyreplicatefindingsreportedintheextant literature.

Fig.1.Voxelwiseanalysisofgroupdifferencesin2-yearchangeinfractionalanisotropyandradialdiffusivity.MultipleregressionanalysisasdescribedforTable2.Color codingindicatescontrastsformultipleregressioneffects:red=fractionalanisotropy,Controls>cannabisusers;green=radialdiffusivity,Controls<cannabisusers.(For interpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)

(7)

Fig.2. Voxelwiseanalysisofgroupdifferencesin2-yearchangeinfractionalanisotropyandradialdiffusivity.MultipleregressionanalysisasdescribedforTable2.Colorcoding indicatescontrastsformultipleregressioneffects:red=fractionalanisotropy,Controls<cannabisusers;green=radialdiffusivity,Controls>cannabisusers;blue=baseline fractionalanisotropy,Controls<cannabisusers;yellow=baselineradialdiffusivity,Controls>cannabisusers.(Forinterpretationofthereferencestocolorinthisfigure legend,thereaderisreferredtothewebversionofthisarticle.)

Morepositivetwo-yearFAchangewasobservedforCUsthan Controls within two clusters in the left hemisphere.One clus-ter(1808mm3)waslocatedprimarilyinanteriorcorpuscallosum (posteriorgenu,rostralbody),withasmalloverlapintothecaudal anteriorcingulum.Theothercluster(1696mm3)waslocatedin posteriorthalamuswhitematter.Themedialaspectofthiscluster overlappedtheCSTandtheATR,whilethelateralaspectextended intotheacousticandopticradiations.

Baselinedifferencesbetweengroupsdidnotfullyaccountfor thesefindings.AnalysisofFAgroupdifferencesatbaselineyielded onesignificantvoxelcluster(2584mm3),inwhichFAlevelswere greaterforCUsthanControls(Table3).Thisclustercoveredthe rightsideofthegenuoftheCC,crossingovertheCCmidlineto extendinto theinferiorpartof the leftgenu. In terms of fiber tracts,theclusteroverlappedthecaudalforcepsminor,withamuch greaterextensionintotherighthemispherethantheleft.

2.3. RDdevelopment

InadditiontotheFAresults,significantdifferencesintwo-year DTIchangebetweenCUsandControlswereidentifiedinthe anal-ysisofRDdata.ItwasgenerallyexpectedthatRDwoulddecrease overtime.RDchangewasmorepositiveforCUsthanControlsina verylargevoxelcluster(6848mm3)thatextendedoverthe poste-riortwo-thirdsoftherighthemisphere,inwhitematteradjacentto precentral,postcentral,supramarginal,inferiorparietal,precuneus, and posterior cingulatecortex (Table 2). The cluster peak was locatedinthemedialaspectofthecluster,whichoverlappedthe posteriorcingulum.Themiddlesectionoftheclusteroverlapped theCST,whilethelateralaspectoverlappedtheSLF.Bothofthese

clustersegmentshadsubstantialoverlapwithasimilarlylocated Controls>CUFA-change cluster(notethat increasedFAoftenis accompaniedbydecreasedRD;seeFig.1).RDchangewasmore positiveforControlsthanCUsinalefthemispherevoxelcluster (2368mm3)locatedalongtheCST,withapeakin theposterior limboftheinternalcapsule.

AnalysisofRDgroupdifferencesatbaselineyieldedone trend-level voxel cluster (1640mm3; trend level at clusterwise FWE p=0.10;Table3),inwhichRDlevelsweregreaterforControlsthan CUs.TheControls>CUsbaselineRDclusterwaslocatedinthesame CCregionastheControls<CUsbaselineFAcluster,butwassmaller insize(seeFig.2).

2.4. Brain–behaviorassociations

Cannabisusecharacteristics.WithinCUs,extentofcannabisuse wasnegatively associated withcluster meanFA-change values, forthreeControls>CUFA-changeclusters(seeTable4)suchthat higherlevelsofreportedusewereassociatedwithlower magni-tudesofchangeovertime. Specifically,thetotal numberofhits duringthepast12monthswassignificantlynegativelyassociated withclustermeanFA-changeintheleftCST(r=−0.259,p<0.05) andleft SLF/CCforcepsmajorjunction(r=−0.525,p<0.01).The maximumcannabishitsfrequencyduringthepastyearwas sig-nificantlynegatively associatedwithclustermeanFA-changein therightSLF/CST junction(r=−0.514,p<0.05).Age ofcannabis useonsetwasnotsignificantlycorrelatedwithclustermean FA-changevalues.Analogouscorrelationswereexaminedinrelation tomeasuresofalcoholandtobaccouseatfollow-upbutwere non-significant.

(8)

Table3

Baselinegroupdifferencesinfractionalanisotropy(FA)andradialdiffusivity(RD).ResultsfromvoxelwisemultipleregressionanalysisofFAandRDatbaseline.Multiple regressioncovariatesincludedgroup,sex,ageatbaseline,andalcoholuse.Allregressionanalyseswereevaluatedaccordingtocluster-levelstatisticalsignificanceafter family-wiseerror(FWE)correctionformultiplecomparisonsbasedonrandomfieldtheory,usinganinputvoxelwiseheightthresholdofp<0.01;trend-levelresult(FWE

p≤0.10)isincludedaswell.Abbreviations:CC=corpuscallosum;CWP=cluster-wiseFWE-correctedp-value.

Measure Contrast PeakZ-stat Peak partialr

Clustersize (mm3)

CWP X Y Z Hemisphere Anatomicalregion

Fractionalanisotropy Controls<CU 4.18 0.60 2584 0.012 4 26 4 Right Genu,forcepsminorofCC Radialdiffusivity Controls>CU 3.86 0.56 1640 0.102 2 26 6 Right Genu,forcepsminorofCC

Table4

Partialcorrelationsofclustermeanfractionalanisotropy(FA)changevalueswithcannabisuse.Partialcorrelationsbetweenclustermeanvaluesandmeasuresofcannabisuse overthepast12months,controllingforinterval,sex,ageatbaseline,andaveragealcoholuse(n=23).Abbreviations:ATR=anteriorthalamicradiation;CC=corpuscallosum; CST=corticospinaltract;FOF=fronto-occipitalfasciculus;SFG=superiorfrontalgyrus;SLF=superiorlongitudinalfasciculus;CU=cannabisuser;L=lefthemisphere;R=right hemisphere.

FAcluster Total#hitsa Maximumcannabishitsfrequencya Ageofregularcannabisuse CU>Controls

LanteriorCC(−14,16,26) 0.209 −0.325 −0.313 Lthalamicwhitematter(−10,−24,−6) −0.004 0.335 −0.087

Controls>CU RSLF/CST(32,−32,40) 0.115 −0.514* 0.281 LCST(−16,−28,50) −0.259* 0.354 −0.016 RATR(26,16,12) −0.152 0.321 −0.010 LSLF/CCforcepsmajor(−38,−44,24) −0.525* 0.345 0.074 LSFGwhitematter(−18,16,42) −0.230 −0.061 0.065

aSquareroottransformed. * p0.05.

Table5

Partialcorrelationsofclustermeanfractionalanisotropy(FA)changevalueswith RAVLTscores.Partialcorrelationsbetweenclustermeanvaluesandtotalscoreson theReyAuditoryVerbalLearningTest(RAVLT),controllingforsex,ageat base-line,timeintervalbetweenbaselineandfollow-upassessments,average12-month alcoholuse,andIQ.Abbreviations:ATR=anteriorthalamicradiation;CC=corpus callosum;CST=corticospinaltract;FOF=fronto-occipitalfasciculus;SFG=superior frontalgyrus;SLF=superior longitudinalfasciculus;CU=cannabis user;L=left hemisphere;R=righthemisphere.

FAcluster RAVLT:Baseline RAVLT:Follow-up CU>Controls

LanteriorCC(−14,16,26) −0.226 −0.504**

Lthalamicwhitematter (−10,−24,−6) −0.407** −0.342* Controls>CU RSLF/CST(32,−32,40) 0.350* 0.364* LCST(−16,−28,50) 0.360* 0.334* RATR(26,16,12) 0.288 0.210 LSLF/CCforcepsmajor (−38,−44,24) 0.319* 0.317* LSFGwhitematter (−18,16,42) 0.315* 0.357* * p0.05. ** p0.01.

Verballearningandmemory.Partialcorrelationsofclustermean FA-changevaluesandtotalRAVLTscores,computedovertheentire sampleofcontrolsandCUs,arelistedinTable5.Forboth base-lineandfollow-upRAVLTtotalscores,whenexaminedseparately, negativepartialcorrelations were obtainedforclusters derived fromtheCUs>Controlscontrastinthevoxelwiseanalysis,while positivepartialcorrelations wereobtainedforclustersfromthe Controls>CUs contrast.No significantpartial correlations were observedbetweenclustermeanFA-changevaluesandRAVLT dif-ferencescores (follow-up scoreminusbaseline score),as RAVLT scoresshowedlittlechangeduringthelongitudinaltestinginterval (F=0.972inarepeatedmeasuresANCOVAthatcontrolledforthe samenuisancevariablesasinthepartialcorrelations).Itshouldbe notedthatCUsperformedsignificantlymorepoorlythanControls atbothtimepoints(F=8.18,p=0.007,partialetasquared=0.173 forthemaineffectofgroupintherepeatedmeasuresANCOVA).

3. Discussion

Thisstudyassessedchangesinaxonalfiberorganizationovera 2-yearperiodofcontinuedheavycannabisuseinyoungadultswho initiateduseduringadolescence.Whitemattermicrostructureis knowntoshowcontinueddevelopmentalchangesduringthisage period(Lebeletal.,2012).Aspredicted,FAchangewasgenerally lowerandRDchangewasgenerallyhigherforCUsascomparedto controls.Specifically,CUsshowedlessFAchangeinthecentraland parietalregionsoftherightandleftSLF(extendingintothe tempo-ral/parietal/occipitalwhitematterjunctionregionontheleft),in whitematteradjacenttotheleftsuperiorfrontalgyrus,intheleft CSTjustmedialtotheSLF,andintherightanteriorthalamic radi-ationlateraltothegenuofthecorpuscallosum.Acorresponding findingofmoreRDchangeforCUsoverlappedtheFAclusterinthe rightSLF,butencompassedamuchlargerwhitematterregionthat includedadjacentmedialstructuressuchastheCSTandposterior cingulum.

These deviations in the development of white matter microstructurewereobservedprimarilyinwidelydistributed cor-tical association fibers, which is consistent with findings from functional MRI and behavioral studies of cannabis users. Cog-nitive studies consistently find diminished CU performance on tasksthatrequiremoreeffortfulperformance,suchasexecutive functioning,verbalfreerecallofpreviouslylearnedinformation, decision-making,abstractreasoning, andcomplexspatial work-ing memory(Beckeret al., 2014; Bolla et al.,2002; Dougherty

et al., 2013; Fontes et al., 2011; Gruber et al., 2012; Hanson

et al., 2010; Lisdahl and Price, 2012; Medina et al., 2007).

In functional MRI studies, CUs tend to show a broader pat-ternofcorticalactivationandrecruitalternativebrainnetworks during task performance, as compared to non-using controls

(Block et al., 2002; Chang et al., 2006; Harding et al., 2012;

Jacobsenetal.,2004;Kanayamaetal.,2004;Padulaetal.,2007;

Schweinsburgetal.,2010;Tapertetal.,2007).Increasedactivation

andrecruitmentofalternativeinformation processingpathways byCUs mayreflect imperfectfunctional compensationfor sub-tledegradationsinstructuralconnectivity,suchasthosereported here.

(9)

SomeofthechangesinCUswhitemattermicrostructurewere dose-dependent, i.e., correlated with self-reported quantity of typicalandheaviercannabisusepatterns,againinaccordwith pre-diction.Thetotalnumberofcannabishitsinthepast12months, asassessedatfollow-up,demonstrated strongnegative correla-tionswithmeanFAchangeintheleftSLF(−38,−44,24)andCST (−16,−28,50)clusters,asdidthemaximumcannabishitsfrequency withtherightSLF(32,−32,40)cluster.Similarmeasuresofalcohol useandameasureoftobaccousedidnotcorrelatewithFAchange. Thus,althoughtheCUgroupusedmorealcoholandtobaccothan controls,onlytheircannabisusewasassociatedinacontinuous quantitativemannerwithFAchange,therebystrengtheningthe inferencethatcannabisuseitselfwasaprimaryfactorunderlying theobservedgroupdifferences.Thisfindingcontradictsexisting longitudinalDTIstudiesthathavenotdocumentedchangesover time attributabletoamountofcannabis use. Bavaet al.(2013)

reporteda largenumber of DTIvoxel clustersdifferentiating a groupofalcohol-plus-cannabisusersfromnon-usersbutfoundthat resultsweredrivenbyalcoholuseoverthe18-monthfollow-up periodratherthancannabisuse.Similarly,Jacobusetal.(2013a)

foundthatovera3-yearinterval,agroupofbingealcoholusers andagroupofcombinedbingealcoholandcannabisusersdiffered fromnon-usingcontrolsin15voxelclusters,but withonlyone exception,theclustersdidnotdifferentiatethetwousergroups. However,a smallerlongitudinalDTIstudy(n=16)onthe initia-tion(ratherthancontinuation)ofheavyalcoholandcannabisuse reported20clustersdistributedthroughoutthebrainthat differ-entiatedparticipantswhoinitiatedcombinedalcoholandcannabis usevs.alcoholusealone,withdecreasedFAshownalmost exclu-sivelybythecombinedalcoholandcannabisusersoverthe3-year follow-upperiod(Jacobusetal.,2013b).Ourresultsareconsistent

withJacobusetal.(2013b)insuggestingthatbrainwhitematter

changesdrivenbyheavycannabisuseduringlateadolescenceand earlyadulthoodcanbedetectedbyDTIdespitethepresenceof concurrentalcoholuse.Thisisanimportantempirical demonstra-tion,giventhatwhitematterdevelopmentextendsintothethird decadeoflife(e.g.,Lebeletal.,2012)andthatcannabisuseduring adolescenceandearlyadulthoodhasbeenfoundtohavelong-term impactsoncognitivefunctioningandbrainstructureand organi-zation,bothinanimalmodels(e.g.,Bambicoetal.,2010;Gleason

etal.,2012;Pistisetal.,2004;Raveretal.,2013;Rubinoetal.,2009;

SchneiderandKoch,2003,2007;Schneideretal.,2008;Stopponi

etal.,2014)andinhumans(e.g.,Battistietal.,2010;Ehrenreich

etal.,1999;Fontesetal.,2011;Lisdahletal.,2013;Meieretal.,

2012;Popeetal.,2003;Solowijetal.,2012;Wagneretal.,2010).

Contrarytoourpredictions,severalfindingsfortheCUgroup involved DTIchanges generally associated with what mightbe interpretedasbetterwhitematterorganization,i.e.,increasedFA anddecreasedRD.Inthelefthemisphere,morepositiveFAchange was observed for CUs in the anterior CC (posterior genu, ros-tralbody)andposteriorthalamicwhitematter.Acorresponding decreased-RDclusteroverlappedtheposteriorthalamicareaand alsoextendedmediallyintotheCST.Inanalysesofbaselinedata, themedial CCgenuregion(caudalforcepsminortract)showed higherFAvalues,andtherewasatrendeffectoflowerRDvaluesin theCUgroup.Filbeyetal.(2014)alsoreportedhigherFAandlower RDforCUsintheforcepsminoroftheCC,andmoregenerallyan earlystudyreportedasimilarpatterninseveralregionsincluding themedialfrontalwhitematter(Delisietal.,2006).

Differencesobservedatbaselinemayreflectareasofrelative strengthintheCUgroup,consistentwithourpreviousbehavioral reportthatdescribedincreasedprocessingspeedinCUsinthe con-textofnumerousrelativecognitivedeficits(Beckeretal.,2014). InterpretationofFAincreasesandRDdecreasesintheCUgroup overthe2-yearfollow-upperiodismorecomplex,asthefunctional implicationsdependonmorenuancedaspectsofthedevelopment

ofaxonalfiberorganizationthatcandriveDTImetricsin oppos-ingdirections.For example,FAincreasesmaybeobservedover timewhenaxonalconnectionsalignmorecohesivelyinasingle fiberpopulation, i.e.,duringdevelopmentwithinaparallel bun-dled tract,while theconverse (FA decreases)may beobserved when neuraldevelopmentoccurswithin multiplefiber popula-tionsinthesameregion,i.e.,areaswherefiberscrossorfanout intograymatter(e.g.,Douaudetal.,2011).Thus,althoughboth groupcontrasts,Controls>CUsandCUs>Controls,generated sig-nificantclustersinthevoxelwiseanalysis,itdoesnotnecessarily followthattheCUs>Controlscontrastclustersrepresentregions ofenhancedwhitematterdevelopmentinCUs.Such determina-tionsultimatelydependoncorrelationswithbehavioralmeasures. AlthoughmappingtheDTIresultstolongitudinalchangesin per-formanceonourfullneuropsychologicalbattery,whichincludes nearlyahundredvariables,wasbeyondthescopeofthisreport, aninformativecorrelationalpatternemergedwhenweanalyzed thedirectionofassociationsbetweenthemeanFA-change clus-tervaluesandperformanceonademandingtestoflearningand memory,theRAVLT.TotalRAVLTscorescorrelatedpositivelywith FAchangewithinfiveofthesevenclusters,namelythose gener-atedbytheControls>CUscontrast.Conversely,totalRAVLTscores correlated negativelywithFAchangewithintheremainingtwo clustersgeneratedbytheCUs>Controlscontrast,i.e.,better per-formancewasassociatedwithmorenegative,ratherthanpositive, FAchangeduringthefollow-upperiodwithinthethalamic/CSTand anteriorCCregions.Thus,itispossiblethatallofourlongitudinal DTIresultsreflectdeleteriouseffectsofcontinuedheavycannabis use,asexpressedinwhitemattermicrostructureinbrainregions spanningdifferentformsoffibertractdevelopment,andin contin-uingfunctionaldecrementsinlearningandmemory.Interpretation islimited,though,bythefactthatCUsdemonstratedconsistently poorperformancerelativetocontrolsintheirverballearningand memoryperformanceanddidnotshowaperformancedeclineover time.

Notably,ourfindingsarealsoinformativeregardingthetiming ofuse-relatedimpactsonbrainstructure.Neurocognitivedeficits

(Fontesetal.,2011;Gruberetal.,2012)aswellasfunctional

alter-ationsobservedintask-basedfMRIparadigms(Tapertetal.,2007) incannabisusershavebeenattributedtoearlyonsetofuse,most typicallyoperationalizedasapatternofregularusebeginningprior totheageof16–17years.Becauseearlyonsetusersarelikelyto useinlargequantitiesandwithhigherfrequenciesthanlateronset users(Sagaretal.,2015),itisunclearwhetherageofonset, fre-quency,andmagnitudeofuseshowindependenteffectsonbrain structureandfunction.Indeed,earlyonsetuse,incombinationwith thehighfrequencyofuseanduseinlargeamountsthattendsto co-occur,mayrepresentadditivevulnerabilitiesassuggestedbySagar

etal.(2015).Heavychronicusehasbeenassociatedwithstructural

braindeviations(Battistellaetal.,2014;Filbey etal.,2014)and neurocognitivechanges(Bollaetal.,2002).Ourfindingssuggest thatimpactsofheavycannabisuseonwhitematter microstruc-turedonotnecessarilydependonageofonset.Furthermore,such impactsmaynotbereliablydetectableuntilafteratleast3–4years ofheavyusebeginningin theadolescentperiodgiven thatfew differencesatourbaselineassessment(after1–2yearsofheavy use)wereobserved.Thispatterningisconsistentwiththeideathat enduringchangesinbrainstructure,particularlynon-lesion-based alterationsthatcanbedetectedwithMRI,emergegraduallydueto theimpactsofexogenouscannabinoidexposureonneurochemical substratesthatultimatelyimpactaxonalintegrity.Largerstudies areneededtodisentangleindividualdifferencesintheeffectsof earlyexposureandlevelsofuseonlaterneuraland behavioral disruptions.

Unlike previouscross-sectional studies,we didnot find DTI differencesbetweenCUsandnon-usingcontrolsinwhitematter

(10)

tracts associated with core limbic structures such as the hip-pocampus(e.g.,fimbria;Zaleskyet al.,2012; Yücelet al.,2010) andamygdala(e.g.,uncinate fasciculus;Jacobuset al.,2013a,b;

Shollenbargeretal.,2015),eitheratbaselineorfollow-up.These

limbic structures are rich in endocannabinoid receptors, and cannabisexposurehasbeenlinkedwithhippocampaldamagein animalmodels(Rubinoetal.,2009).Otherstudiesmayhavehad greatersensitivitytodetecteffectsinthesesmallertortuouswhite mattertractsduetotheuseoffibertractography(Shollenbarger

etal.,2015;Zaleskyetal.,2012)andincorporationofsubstantially

lowercluster-sizethresholdsinvoxel-basedanalysis(Jacobusetal.,

2013a,b).

Limitationsofourstudymustbenoted.Thesamplewas rela-tivelysmallandcomprisedmostlyofCaucasianparticipants.While theDTIresultsweresimilartothosereportedinothersamples, additionalresearchisneededamongmoreraciallyandethnically diversesamplestoimprovegeneralizabilityofthefindings. Sim-ilarly,oursampleischaracterizedbyhighaverageIQestimates, which againlimits thegeneralizability of findings and perhaps skewsoverallFAlevelstowardhighervalues(Navas-Sánchezetal., 2014).Thatsaid,thesefindingsaresoberingregardingpotential impactsofchroniccannabisuseacrosslateadolescent develop-mentin anotherwiselow-risk sample.Anadditionallimitation withoursampleisthegreaterrangeofagesinthecontrolgroup relativetoCUs.Groupswerematchedonmeanage,andwetookthe additionalstepofcontrollingforageatbaselineinstatistical analy-sestomitigatepotentialconfoundingbyagevariance.Baselineage didnotproducesupra-thresholdclustersinanalysesofFAandRD changeoverthetwo-yearfollow-upperiod,whichindicatedthat thegroup-wiseDTIfindingswerenotconfoundedsignificantlyby theagerangedifferencebetweengroups.

Perhapsmoreimportantly,CUsandcontrolsdiffered substan-tiallyinalcohol use, whichis commonlyobservedin studiesof adolescent and young adultcannabis users (e.g.,Arnone et al.,

2008;Bavaetal.,2009,2013;Gruberetal.,2014;Jacobusetal.,

2013a,b;Shollenbargeretal.,2015;Thatcheretal.,2010).Toassess

thissamplingconfound,past12-monthalcoholuseatboth base-lineand follow-upwereincludedin preliminaryanalysesofFA andRDchange,whichproducednosupra-thresholdclustersfor thetwo alcohol usevariables. To providestatistical control for sub-thresholddifferences,weincludedtheaverageofthe base-lineandfollow-upalcoholusevariablesinthefinalDTIanalyses. Thedose-dependentfindingsalsoindicatedthatcannabisusewas theprimaryfactorinourDTIclusterresults,asmeanvaluesfrom severaloftheFAchangeclusterscorrelatedwithpast12-month usevariablesforcannabisbutnotalcohol.Nevertheless,the opti-maldesigntodetectandisolatelongitudinalcannabiseffectswould employcontrastinggroupsthatusedalcoholatequivalentlevels.

While we requested that participants be free of substance use for at least 24h prior to scanning, we did not assess for recencyofcannabis,alcohol,orotherdruguseanddidnotemploy drugtestingtoverifyreportedlevelsofcannabisandother sub-stance use among CUs or abstinence in both groups. CU and controlparticipantscompletedmultipleself-reportandinterview measuresofsubstanceuse,andresponsesacrossmeasureswere comparedforreliabilityofinformation.While itispossiblethat participantsendorsedfalseuselevels,thisappearsunlikelygiven consistencyof reports across measureswithin time points and across the follow-up interval. Further, CUs’ self-reported level ofcannabisusewassupportedbyevidenceofDSM-IV cannabis dependenceinthesampleaswellastheneurocognitive impair-mentsnotedatbaselinefortheseparticipants(seeBeckeretal.,

2014).

Finally, while reports have described alterations in cerebel-lar white matter in CUs, our field-of-view placement did not allowforequivalentexaminationofthisregionacrossparticipants.

Therefore, the current study could not address the impact of cannabisuseoncerebellarwhitematterdevelopment.

4. Conclusions

Ourfindingsareconsistentwithcross-sectional DTIresearch indicating that heavy cannabis use is associated with deleteri-ousalterationsinwhitemattermicrostructurewithinmajorfiber tracts,suchasthesuperiorlongitudinalfasciculus,corticospinal tract,andcorpuscallosum.Ourlongitudinalstudydesignextends theexistingliteraturebyassessingwhitemattermicrostructural changeinthecontextofcontinuingregularcannabisuseby high-functioning,non-treatment-seekingyoungadultswhowerefree ofcomorbiditiesotherthanalcoholuse.Thisdesignallowedusto distinguisha smallnumberofbaseline DTIdifferences between CUsandnon-usingcontrolsfromalargernumberofgroup differ-encesinFAandRDchangeovertime.Thesedifferencessuggest aberrantpatternsofneurodevelopmentasaconsequenceofheavy cannabisusegiventhatseveralDTIindicatorsofdisruptedwhite matterdevelopmentwerecorrelatedinadose-dependentpattern withthequantityandfrequencyofcannabisuse.Thesedistributed alterationsinwhitematterfiberdevelopmentmayunderliethe inefficient functional MRI activation patterns shown by heavy cannabisusers,aswellastheirdeficitsincomplexcognitionand effortfulprocessing.

Conflictofinterest

Theauthorsreportnoconflictsofinterest. Acknowledgements

Thiswork wassupported bythe NationalInstitute onDrug AbuseunderGrantR01DA017843awardedtoM.Luciana,bythe NationalInstituteonAlcoholAbuseandAlcoholismunderGrant R01AA020033awardedtoM.Luciana,andbyBTRCP41EB015894 andP30NS076408grantsawardedtotheUniversityofMinnesota’s CenterforMagneticResonanceResearch.Supportfromthe Univer-sityofMinnesota’sCenterforNeurobehavioralDevelopmentand theMinnesotaSupercomputingInstituteisalsogratefully acknowl-edged.

References

Achenbach,T.M.,Rescorla,L.A.,2003.ManualfortheASEBAAdultForms&Profiles. UniversityofVermont,ResearchCenterforChildren,Youth,&Families, Burlington,VT.

Allison,P.D.,1990.Changescoresasdependentvariablesinregressionanalysis. Sociol.Methodol.20,93–114.

Arnone,D.,Barrick,T.R.,Chengappa,S.,Mackay,C.E.,Clark,C.A.,Abou-Saleh,M.T., 2008.Corpuscallosumdamageinheavymarijuanause:preliminaryevidence fromdiffusiontensortractographyandtract-basedspatialstatistics. NeuroImage41(3),1067–1074,http://dx.doi.org/10.1016/j.neuroimage.2008. 02.064.

Ashtari,M.,Avants,B.,Cyckowski,L.,Cervellione,K.L.,Roofeh,D.,Cook,P.,Kumra, S.,2011.Medialtemporalstructuresandmemoryfunctionsinadolescents withheavycannabisuse.J.Psychiatr.Res.45(8),1055–1066,http://dx.doi.org/ 10.1016/j.jpsychires.2011.01.004.

Ashtari,M.,Cervellione,K.,Cottone,J.,Ardekani,B.A.,Sevy,S.,Kumra,S.,2009. Diffusionabnormalitiesinadolescentsandyoungadultswithahistoryof heavycannabisuse.J.Psychiatr.Res.43(3),189–204,http://dx.doi.org/10. 1016/j.jpsychires.2008.12.002.

Ashtari,M.,Cervellione,K.L.,Hasan,K.M.,Wu,J.,McIlree,C.,Kester,H.,...,Kumra, S.,2007.Whitematterdevelopmentduringlateadolescenceinhealthymales: across-sectionaldiffusiontensorimagingstudy.NeuroImage35(2),501–510,

http://dx.doi.org/10.1016/j.neuroimage.2006.10.047.

Bambico,F.R.,Nguyen,N.T.,Katz,N.,Gobbi,G.,2010.Chronicexposureto cannabinoidsduringadolescencebutnotduringadulthoodimpairsemotional behaviourandmonoaminergicneurotransmission.Neurobiol.Dis.37(3), 641–655.

(11)

Bartzokis,G.,Beckson,M.,Lu,P.H.,Nuechterlein,K.H.,Edwards,N.,Mintz,J.,2001.

Age-relatedchangesinfrontalandtemporallobevolumesinmen:amagnetic resonanceimagingstudy.Arch.Gen.Psychiatry58(2001),461L465.

Batalla,A.,Bhattacharyya,S.,Yücel,M.,Fusar-poli,P.,Crippa,J.A.,Nogué,S.,...,Yu, M.,2013.Structuralandfunctionalimagingstudiesinchroniccannabisusers: asystematicreviewofadolescentandadultfindings.PLOSONE8(2),e55821,

http://dx.doi.org/10.1371/journal.pone.0055821.

Battistella,G.,Fornari,E.,Annoni,J.-M.,Chtioui,H.,Dao,K.,Fabritius,M.,..., Giroud,C.,2014.Long-termeffectsofcannabisonbrainstructure. Neuropsychopharmacology,1–8,http://dx.doi.org/10.1038/npp.2014.67. Battisti,R.A.,Roodenrys,S.,Johnstone,S.J.,Pesa,N.,Hermens,D.F.,Solowij,N.,

2010.Chroniccannabisusersshowalteredneurophysiologicalfunctioningon Strooptaskconflictresolution.Psychopharmacology212(4),613–624,http:// dx.doi.org/10.1007/s00213-010-1988-3.

Bava,S.,Frank,L.R.,McQueeny,T.,Schweinsburg,B.C.,Schweinsburg,A.D.,Tapert, S.F.,2009.Alteredwhitemattermicrostructureinadolescentsubstanceusers. PsychiatryRes.:Neuroimaging173(3),228–237,http://dx.doi.org/10.1016/j. pscychresns.2009.04.005.

Bava,S.,Jacobus,J.,Thayer,R.E.,Tapert,S.F.,2013.Longitudinalchangesinwhite matterintegrityamongadolescentsubstanceusers.Alcoholism:Clin.Exp.Res. 37(S1),E181–E189,http://dx.doi.org/10.1111/j.1530-0277.2012.01920.x. Becker,M.P.,Collins,P.F.,Luciana,M.,2014.Neurocognitionincollege-ageddaily

marijuanausers.J.Clin.Exp.Neuropsychol.36(4),379–398,http://dx.doi.org/ 10.1080/13803395.2014.893996.

Block,R.I.,O’Leary,D.S.,Ehrhardt,J.C.,Augustinack,J.C.,Ghoneim,M.M.,Arndt,S., Hall,J.A.,2000.Effectsoffrequentmarijuanauseonbraintissuevolumeand composition.Neuroreport11(3),491–496.

Block,R.I.,O’Leary,D.S.,Hichwa,R.D.,Augustinack,J.C.,BolesPonto,L.L.,Ghoneim, M.M.,...,Andreasen,N.C.,2002.Effectsoffrequentmarijuanauseon memory-relatedregionalcerebralbloodflow.Pharmacol.Biochem.Behav.72, 237–250.

Bolla,K.I.,Brown,K.,Eldreth,D.,Tate,K.,Cadet,J.-L.,2002.Dose-related neurocognitiveeffectsofmarijuanause.Neurology59(9),1337–1343.

Budde,M.D.,Kim,J.H.,Liang,H.-F.,Schmidt,R.E.,Russell,J.H.,Cross,A.H.,Song, S.-K.,2007.Towardaccuratediagnosisofwhitematterpathologyusing diffusiontensorimaging.Magn.Reson.Med.57(4),688–695,http://dx.doi. org/10.1002/mrm.21200.

Budde,M.D.,Xie,M.,Cross,A.H.,Song,S.K.,2009.Axialdiffusivityistheprimary correlateofaxonalinjuryintheexperimentalautoimmuneencephalomyelitis spinalcord:aquantitativepixelwiseanalysis.J.Neurosci.29(9),2805–2813,

http://dx.doi.org/10.1523/JNEUROSCI.4605-08.2009.

Chang,L.,Yakupov,R.,Cloak,C.,Ernst,T.,2006.Marijuanauseisassociatedwitha reorganizedvisual-attentionnetworkandcerebellarhypoactivation.Brain 129,1096–1112,http://dx.doi.org/10.1093/brain/awl064.

Churchwell,J.C.,Lopez-Larson,M.P.,Yurgelun-Todd,D.A.,2010.Alteredfrontal corticalvolumeanddecisionmakinginadolescentcannabisusers.Front. Psychol.1,225,http://dx.doi.org/10.3389/fpsyg.2010.00225.

Cohen,J.,Cohen,P.,1983.AppliedMultipleRegression/CorrelationAnalysisforthe BehavioralSciences,2nded.Erlbaum,Hillsdale,NJ.

Colby,J.B.,VanHorn,J.D.,Sowell,E.R.,2011.Quantitativeinvivoevidencefor broadregionalgradientsinthetimingofwhitemattermaturationduring adolescence.NeuroImage54,25–31,http://dx.doi.org/10.1016/j.neuroimage. 2010.08.014.

Cousijn,J.,Wiers,R.W.,Ridderinkhof,K.R.,vandenBrink,W.,Veltman,D.J., Goudriaan,A.E.,2012.Greymatteralterationsassociatedwithcannabisuse: resultsofaVBMstudyinheavycannabisusersandhealthycontrols. NeuroImage59(4),3845–3851,http://dx.doi.org/10.1016/j.neuroimage.2011. 09.046.

Cronbach,L.J.,Furby,L.,1970.Howshouldwemeasurechange–orshouldwe? Psychol.Bull.74,68–80.

Delisi,L.E.,Bertisch,H.C.,Szulc,K.U.,Majcher,M.,Brown,K.,Bappal,A.,Ardekani, B.A.,2006.ApreliminaryDTIstudyshowingnobrainstructuralchange associatedwithadolescentcannabisuse.HarmReduct.J.3,17,http://dx.doi. org/10.1186/1477-7517-3-17.

Demirakca,T.,Sartorius,A.,Ende,G.,Meyer,N.,Welzel,H.,Skopp,G.,...,Hermann, D.,2011.Diminishedgraymatterinthehippocampusofcannabisusers: Possibleprotectiveeffectsofcannabidiol.DrugAlcoholDepend.114(2–3), 242–245,http://dx.doi.org/10.1016/j.drugalcdep.2010.09.020.

Douaud,G.,Jbabdi,S.,Behrens,T.E.J.,Menke,R.A.,Gass,A.,Monsch,A.U.,...,Smith, S.,2011.DTImeasuresincrossing-fibreareas:increaseddiffusionanisotropy revealsearlywhitematteralterationinMCIandmildAlzheimer’sdisease. NeuroImage55(3),880–890,http://dx.doi.org/10.1016/j.neuroimage.2010.12. 008.

Dougherty,D.M.,Mathias,C.W.,Dawes,M.A.,Furr,R.M.,Charles,N.E.,Liguori,A.,

...,Acheson,A.,2013.Impulsivity,attention,memory,anddecision-making amongadolescentmarijuanausers.Psychopharmacology226,307–319,

http://dx.doi.org/10.1007/s00213-012-2908-5.

Ehrenreich,H.,Rinn,T.,Kunert,H.J.,Moeller,M.R.,Poser,W.,Schilling,L.,..., Hoehe,M.R.,1999.Specificattentionaldysfunctioninadultsfollowingearly startofcannabisuse.Psychopharmacology142(3),295–301.

Falck,R.S.,Nahhas,R.W.,Li,L.,Carlson,R.G.,2012.Surveyingteensinschoolto assesstheprevalenceofproblematicdruguse.J.SchoolHealth82(5),217–224.

Filbey,F.M.,Aslan,S.,Calhoun,V.D.,Spence,J.S.,Damaraju,E.,Caprihan,A.,Segall, J.,2014.Long-termeffectsofmarijuanauseonthebrain.Proc.Natl.Acad.Sci. U.S.A.,1–6,http://dx.doi.org/10.1073/pnas.1415297111.

Filbey,F.M.,Yezhuvath,U.,2013.Functionalconnectivityininhibitorycontrol networksandseverityofcannabisusedisorder.Am.J.DrugAlcoholAbuse39 (6),382–391,http://dx.doi.org/10.3109/00952990.2013.841710.

Fontes,M.A.,Bolla,K.I.,Cunha,P.J.,Almeida,P.P.,Jungerman,F.,Laranjeira,R.R.,..., Lacerda,A.L.T.,2011.Cannabisusebeforeage15andsubsequentexecutive functioning.Br.J.Psychiatry198(6),442–447,http://dx.doi.org/10.1192/bjp. bp.110.077479.

Fride,E.,2008.Multiplerolesfortheendocannabinoidsystemduringtheearliest stagesoflife:pre-andpostnataldevelopment.J.Neuroendocrinol.20,75–81,

http://dx.doi.org/10.1111/j.1365-2826.2008.01670.x.

Giedd,J.N.,2004.Stucturalmagneticresonanceimagingoftheadolescentbrain. Ann.N.Y.Acad.Sci.1021,77–85.

Giedd,J.N.,Blumenthal,J.,Jeffries,N.O.,Castellanos,F.X.,Liu,H.,Zijdenbos,A.,..., Rapoport,J.,1999.Braindevelopmentduringchildhoodandadolescence:a longitudinalMRIstudy.Nat.Neurosci.2(10),861–863,http://dx.doi.org/10. 1038/13158.

Gilman,J.M.,Kuster,J.K.,Lee,S.,Lee,M.J.,Kim,B.W.,Makris,N.,...,Breiter,H.C., 2014.Cannabisuseisquantitativelyassociatedwithnucleusaccumbensand amygdalaabnormalitiesinyoungadultrecreationalusers.J.Neurosci.34(16), 5529–5538,http://dx.doi.org/10.1523/JNEUROSCI.4745-13.2014.

Giorgio,A.,Watkins,K.E.,Douaud,G.,James,A.C.,James,S.,DeStefano,N.,..., Johansen-Berg,H.,2008.Changesinwhitemattermicrostructureduring adolescence.NeuroImage39(1),52–61,http://dx.doi.org/10.1016/j. neuroimage.2007.07.043.

Gleason,K.A.,Birnbaum,S.G.,Shukla,A.,Ghose,S.,2012.Susceptibilityofthe adolescentbraintocannabinoids:long-termhippocampaleffectsand relevancetoschizophrenia.Transl.Psychiatry2(11),e199,http://dx.doi.org/ 10.1038/tp.2012.122.

Gogtay,N.,Giedd,J.N.,Lusk,L.,Hayashi,K.M.,Greenstein,D.,Vaituzis,A.C.,..., Thompson,P.M.,2004.Dynamicmappingofhumancorticaldevelopment duringchildhoodthroughearlyadulthood.Proc.Natl.Acad.Sci.U.S.A.101(21), 8174–8179,http://dx.doi.org/10.1073/pnas.0402680101.

Gonzalez,R.,Schuster,R.M.,Mermelstein,R.J.,Vassileva,J.,Martin,E.M.,Diviak, K.R.,2012.Performanceofyoungadultcannabisusersonneurocognitive measuresofimpulsivebehaviorandtheirrelationshiptosymptomsof cannabisusedisorders.J.Clin.Exp.Neuropsychol.34(9),962–976,http://dx. doi.org/10.1080/13803395.2012.703642.

Grant,I.,Gonzalez,R.,Carey,C.L.,Natarajan,L.,Wolfson,T.,2003.Non-acute (residual)neurocognitiveeffectsofcannabisuse:ameta-analyticstudy.J.Int. Neuropsychol.Soc.9(5),679–689,http://dx.doi.org/10.1017/

S1355617703950016.

Gruber,S.A.,Dahlgren,M.K.,Sagar,K.A.,Gönenc¸,A.,Lukas,S.E.,2014.Worththe wait:effectsofageofonsetofmarijuanauseonwhitematterandimpulsivity. Psychopharmacology231(8),1455–1465, http://dx.doi.org/10.1007/s00213-013-3326-z.

Gruber,S.A.,Sagar,K.A.,Dahlgren,M.K.,Racine,M.T.,Lukas,S.E.,2012.Ageofonset ofmarijuanauseandexecutivefunction.Psychol.Addict.Behav.26(3), 496–506,http://dx.doi.org/10.1037/a0026269.

Gruber,S.A.,Silveri,M.M.,Dahlgren,M.K.,Yurgelun-Todd,D.A.,2011.Whyso impulsive?Whitematteralterationsareassociatedwithimpulsivityinchronic marijuanasmokers.Exp.Clin.Psychopharmacol.19(3),231–242,http://dx.doi. org/10.1037/a0023034.

Hanson,K.L.,Luciana,M.,2010.NeurocognitiveimpairmentsinMDMAandother drugusers:MDMAalonemaynotbeacognitiveriskfactor.J.Clin.Exp. Neuropsychol.32(4),337–349,http://dx.doi.org/10.1080/

13803390903042361.

Hanson,K.L.,Winward,J.L.,Schweinsburg,A.D.,Medina,K.L.,Brown,S.A.,Tapert, S.F.,2010.Longitudinalstudyofcognitionamongadolescentmarijuanausers overthreeweeksofabstinence.Addict.Behav.35(11),970–976,http://dx.doi. org/10.1016/j.addbeh.2010.06.012.

Harding,I.H.,Solowij,N.,Harrison,B.J.,Takagi,M.J.,Lorenzetti,V.,Lubman,D.I.,..., Yücel,M.,2012.Functionalconnectivityinbrainnetworksunderlying cognitivecontrolinchroniccannabisusers.Neuropsychopharmacology37(8), 1923–1933,http://dx.doi.org/10.1038/npp.2012.39.

Harvey,M.A.,Sellman,J.D.,Porter,R.J.,Frampton,C.M.,2007.Therelationship betweennon-acuteadolescentcannabisuseandcognition.DrugAlcoholRev. 26(3),309–319,http://dx.doi.org/10.1080/09595230701247772.

Henley,G.A.,Winters,K.C.,1989.Developmentofpsychosocialscalesforthe assessmentofadolescentsinvolvedwithalcoholanddrugs.Int.J.Addict.24, 973–1001.

Huizink,A.C.,2014.Prenatalcannabisexposureandinfantoutcomes:overviewof studies.Prog.Neuro-Psychopharmacol.Biol.Psychiatry52,45–52,http://dx. doi.org/10.1016/j.pnpbp.2013.09.014.

Jacobsen,L.K.,Mencl,W.E.,Westerveld,M.,Pugh,K.R.,2004.Impactofcannabis useonbrainfunctioninadolescents.Ann.N.Y.Acad.Sci.1021,384–390,

http://dx.doi.org/10.1196/annals.1308.053.

Jacobus,J.,Squeglia,L.M.,Bava,S.,Tapert,S.F.,2013a.Whitematter

characterizationofadolescentbingedrinkingwithandwithoutco-occurring marijuanause:a3-yearinvestigation.PsychiatryRes.:Neuroimaging214(3), 374–381,http://dx.doi.org/10.1016/j.pscychresns.2013.07.014.

Jacobus,J.,Squeglia,L.M.,Infante,M.A.,Bava,S.,Tapert,S.F.,2013b.Whitematter integritypre-andpostmarijuanaandalcoholinitiationinadolescence.Brain Sci.3(1),396–414,http://dx.doi.org/10.3390/brainsci3010396.

Jager,G.,vanHell,H.H.,DeWin,M.M.L.,Kahn,R.S.,VanDenBrink,W.,VanRee, J.M.,Ramsey,N.F.,2007.Effectsoffrequentcannabisuseonhippocampal

References

Related documents

The study determined the level of awareness, perceived oral status and practices of oral health care among PLWHA attending Federal Medical Centre Abeokuta

Moving investments out of Canada is also part of the tax planning process as any assets left in Canada that are generating interest or dividends will be subject to Canadian

The instructions for Form 1040 Schedule A regarding Real Estate Taxes state “include taxes (state, local, or foreign) you paid on real estate you own that was not used for

Charge reversal reveals that loop 3 Lys-rich region is involved in electrostatic interactions needed by ribotoxins to cross cell membranes.. Loop 2 loop is responsible for

On line 18, enter the sum of the amount of the start-up credits allowed in the initial tax year, prior carryover tax years, and any refunds (these are the amounts claimed as

The non-linear analysis considering dead load &amp; prestress load (as constant load) and dead load, prestress load &amp; load due to accident temperature (as constant load) have

When we include only the linear term of this variable, its parameter estimate is highly significant (p-value below 0.01) with a negative sign – hence, emission intensities

Elsker, elsker det og tenker, På vår far og mor Og den saganatt som senker, Drømme på vå jord Og den saganatt som senker, Senker drømme på vår jord.. Canadian