Saguenay Youth Study : a multi-generational approach to studying virtual trajectories of the brain and cardio-metabolic health

(1)

ContentslistsavailableatScienceDirect

Developmental

Cognitive

Neuroscience

jo u r n al ho me p ag e :htt p : / / w w w . e l s e v i e r . c o m / l o c a t e / d c n

Saguenay

Youth

Study:

A

multi-generational

approach

to

studying

virtual

trajectories

of

the

brain

and

cardio-metabolic

health

T. Paus

a,∗,1

_,

_Z.

_Pausova

b,∗∗,1

_,

_M.

_Abrahamowicz

c

_,

_D.

_Gaudet

d

_,

_G.

_Leonard

e

_,

G.B.

Pike

f

_,

_L.

_Richer

g

a_Rotman_Research_Institute,_University_of_Toronto,_Toronto,_Canada b_Hospital_for_Sick_Children,_University_of_Toronto,_Toronto,_Canada c_McGill_University_Health_Centre,_McGill_University,_Montreal,_Canada

d_Community_Genomic_Medicine_Centre,_Department_of_Medicine,_Université_de_Montréal,_Chicoutimi,_Canada e_Montreal_Neurological_Institute,_McGill_University,_Montreal,_Canada

f_Hotchkiss_Brain_Institute,_University_of_Calgary,_Calgary,_Canada

g_Department_of_Health_Sciences,_University_of_Quebec_in_Chicoutimi,_Chicoutimi,_Canada

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received15June2014

Receivedinrevisedform3October2014 Accepted10October2014

Availableonline23October2014 Keywords: Adolescence Middleage MRI Brain Mentalhealth Addiction

a

b

s

t

r

a

c

t

ThispaperprovidesanoverviewoftheSaguenayYouthStudy(SYS)anditsparentalarm.The overarchinggoalofthiseffortistodeveloptrans-generationalmodelsofdevelopmental cascadescontributingtotheemergenceofcommonchronicdisorders,suchas depres-sion,addictions,dementiaandcardio-metabolicdiseases.Overthepast10years,wehave acquireddetailedbrainandcardio-metabolicphenotypes,andgenome-widegenotypes,in 1029adolescentsrecruitedinapopulationwithaknowngeneticfoundereffect.Atpresent, weareextendingthisdatasettoacquirecomparablephenotypesandgenotypesinthe bio-logicalparentsoftheseindividuals.Afterprovidingconceptualbackgroundforthiswork (transactionsacrosstime,systemsandorgans),wedescribebrieﬂythetoolsemployedin theadolescentarmofthiscohortandhighlightsomeoftheinitialaccomplishments.We thenoutlineindetailthephenotypingprotocolusedtoacquirecomparabledatainthe parents.

1. Introduction

Thelast centurywitnesseda dramaticgrowthin life

expectancy.IntheUnitesStatesofAmerica,lifeexpectancy

increasedfromanaverageof44.8/47.8(men/women)years

∗ Correspondingauthor.Tel.:+14162852500x2957. ∗∗ Correspondingauthor.Tel.:+14168137654x304340.

E-mailaddresses:tpaus@research.baycrest.org(T.Paus), zdenka.pausova@sickkids.ca(Z.Pausova).

1 _Equal_contributors_and_co-directors_of_the_Saguenay_Youth_Study.

in1900toanaverageof73.9/79.4yearsin1998(Smith

andBradshaw,2006),owingmainlytothedevelopmentof

treatmentsofinfectiousdiseasesandthemanagementof

cardiovasculardisordersandcancers(Guyeretal.,2000).

Unfortunately,thisincreaseinlifeexpectancyhasnotbeen

paralleledbyincreasesinhealthylifeexpectancy,deﬁned

as years lived without a disability. In 2002, the global

(194countries)gapbetweenlifeexpectancyand

health-adjusted life expectancy was 7.5 years (Mathers et al.,

2004).In developedcountries,themaincausesofYears

LivedwithDisability(YLD) –a metricused tocalculate

health-adjustedlifeexpectancy–arenon-communicable

http://dx.doi.org/10.1016/j.dcn.2014.10.003

(2)

diseases(86.2%ofallcauses),withpsychiatricconditions

contributingthemost(41.9%)totheoverallhealth

bur-den(Matherset al.,2004).Amongthelatterconditions,

unipolardepressive disorder(15%),alcohol abuse(6.8%)

andAlzheimer’sdiseaseandotherdementias(4.2%)stand

outasthemajorcausesof YLD.Althoughtheincreased

prevalenceofdementiasisinpartareﬂectionoflongerlife

span,acumulativeimpactofpoorcardio-metabolichealth

onbrainhealthisalsooneofthekeymechanisticpathways

leadingtodementia(seebelow).

Oneof themain reasonsfor the highhealth-burden

associated with psychiatric disorders, such as

depres-sionandsubstanceuse,butalsoschizophrenia(2.3%)and

bipolardisorder(2.2%), is theirearly onset and chronic

course, resulting in a large accumulation of YLD over

time(Fig.1).

For this reason, ourquest to understandthe causes

andpathwaysleadingtopsychiatricdisordersmusttake

adevelopmental perspective.Thisperspective

acknowl-edgesthe complexity of developmentalcascades – and

ensuingtransactions–playingoutovertime,across

lev-els and between organs (Masten and Cicchetti, 2010).

Inthefollowingtext,wewillreviewbrieﬂythesethree

elementsofdevelopmentalcascadesinordertoprovide

contextforthedesignoftheSaguenayYouthStudy(SYS).

We will conclude this section by providing motivation

for expanding the SYS to include a multi-generational

arm.

1.1. Developmentalcascadesandtransactions

1.1.1. Transactions over time is perhaps the

best-understoodlong-termdriverofhealth:acascadestarting

in pregnancy and early post-natal development,

pro-gressing through “preclinical” stages of a disease (e.g.,

in adolescence), and ending in a disease with its fully

expressed manifestations (e.g., in adulthood). Let us

illustrate this concept withtwo seminal studies of the

relationshipbetweenearlyenvironmentandadulthealth.

In experimental settings, Meaney and colleagues have

demonstrated that maternalcare(licking and grooming

of pups) has a striking impact on the

hypothalamus-pituitary-adrenal axis (HPA) and stress reactivity of

the offspring (Liu et al., 1997). This, in turn, sets in

motionanumberofpathophysiologicalprocessesleading

ultimately to poor mental and physical health (Fig. 2).

Using an epidemiological approach, Barker discovered

an associationbetweenbirthweightand cardiovascular

mortalityinadulthood(Barkeretal.,1989).Here,acascade

initiatedinuteroandinearlypost-natallifemayexerta

dominoeffectoncardio-metabolichealthoftheoffspring

for therestofhis/her life(Plagemann,2006).Theseare

two examplesof early“programming”of thebrain and

body systems, with powerful long-term consequences

formentalandcardio-metabolichealth.Notsurprisingly,

many scholars interested in long-term consequences

of early adversity have embraced the hypothesis of

Fig.1. Rangesofonsetage(top)andyearslivedwithdisability(bottom)forcommonpsychiatricdisorders.Topﬁgureadaptedfrom(Pausetal.,2008a). Bottomﬁgurebasedondatafrom(Whitefordetal.,2013).Percentagesindicateproportionofyearslivedwithdisabilityexplainedbyeachmentaland substanceusedisordergroupin2010(100%=allpsychiatricdisorders).

(3)

Fig.2.Associationsbetweenmaternalbehavior(lickingandgrooming),expressionoftheglucocorticoidreceptorinthehippocampus,regulationofthe hypothalamus-pituitary-adrenalaxisandpsychopathology(rightside).From(Zhangetal.,2013).LG,lickingandgrooming;ACTH,adrenocorticotropin; CRF,corticotropinreleasingfactor;5-HT,serotonin;camp,cyclicadenosinemonophosphate;PKA,proteinkinaseA;NGFI-A,nervegrowthfactor-inducible factorA;CBP,CREB-bindingprotein;GR,glucocorticoidreceptor.

developmentaloriginsofdisease(Wadhwa etal.,2009).

Given this perspective, longitudinal studies of health

trajectories are clearly the most suitable approach for

investigating developmentalcascadesovertime, though

thecomplexityof suchcascadesrequires largesamples.

Although a number of birth cohorts (_∼831 to 100,000

participants each) have been studying antecedents of

mental health in aprospective fashion,a recent review

suggests that many of these cohorts fall short due to

thelimited“breadthanddepthofmeasurement”

neces-saryfor enhancingourunderstandingof“howpre- and

perinatal factors and earlyneurodevelopment relate to

child psychopathology” (Thompson et al., 2010). From

theperspectiveofthedevelopmentalcognitive

neurosci-entist,theuseofmagnetic resonanceimaging(MRI) for

quantifyingbraindevelopmentprovidesboththebreadth

anddepthofquantitativephenotypesrelevantformental

health.Inthisregard,theGenerationRStudystandsout

astheonlybirthcohortthathasbegunserialMRscanning

of its members. Out of the total of 7893 Generation-R

children,1000ofthemhavebeenscannedbetween6and

8yearsofageand5000childrenwillbescannedbetween

10and12yearsofage(Whiteetal.,2013).

1.1.2.Transactionsacrosslevels(molecules,physiological

systems,individualbehavior,andsocialgroups)arecritical

forourunderstandingofpathwaysunderlyinggivenhealth

trajectories. The above example of maternal care

(lick-ingandgrooming)andHPAreactivityillustratescogently

thisperspective(Zhangetal.,2013).AsshowninFig.2,

thiscascadebeginswithanenvironmentalmanipulation–

maternalbehavior–measuredatthebehaviorallevel

(lick-ingandgroomingbehaviors),butitseffectontheoffspring

HPA system can beassessed at a systems level

(circu-latinglevelsofglucocorticoids),andthemolecularlevel

(expressionofglucocorticoidreceptorsinthe

hippocam-pus). And consequences on offspring behavior can be

assessed at the behavioral level (self-administration of

cocaine).

Thisexamplefromexperimentalworkcarriedoutin

rodents highlights one of the challenges of the human

work:theabsenceof direct observationsof behaviorin

largeepidemiologicalstudies–whetherinthecontextof

“exposures”(e.g.,maternalbehavior)or“outcomes”(i.e.,

offspringbehavior).Instead,werelyonself-reportsbythe

parentsandtheoffspring.Ontheotherhand,MR

imag-ingprovidesarichsourceofquantitativephenotypesthat

can be usedto characterize the state of structural and

functionalorganizationoftheoffspringbrainatasystem

level(phenomics).Inaddition,carefullydesignedcognitive

batteriescanprovidesystemlevelassessmentsofkey

pro-cessesunderlyingagivenbehavior(e.g.,decision-making

orrewardsensitivitytestedinthelaboratory).Finally,a

sampleofblood(orsaliva)canbethesourceof

biologi-calmaterialforboththesystemlevel(e.g.,stressandsex

hormones) andmolecular-level(e.g.,genetic[genomics]

andepigenetic[epigenomics]variations)assessments.We

havereviewedtheuseofthedifferent“omics”sciencesin

population-basedstudieselsewhere(Paus,2013).

1.1.3. Transactionsacross organshighlightthe

impor-tanceof anintegratedapproach tomental andphysical

health. As pointed out above, Alzheimer’sdisease (AD)

isanexcellentexampleoftheinterplaybetween

cardio-metabolicand brainhealth. Over50% ofthepopulation

attributable risk of AD is modiﬁable: diabetes, midlife

hypertension,midlifeobesity,smoking,depression,

cog-nitive inactivity or low educational attainment, and

physical inactivity (Barnes and Yaffe, 2011). As shown

(4)

Fig.3.Transactionsacrossorgans:fromeatingbehavior,throughfatcellstocardiovascularregulation.Basedon(Pausova,2006).From(Paus,2013).HFD, high-fatdiet;FFA,freefattyacids;AGT,angiotensinogen;TNF␣,tumornecrosisfactoralpha;IL-6,interleukin6;ROS,reactiveoxygenspecies;RAAS, rennin-angiotensin-aldosteronesystem.

from fat tissue to the endocrine system (e.g., insulin

signaling)andlow-gradeinﬂammation(e.g.,cytokines)to

sympatho-activation,kidneyfunction,bloodpressureand

cerebro-vascularreactivity, allleadingeventually tothe

deteriorationofbrainperfusionandmetabolism,lossof

grayandwhite-matterandcognitivedecline(Paus,2013).

Furthermore,earlyenvironment–andtherefore

trans-actions over time – may play an important role even

in the case of AD; not only in terms of metabolic

programming mentioned above but also vis-à-vis early

brain developmentand, in turn,formation of cognitive

andmental-health“reserves”(Stern,2012).Alife-course

approachis thereforewarrantednotonlyin thecaseof

neuro-developmentaldisorders, suchas depression and

addictions,butalsointhecaseofneuro-degenerative

dis-orders,suchasAD(MillerandO’Callaghan,2008).

1.2. Virtualhealthtrajectories

Given theneed for a life-span approach in studying

health trajectories, birth cohorts are a logical solution.

Without doubt, such cohorts are invaluable sources of

knowledgeaboutdevelopmentalcascades.Butaspointed

out above, very few birth cohorts have the necessary

depthof phenotyping(Thompsonet al.,2010).Manyof

thecutting-edgetoolsavailabletodaydidnotexistwhen

theoriginalbirthcohortswereinitiated.Atthesametime,

longevityofphenotypingtoolsisinevitablyshorterthan

thatofthehumanlifespan:whatisstate-of-the-arttoday

maybeobsolete(orunavailable)tomorrow.

Thetrans-generationalapproachrepresentsa possible

“shortcut”toachieving(andvalidating)long-term

mod-elsofhealthtrajectories,whileovercomingthechallenges

associatedwithimplementing andsustaininglong-term

longitudinal studies. In a multi-generational study, the

successivegenerationsrepresentstagesofdisease

trajecto-ries.Inotherwords,thetrans-generationalcommonalities

–basedonsharedgenes(25–50%ina3-generationfamily

vs.∼1%ingeneralpopulation)andthesharedfamily

envi-ronment(e.g.,geographical location,lifestyle)–become

a “signature” of a given family on which non-shared

elements (e.g., individual lifestyle, treatments) operate.

By comparingindividuals of thesame age,but coming

fromfamilies withdifferent“signatures,”we maybein

a positiontoidentifylong-term predictors ofbrain and

cardio-metabolichealth.

We have hypothesized that the accuracy of

dis-criminating between a descendant (e.g., daughter or

granddaughter)whowilldevelopadiseaseandonewho

willnot,testedagainsttheproﬁle(and/ordiseasestatus)

ofherancestor(e.g.,motherorgrandmother),willbe

com-parabletothediscriminativeaccuracyobservedinrecent

epidemiologicalstudiesbutonashortertime-scale(Paus,

2013).Thishypothesisisbasedontwogeneral

observa-tions: (1) most complex traits (and diseases) are likely

causedbyahostoffactors:multiplegenes,various

environ-mentalinﬂuencesand,ofcourse,combinationsofthetwo;

and(2)geneticandenvironmentalfactorsclusterin

fam-ilies,furtherenhancingsimilarityincomplextraitsacross

generations(Fig.4).

Inthecaseofmortalityrisk,asshownbyothersfrom

theRotterdamStudy(Hofmanetal.,2011),thepredictive

valueofanumberoflifestyleandphysiological

characteris-tics(162variablesintotal)ishighvis-à-vistheshort-term

(<1year)predictionofmortality(∼0.80)anditdecreases

to ∼0.70 when death occurs 15 years post-assessment

(Walteretal.,2012);thisisillustratedbythedashedlinesin

Fig.5.Ifourhypothesisiscorrect,thenPrognosisby

Ances-tor/Pedigreemaybeequalto,orbetterthan,along-term

(5)

Fig.4. Shapingthechild’sbrain.Thisschematicdiagramillustrateshowparents’brainscontributetothevariationsintheirchildren’senvironments– internal(e.g.,nutrition),physical(e.g.,pollution,noise,accesstoparks)andsocial(neighborhoodsafety,schoolfriendsandrolemodels).

solid lines). If so, Prognosis by Ancestor/Pedigree will

provideaglimpseintothefuturefordescendants,aswellas

providingthebasisforspeciﬁcinterventionsand

preven-tivemeasures,soastoavoidpredictedadverseoutcomes.

2. SaguenayYouthStudy:overalldesign

The ﬁrst wave of the SYS cohort (2003–2012) has

focusedonestablishingacommunity-basedsampleof

ado-lescents (12 to 18 years of age) in which to evaluate

associationsbetweentheexposuretoanadverse

prena-talenvironment,braindevelopmentandcardio-metabolic

health (Pausovaetal.,2007).Inkeeping withtheabove

principlesofdevelopmentalcascades,thecohortwassetup

sothatdetailedinformationcouldbecollectedatdifferent

levels(behavioral,systemic,molecular)andorgans(brain,

adiposetissue,cardiovascularsystem,endocrinesystem).

Overaperiodof10years,wehavecollectedarichdatasetin

1029adolescentsfromtheSaguenayLacSaintJeanregion

(Quebec,Canada).Thisregionisthehomeofthelargest

Fig.5. Risk/resilienceproﬁling:virtualvalidation (dashedlines) and prognosis by ancestor/pedigree (solid lines). The numbers indicate hypothesizeddiscriminativeaccuracy(0.5=chance,1=perfect discrim-ination).

From(Paus,2013).

populationwithaknowngeneticfounder-effectinNorth

America(DeBraekeleer,1991;DeBraekeleeretal.,1998;

Gradieetal.,1988;Grompeetal.,1994),makingit

partic-ularlysuitableforstudiesofcomplextraits.

Asamodelof prenataladversity,wechosematernal

cigarette smoking during pregnancy (MSP). This choice

reﬂects highprevalence of MSP in thegeneral

popula-tion; the latest resultsof theNational Survey onDrug

UseandHealth(U.S.A.)suggestthatsmokingduring

preg-nancyhasnotchangedsigniﬁcantlybetween2002/2003

and2011/2012,rangingbetween18%and15.9%

respec-tively(Anon.,2013).Smokingduringpregnancyhasbeen

associatedwithanumberofbehavioralsequelae(Cornelius

andDay,2009;Gaysinaetal.,2013;Kandeletal.,2009; Wakschlagetal.,2011),inthatoffspringofmotherswho

smokedcigarettesduringpregnancyaremorevulnerable

todevelopingaddictions,likelyduetothecombinationof

prenatalexposurewithotherfamilial(geneticand

envi-ronmental)risks.

Werecruitedadolescentsinhighschools.Overaperiod

of10years,ourteamhasmade28visitstoschoolsand,in

thisway,contactedatotalof27,190students(18,127

fami-lies).Ofthe18,127families,5570(33%)sentaresponsecard

indicatingtheirinterestinthestudy(3269families;59%

ofallresponses)ordecliningfurtherparticipation(2301

families;41%ofallresponses).Basedontheinclusion(e.g.,

maternalsmokingduringpregnancy, 2ormore siblings

perfamily)andexclusion(MRcontraindications)criteria

(SupplementaryTableIinPausovaetal.,2007),atotalof

1801families(55%oftheinterestedfamilies)wereeligible

toparticipateinthestudy;theeligibilitywasdetermined

byaresearchnurseviaastructuredtelephoneinterview.

Individualswhosemotherssmokedduringpregnancy(at

least1cigaretteperdayduringthesecondtrimester)were

identiﬁedﬁrst(viaatelephoneinterviewwiththemother);

thenweselectednon-exposedindividuals(nosmokingfor

(6)

Table1

SaguenayYouthStudy:Baselineinadolescence(completed).

Domain Tool Phenotypes

Brain MRI Globalandregionalvolumes;corticalsurface&thickness;MTR

Cognition 6-hbattery FSIQ,VIQ,PIQ;verbal,visuospatial,workingmemory;

executivefunctioning,problemsolving,ﬂuency,language, phonologicalandmotorskills;socialcognition

Mentalhealth DPS,GRIP Epidemiologicaldiagnoses;symptomcounts

Substanceuse GRIPado Cigarettesmoking,cannabis,alcoholuse,drug

experimentation(ageofinitiation,last30-days,binge drinking)

Personality NEO-PI-R Neuroticism,extroversion,openness,agreeableness,

conscientiousness

Sexualmaturation PDS Stagesofpubertaldevelopment(Tannerstages)

Lifestyle Lerner,24-h

foodrecall

Sleep,energyandnutrientintake,physicalactivity, extracurricularactivities,sexuality,academic/vocational aspirations

Family environment

FamEnvi Stressfullifeevents,ﬁnancialdifﬁculties,SES(familyincome, parentaleducation)

Bodycomposition Anthropometry,

MRI, Bioimpedance

Height,weight,circumferences,skinfolds;subcutaneous, visceralfatandmusclevolumes;fat&musclemass Cardiovascular Finometer Beat-by-beatbloodpressureandheartrateatrestandin

responsetophysicalandmentalchallenges,sympathetic& parasympathetictone

Hormones Blood Testosterone,estrogen,cortisol

Biochemistry Blood Glucose,insulin,cholesterol,HDL-cholesterol,triglycerides, leptin,C-reactiveprotein,glycerol,freefattyacids

Lipidomics LC-ESI-MS ∼700lipidspecies

MTR,magnetizationtransferratio;DPS,DISCPredictiveScales;GRIP,GroupedeRecherchesurl’InadaptationPsychosociale,adolescentself-assessmentof mentalhealthandsubstanceusedevelopedfortheSYSbyJ.SéguinbasedonvalidatedNationalLongitudinalSurveyofChildrenandYouth(NLSCY)and Que-becLongitudinalStudyofChildDevelopment(QLSCD)protocols;Lerner,adolescentself-assessmentdevelopedbyRichardLerner.FSIQFullScaleIQRating; VIQ,VerbalIQRating;PIQ,PerformanceIQRating;PDS,PubertyDevelopmentScale;HDL,high-densitylipoprotein;LC-ESI-MS,liquid-chromatography electrospray-ionizationmass-spectrometry;NEO-PI,Neuroticism,Extraversion,Openness–PersonalityInventory.Fordetails,seePausovaetal.,2007.

werematchedtotheexposedonesbymaternaleducation

andschoolattended.Notethatthis matchingprocedure

wasusedonanongoingbasisthroughoutthestudy;itwas

appliedateachhighschoolsothatanequivalentnumberof

“exposed”and(matched)non-exposedadolescentswere

recruitedat each school.Inthis manner,weminimized

differences between the “exposed” and “non-exposed”

adolescentsintheirfamilies’socio-economicstatus. We

usedafamily-baseddesign,recruitingaminimumoftwo

siblingsperfamily;notethatsiblingswereconcordantfor

theexposurestatusinthemajorityoffamilies(446/481

families;93%).Phenotypingoftheadolescentstookplace

overseveralsessions(∼15hintotal)andincludeda

num-ber of domains detailed in Table 1 (further details in

Pausovaetal.,2007andwww.saguenay-youth-study.org);

eachadolescentprovidedafasting(morning)blood

sam-ple.

Duringthisinitialphase(Wave1),biologicalparentsof

theadolescentsﬁlledoutaseriesofquestionnairesabout

thefamilyenvironmentandtheirmentalhealth;thelatter

includedquestionsaboutcigarettesmoking,alcoholuse,

anddrugexperimentationthroughouttheirlife(including

currenthabitsandageofonset),andthepresenceof

anti-socialbehavior(atpresentandduringtheiradolescence;

Table2).Parentsalsoprovidedabloodsampleforgenetic analyses.

InTable3,weprovidebasicdemographicinformation

aboutthesampleofadolescentswhounderwentthefull

assessmentduringWave1.InTable4,weprovide

infor-mationaboutthelifetimehistoryandcurrent(last30days)

Table2

SaguenayYouthStudy:Baselineinparents(completed).

Familyenvironment FamEnvi Stressfullifeevents,ﬁnancial difﬁculties,SES(familyincome, parentaleducation)

Mentalhealth GRIPadult Symptomcounts(depression, anxiety,anti-socialbehavior) Substanceuse GRIPadult Cigarettesmoking,alcoholuse,

drugexperimentation

FamEnvi,questionnaireonfamilyenvironmentdevelopedbytheSYS team;GRIPAdult,self-assessmentofmentalhealthandsubstanceuse, asadaptedbycolleaguesattheGroupedeRecherchesurl’Inadaptation PsychosocialeoftheUniversityofMontreal.

Table3

Wave1:Baselineinadolescents(completed).

Measure Distribution

N 1029

Numberoffamilies 481

Age(years) Mean=15.02;SD=1.84

Sex 48%male;52%female

ExposuretoMSP 48%exposed;52%non-exposed Householdincome ≤$20,000–13%

$30,000–40,000–19% $50,000–60,000–24% $70,000–80,000–20% ≥$85,000–24% FullscaleIQ Mean=104.42;SD=12.14 SD=Standarddeviation;IQ=intelligencequotient;MSP=maternal smok-ingduringpregnancy.

(7)

Table4

Percentofadolescentsreportingdrugexperimentation(lifetime)andcurrentuse(last30days)forcannabis,alcoholandcigarettesintheSYSadolescents. N Cannabis (lifetime) Cannabis(last 30days) Alcohol (lifetime) Alcohol(last 30days) Cigarettes (lifetime) Cigarettes(last 30days) Earlyadolescence (12–15.9years) 704 22% 9% 52% 21% 21% 8% Lateadolescence (16–18years) 324 55% 20% 91% 65% 38% 22%

useofthethreemostcommonsubstances:cannabis,

alco-holandcigarettes.Notethatthecurrentuseofcannabis

inolderadolescents(16–18years)iscomparabletothat

found in other countries among high-school students

(Hibelletal.,2012).

Using deoxyribonucleic acid (DNA) extracted from

theblood samplesoftheadolescentsand theirparents,

we have acquired information about single nucleotide

polymorphisms (SNPs) using a genome-wide approach.

The ﬁrst 600 adolescents were genotyped with the

Illumina Human610-Quad BeadChip (610K SNPs). The

remaining 424 adolescents and all 971 parents were

genotypedwiththeIllumina HumanOmniExpress

Bead-Chip(700KSNPs).Wehaveusedimputationstogenerate

the same set of markers for the two samples; we

employed an imputation protocol developed by the

ENIGMA WorkingGroup, and imputedgenotypes using

IMPUTE(www.mathgen.stats.ox.ac.uk/impute),witha

ref-erenceﬁlecreatedbytheENIGMA2GeneticsSupportTeam.

Thisreferenceﬁleisbasedonthemostrecentversionsof

the1000GenomesProjectset(Phase1,Releasev3;∼41M

SNPs)butincludesonly∼13MSNPsthatarepolymorphic

inCaucasiansandhavebeenobservedmorethanoncein

Europeanpopulations.

Inaddition,wehavequantiﬁedtherateofDNA

methyla-tionsacross450,000CpGsitesinasubsetoftheadolescents

(n=132)andtheirparents(n=280);thiswasaccomplished

byhybridizingDNAtotheInﬁniumHumanMethylation450

BeadChip(Illumina,SanDiego,CA).Thischipinterrogates

methylationat>485,000CpGsites,providingcoverageof

>99%RefSeqgenes;theCpGsitesaretargetedacrossgene

regions including the promoter, 5UTR, ﬁrst exon, gene

body,and3UTR,aswellasintergenicsequences(Sandoval

etal.,2011).

Finally,targetedlipidomicsproﬁlingiscurrently

con-ducted in all adolescents and their parents. Liquid

chromatography,electrosprayionizationmass

spectrom-etry (LC-ESI-MS) is used to assess plasma proﬁles of

>700 glycerolipid,glycerophospholipidand sphingolipid

speciesaspotentialbiomarkersofcardiovascularand

men-talhealth.

3. SaguenayYouthStudy:highlights

Beforeproceedingwiththedescriptionoftheparent

arm of theSYS cohort (Wave2: Parents), letus brieﬂy

highlight some of the published observations made in

Wave1:Adolescents.Wewillfocushereonﬁndings

rele-vanttomaternalsmokingduringpregnancyandaddictive

behavior;ourworkonsexdifferencesinthematuration

of white matter(Herveet al., 2009;Perrin et al.,2008,

2009), puberty-related changes in the face morphology

(Mareckovaetal.,2011,2013)andcardio-metabolichealth (Goodwinetal.,2013;Melkaetal.,2013a,2012;Pausova etal.,2010,2012;Symeetal.,2008,2009)canbefound elsewhere.

One of the most common consequences of MSP is

intra-uterinegrowthretardation(Lowe,1959);thisisnot

surprisinggiven the multipleeffects ofcigarette

smok-ingon thesupply of nutrientsand oxygen tothefetus

(reviewedinPausovaetal.,2007;Slotkin,1998).Asshown

withfetalimaging,braingrowthdoesnotappeartoescape

thisglobalphenomenon(Anblaganetal.,2013).Bythetime

theexposedoffspringreachesadolescence,however,the

brainsizeappearstobethesameasthatofnon-exposed

adolescents.Nonetheless, we askedwhether this is the

casealsoforindividualswithaparticulargeneticvariation

associatedwithbrainsize,asrevealedinagenome-wide

association study (GWAS) in the SYS adolescents. We

foundthatthis wasnot so:exposed femaleadolescents

withtheKCTD8risk-varianthadsmallersurface areaof

the cerebral cortex than non-exposed females without

this variant (Fig. 6; Paus et al., 2012).We have

specu-latedthat this gene-environmentinteractionreﬂects an

acceleratedapoptosisofprogenitorcellsin the

develop-ingbrainsofembryos/fetuseswhopossessthisparticular

geneticvariant(Pausetal.,2012).Aboveandbeyondglobal

braingrowth,wehaveobserveddifferencesbetweenthe

exposedandnon-exposed(female)adolescentsinthe

(rel-ative) size of the corpus callosum (Paus et al., 2008b)

andthethickness oftheorbitofrontal cortex,OFC(Toro

etal.,2008).Wefollowedupthelatterﬁndingandasked

whetherthereisarelationshipbetweentheOFCthickness

anddrugexperimentation;inthiscontext,wehave

exam-inedtheroleoftheknownfunctionalpolymorphisminthe

BDNFgene(andthemethylationstatusofitspromoters)in

moderatingthisrelationship(Lotﬁpouretal.,2009;

Toledo-Rodriguezetal.,2010).Wealsoinvestigatedarelationship

betweengeneticvariationsinalpha6nicotinicreceptor

gene(CHRNA6),striatalvolumeanddrugexperimentation

(Lotﬁpouretal.,2010).

Prenatal exposuretomaternalsmokingduring

preg-nancy is a well-established risk factor for obesity (Al

Mamunetal.,2006;Ino, 2010;Learyetal.,2006; Oken etal.,2005,2008;Poweretal.,2010;PowerandJefferis, 2002; Syme et al., 2010; von Kries et al., 2002; Weng

et al., 2012).In an exposed individual,it increases the

likelihoodfordevelopingobesityby50%,(Ino,2010;Weng

et al., 2012).Given this higher risk (odds ratio of 1.5),

andtheprevalenceofMSPinthe1960sand1970s(40%)

and currently (16%), we estimated that – at present –

(8)

Fig.6.Age-adjustedvalues(Mean±SE)ofbrainvolume(top),totalcorticalarea(middle)andcorticalfolding(bottom)infemaleadolescentsexposed(left column)andnotexposed(rightcolumn)tomaternalcigarettesmokingduringpregnancy,allplottedasafunctionoftheKCTD8genotype(rs716890;GG, GTandTTgenotypes;G,guanine;T,thymine).Theamountofvariance(r2₎_explained_by_the_genotype_and_statistical_{signiﬁcance}_of_the_genotype_effect_of eachphenotypeareindicated.NotethatweobservedsigniﬁcantinteractionbetweentheKCTD8genotype(rs716890)and“exposure”ontotalcorticalarea andcorticalfoldingbutnotonbrainvolume.

From(Pausetal.,2012).

obesityinchildrenisattributabletoMSP.Theunderlying

mechanisms of the link between MSP and obesity are

notclear,however.Our ﬁndingsintheSYSsuggestthat

reward-relatedmechanismsmaybeatplay.Weshowed

thatMSPisassociatedwithsubstantialincreasesinbody

adiposity(Symeetal.,2010),andhigherpreferenceforfat

accompanied by smaller amygdalae volumes (Haghighi

etal.,2013).Inagenome-wideassociationstudy,wealso

showedthatdietarypreference for fat (aswellasbody

adiposity)isassociatedwithgeneticvariationintheopioid

receptormu1gene(OPRM1)(Haghighietal.,2014).Finally,

wehavedemonstratedthatMSPisassociatedwith

modiﬁ-cationsofDNAmethylationthatpersistintoadolescenceof

theexposedoffspring(Leeetal.,2014a),andthatsomeof

thesemodiﬁcationsarepresentinOPRM1,andmayinhibit

expressionoftheprotective(fatintake-lowering)alleleof

thisgene(Leeetal.,2014b).Takentogether,these

observa-tionssuggest(a)thepresenceofrelationshipsbetweenthe

brain-rewardsystem,dietarypreferenceforfatand

obe-sity;(b)perturbationsoftheserelationshipsbyMSPand

geneticvariationsinOPRM1;and(c)DNAmethylationas

apossiblemolecularmechanismunderlyinginteractions

betweenenvironment(MSP)andgenes(OPRM1).

Apart from the above fat-preference angle, we also

found that a genetic variant of the FTO gene, the best

knowngeneticriskfactorforobesity,predictsaninverse

relationship between total brain volume and fat body

(9)

given the completion of overall brain growth in early

childhood, theseeffects mighthavetheirorigins during

earlydevelopmentandreﬂectadifferentialcommitment

ofprogenitorcellstoectoderm(brain)ormesoderm(fat

tissue).Finally,webeganexploringthedifferentialimpact

ofvisceralfatontheadolescent’sbrain(Schwartzetal.,

2014)andcognition(Schwartzetal.,2013).

Overall,theinitialﬁndingsobtainedinWave1ofthe

SYS cohort revealed a number of associations between

MSP,brain,andaddictivebehaviorofadolescentsinthe

contextofbothdrugexperimentationanddietaryfat

pref-erence. Although observational studiescannot attribute

solely suchassociationstoMSP (D’Onofrio etal., 2012),

humanstudieswithgeneticallysensitiveresearchdesigns

(Gaysina et al., 2013) and preclinical studies of

prena-talexposuretonicotine(Frankeetal.,2008)suggestthat

theseprenatalexposuresdoplayaroleinshaping

mental-healthtrajectories.Furthermore,carefulconsiderationof

potential confounders, suchas maternal education and

alcoholuseduringpregnancy,whenexaminingthe

rela-tionsbetweenexternalizing behaviorandsubstanceuse

(Lotﬁpouretal., 2014)or thespeciﬁcityof someofthe

observed structure–function relationships (e.g.,

correla-tion betweentheamygdalavolume withfat preference

but notwithalcoholuse;Haghighietal.,2013)suggest

thattheseassociationsarenotduetoglobalphenomena

(such as “poverty” or an “addictive personality” of the

parents). Finally, it is important to note that we have

observed a number of gene-exposure interactions that

argueagainstgeneralconfounding:withtheequal

distribu-tionofagivengeneticvariantbetweenthe“exposed”and

“non-exposed”adolescents,suchgene–environment

inter-actions point tospeciﬁc molecularpathways mediating

associationsbetweenMSPandagivenphenotype(CHRNA6

(Lotﬁpouretal.,2010);KCTD8(Pausetal.,2012);OPMR1 (Leeetal.,2014a)).Ingeneral,theuseofgeneticvariations

toascribecausalitytoobservationsmadeinlarge

epidemi-ologicalstudieshasbeenhelpfulinanumberofdomains

(Smithand Ebrahim, 2003).In thecontext ofaddiction,

we employthis approach in secondary analysescarried

outbythemembersofa consortiumcalled“The Causal

AnalysisResearchinTobaccoandAlcohol(CARTA)”,which

includes30 studies,withatotal of150,000participants

(http://goo.gl/dkXQyH).

4. SaguenayYouthStudy:parentarm

Aspointed out above,wehave obtainedbasic

infor-mationaboutthementalhealthandsubstanceuseofthe

biologicalparentsatthetimeofrecruitmentofthe

adoles-cents,aswellasabloodsampleforgeneticandepigenetic

analyses.InTables5and6,respectively,weprovide

summ-ariesofdemographicinformationabouttheparents,their

mentalhealthandsubstanceuseatthetimeoftheinitial

recruitment(Wave1).

In2012,weinitiatedWave2:ParentsoftheSYScohort.

Thiswavefocusesondeepphenotypingofthebiological

parentsoftheSYSadolescents.Parentscompleteaseries

ofon-linequestionnairesandvisit ourphenotypingunit

forMRIscansofthebrainandabdominalfat,bloodsample

for biochemistryand lipidomicsanalyses, aswellas for

Table5

Wave1:Baselineinparents(completed).

Measure Distribution

N 962

Numberoffamilies 481

Age(years) Mean=43.33;SD=4.58

Sex 50%male;50%female

Householdincome ≤$20,000–13% $30,000–40,000–19% $50,000–60,000–24% $70,000–80,000–20% ≥$85,000–24%

Education Nohighschool–1%

Somehighschool–15% Highschool–52% Collegedegree–19% Bachelors–9% Mastersordoctorate–3% Unknown–1% SD=standarddeviation.

cardio-metabolicandcognitiveassessments(Table7).The

visitlasts∼4h.

4.1. Internet-basedassessments

AsindicatedinTable7,wearere-administering

fam-ily environment, lifestyle and mental-health/substance

use questionnaires, as well as several new

question-nairesfocusingonparenting,sleep,personalityandvarious

addictivebehaviors,suchasfood,gamblingandinternet

addictions.

4.2. Face-to-faceassessment

Thevisit takesplace in the morningand lasts∼4h:

it includesa drawof blood(after overnightfasting) for

future “omics” analyses, a structured psychiatric

inter-view,cognitiveassessment,MRIand

cardiovascular/body-composition sessions, each lasting ∼60min. For the

psychiatricinterview,weusetheMini-International

Psy-chiatric Interview (MINI Plus, Sheehan et al., 1998)

administeredbyatrainedresearchassistant.MINIPlusisa

validated,structuredpsychiatricinterviewforcurrentand

lifetimeDSM-IVandICD-10psychiatricdisorders.Wehave

alsoaddedthenicotinedependencemodulefrom

Semi-Structured Assessment for the Genetics of Alcoholism,

SSAGA(Hesselbrocketal.,1999).

4.3. Cognitiveabilities

Cognitive abilities are assessed using a

vali-dated web-based battery for the assessment of

Table6

Wave1:Parents.Substanceuseandmentalhealth.

Mothers(%) Fathers(%) Smoking(current/former/never) 32/42/26 31/41/29

Cannabis(last12months) 5.2 11.2

Alcohol(bingedrinkinga₎ _49.8 _71.15

Depressionsymptoms(90thpercentile) 10.6 9.7 Anxietysymptoms(90thpercentile) 12.5 6.9 a₅_or_more_drinks_on_one_occasion_(at_least_once_in_the_last₁₂_months).

(10)

Table7

Wave2:parents(ongoing).

Brain MRI Globalandregionalvolumes;corticalsurface&thickness;

white-matterhyperintensities;magnetizationtransfer ratio;diffusiontensorimaging;resting-statefunctional MRI

Cognition CambridgeBrainSciencesPlatform Executivefunctioning;attention;learning&memory; reasoning;spatialskills

Mentalhealth MINIInternationalNeuropsychiatricInterview; MentalHealthandAddictionQuestionnaire;ASR; CES-D;FamilyHistoryScreen

Depression,anxiety,attentiondeﬁcithyperactivedisorder, antisocialpersonalitydisorder;post-traumaticstress disorder;obsessivecompulsivedisorder;alcoholand substancedependence,bulimia,anorexia;familyhistoryof psychiatricdisorders

Substanceuse&addiction MentalHealthandAddictionQuestionnaire;YFAS; FNDS;AUDIT;SRE;ESPAD;IAT;SOGS

Cigarettesmoking,alcoholanddruguse,gambling, internetaddition,foodaddiction

Personality NEO-FFI Neuroticism,extroversion,openness,agreeableness,

conscientiousness Lifestyle LifeExperiencesQuestionnaire;PBI;Hand

Preference

Familycharacteristics;education;socio-economicstatus; physicalactivity;sexualactivity;parentalstyle;hand laterality

Sleep PSQI;ESS Sleepquality,latency,duration,efﬁciencyand

disturbances;daytimesleepiness

Bodycomposition Anthropometry,MRI,Bioimpedance Height,weight,circumferences,skinfolds;subcutaneous, visceralfatandmusclevolumes;fat&musclemass Cardiovascular Finometer Beat-by-beatbloodpressureandheartrateatrestandin

responsetophysicalandmentalchallenges,sympathetic& parasympathetictone

LungFunction Spirometer Forcedvitalcapacity,forcedexpiratoryvolume

Diet 24-hfoodrecall Energyandnutrientintake

Medicalhistory MedicalQuestionnaire Personalandfamilyhistoryof:cancer,hypertension, diabetes,heartdisease,lipiddisease,psychiatricdisorders, addiction;reproductiveandsexualhealth;medications

Hormones Blood Testosterone,estrogen,cortisol

Biochemistry Blood Lipidproﬁle(TG,TC,HDL-C,LDL-C),glucose,insulin,free

fattyacids,glycerol,C-reactiveprotein

Lipidomics LC-ESI-MS(Blood) >700lipidspecies

MRI=magneticresonanceimaging;CambridgeBrainSciencesPlatform(Hampshire,Highﬁeld,Parkin,&Owen,2012);MINIInternationalNeuropsychiatric Interview(Sheehanetal.,1998);MentalHealthandAddictionQuestionnaire(AdaptedfromtheOntarioHealthStudywww.ontariohealthstudy.caandthe Wave-1questionnairedevelopedbytheSYSteam);ASR=AdultSelfReport(Achenbach&Rescorla,2003);CES-D=CenterforEpidemiologyStudies Depres-sionScale(Radloff,1977);FamilyHistoryScreen(Weissmanetal.,2000);YFAS=YaleFoodAddictionScale(Gearhardtetal.,2009);FNDS=Fagerström’s NicotineDependenceScale(Heathertonetal.,1991);AUDIT=AlcoholUseDisorderIdentiﬁcationTest(Barboretal.,1992);SRE=SubjectiveResponseto Ethanol(Schuckitetal.,1997);ESPAD=EuropeanSchoolSurveyProjectonAlcoholandOtherDrugs(Hibelletal.,2012);IAT=InternetAddictionTest (WidyantoandMcMurran,2004);SOGS=SouthOaksGamblingScreen(Lesieur&Blume,1987);NEO-FFI=NEO-FiveFactorInventory(CostaandMcCrae, 1992);LifeExperiencesQuestionnaire(AdaptedfromtheOntarioHealthStudywww.ontariohealthstudy.caandtheWave-1questionnairedevelopedbythe SYSteam);PBI=ParentalBondingInstrument(Parkeretal.,1979);HandPreference(AdaptedfromCrovitzandZener(1962));PSQI=PittsburghSleep Qual-ityIndex(Buysseetal.,1989);ESS=EpsworthSleepinessScale(Johns,1991);24-hFoodRecall(Buzzardetal.,1996);MedicalQuestionnaire(Adaptedfrom theOntarioHealthStudywww.ontariohealthstudy.ca);TG=triglycerides;TC=TotalCholesterol;HDL-C=Highdensitylipoprotein-cholesterol;LDL-C=low densitylipoprotein-cholesterol;LC-ESI-MS=liquidchromatographyelectrosprayionizationmassspectrometry.

cognition developed by Drs. Owen and Hampshire

(www.cambridgebrainscience.com).The battery is

com-prisedof12computer-basedtestsofexecutivefunction,

memory, learning and attention, and takes ∼35min

to complete. Population norms are available from two

large-scale public trials involving more than 100,000

participants(Hampshireetal.,2012).

4.4. TheMRIsession

The MRI session takes place at a private MR clinic

in Chicoutimi, which is equipped with a Siemens 1.5T

(Avanto)scanner.A60-minMRsessionincludesthe

fol-lowingimagingprotocols.StructuralMRIofthebrain:T1W,

1-mm,isotropicimagesacquiredwitha3DfastRF-spoiled

gradientechoscan.Magnetization-transfer(MT)ratio:MT

dataareacquiredusingadualacquisition(3DRF-spoiled

gradient echoscan)withand withoutan MTsaturation

pulse (3-mm thick axial slices, 1×1mm in-plane

res-olution). Diffusion Tensor Imaging (DTI): DTI is used to

assess thestructural properties of white matter.

Diffu-sionencodingisachievedusingasingle-shot,spin-echo,

echoplanarsequencewithtwice-refocusedbalanced

diffu-sionencodinggradients(64diffusion-encodingdirections,

3-mm thick axial slices, 2.3mm×2.3mm in-plane

res-olution). Resting-state functional MRI is acquired over a

period of12min(260 volumes,4-mm thickaxialslices,

3.5mm×3.5mmin-planeresolution,eyesclosed).Finally,

abdominalscansareacquiredusingheavilyT1-weighted,

spin-echo scans (30 axial 10-mm thick slices, 1-mm

gap,0.9mm×0.9mmin-planeresolution)extendingfrom

(11)

4.5. Thecardiovascular/bodycompositionsession

This session takes place at ECOGENE-21/Community

GenomicMedicineCenterinChicoutimi.Thefollowing

pro-tocolisidenticaltotheoneusedtocollectequivalentdatain

adolescentsduringWave1.Bloodpressure(BP)ismeasured

bothunderstandardclinicalconditionsandduringa

52-mincardiovascularreactivityprotocol.(a)Standardclinical

conditions:BPismeasuredatrestforatleast10minwhile

seatedwithastandardocclusioncuff(“Thefourthreport

onthediagnosis,evaluation,andtreatmentofhighblood pressureinchildrenandadolescents,”2004).This

measure-mentisdoneatthebeginningandendofthecardiovascular

reactivityprotocoldescribednext.(b)Cardiovascular

reac-tivityprotocollasts52minandmimicsdaily-lifeactivities,

includingchangesinpostureandmentalstress;BPanda

numberofothercardiovascularparametersaremeasured

beat-by-beatusinganon-invasivehemodynamicmonitor,

Finometer (FMSFinapres,Amsterdam,TheNetherlands;

see below).The posture test consistsof 3 periods

dur-ing which theparticipant rests ina supineposition for

10min,standsfor10min,andsitsfor10min.Themental

stresstestinvolvesa30-sexplanationadministered5min

priortoa2-minsequenceof23simplearithmetic

prob-lems,eachpresentedfor5s;theproblemsincludesimple

mathadditions orsubtractionsfollowedbysimple

mul-tiplications ordivisions.The levelof difﬁcultyincreases

progressivelywithtimetoensuresomefailuresforall

par-ticipants;allanswersarerecorded.Themathsequenceis

followedbya10-minperiodofrestinginasitting

posi-tion. Throughout this protocol, a Finometer is used to

recordcontinuouslytheﬁngerbloodﬂow.TheFinometer

derivesbeat-by-beatbrachialsystolicanddiastolicBPfrom

the reconstructed and level-corrected ﬁngerblood-ﬂow

waveform. TheFinometer is a reliabledevice for

track-ingBPinadultsandchildrenolderthansixyears(Parati

etal.,1989;Tanakaetal.,1994)andtheprecisionofBP

measurementwiththisdevicemeetstherequirementsof

theAmericanAssociationfortheAdvancementofMedical

Instruments(Guelenetal.,2008;Westerhofetal.,2002).

BodyComposition:Bodyweight,height,sixcircumferences

(upperarm,waist,hips,proximalthigh,middlethighand

distalthigh),andﬁveskinfolds(triceps,biceps,

subscapu-lar,suprailiacandmid-thigh)aremeasuredaccordingto

standard procedures(Pausovaetal., 2001).Bioelectrical

impedanceisusedtomeasuretotalbodyfat,totalbody

waterandfat-freemass.Participantsareaskedtorefrain

fromcaffeine, alcohol,and vigorous activity24hbefore

thetest.Theactualmeasurementismadeaftera20-min

stabilizationperiodduringwhichtheparticipantsrestin

asupineposition.Thesemeasuresarecomplementedby

thequantiﬁcationofsubcutaneousandvisceralfatderived

fromabdominalMRIs(seeabove).Giventhehigh

preva-lenceofsmokersinthiscohort(duetotheascertainment

oftheoriginalsample,withhalfofthemotherssmoking

duringpregnancy),wehavealsoincludedassessmentof

lungfunctionusingspirometry.Usingaspirometer

(Min-iSpirbyMedicalInternationalResearch[MIR],Rome,Italy

(www.spirometry.com);MiniSpir User’sManual.Manual

revision1.3.2006.MedicalInternationalResearch;User’s

Manual Code 980255)wemeasure theamountandthe

rateatwhichaparticipantexhalesinasinglebreath.The

standardspirometrictestrequirestheparticipanttoexhale

asforcefullyaspossibleaftertakingafullinspiration.This

measurestheparticipant’sforcedvitalcapacityaswellas

theirforcedexpiratoryvolumeinonesecond.

AsofSeptember2014,wehaveacquiredfulldatasetsin

573parents.Wearealsointheprocessofcontactingthe

livinggrandparentsoftheadolescentsinordertoobtain

asalivasample(forgeneticandepigeneticanalyses)and

basicinformationabouttheirmentalhealthandsubstance

use.Finally,wehaveobtainedapprovalforreceiving

infor-mationcontained in deathcertiﬁcates for thedeceased

parentsandgrandparentsoftheSYSadolescents.

5. Largedevelopmentalcohorts:designstrategies andchallenges

TheSaguenayYouthStudy(anditsparentalarm)isone

amongagrowingnumberoflargecommunity-based

stud-ies(IMAGEN(Schumannetal.,2010);GenerationR(White

etal.,2013);PING(Fjelletal.,2012))thatcombinebrain

imagingwithgenetics,aswellaswithadetailed

assess-mentofcognition,mentalhealthandfamilyenvironment.

Theoverarchinggoalofthesestudiesistogain insights

intofactors(andmechanisms)shapingthebrainin

typ-icallydevelopingchildrenandadolescents.Thesestudies

facesimilarchallengeswithregardstotheascertainment

oftheirparticipants(andrepresentativenessofthe

sam-ples),thechoiceofneuroimagingprotocols(structuralvs.

functional),assessmentsofcognitionandmentalhealth,

andinvolvementofotherfamilymembers.Wewilladdress

brieﬂysomeoftheseissuesinthefollowingtext(fordetails,

seePaus,2010,2013).

Ideally, ascertainment of participants in

population-based studies should be free of selection biases, thus

creatingconditionsforgeneratingdatarepresentativeof

the generalpopulation – “a representative brain” (Falk

etal.,2013).Aswepointedoutelsewhere,previous

imag-ingcohortsuseddifferentrecruitmentstrategies(samples

ofconveniencevs.census-basedsampling)andexclusion

criteria(MRcontraindicationsonlyvs.screeningout

chil-drenwithanypersonalandfamily-basedriskfactors);not

surprisingly,someofthesestrategies yielded

“supernor-mal”samples(Paus,2010).IntheSaguenayYouthStudy,

wehavecarriedoutrecruitmentinallpublichighschools

intheregionandexcludedfromparticipationonly

adoles-centswithMRIcontraindicationsand seriousconditions

likelytoaffectthebrain(e.g.,epilepsy)orheart(e.g.,heart

defects)development.Bydesign, thesampleisenriched

byindividualsborntomotherssmokingcigarettesduring

pregnancy(50%vs.∼20%expectedingeneralpopulation).

Other–moresubtle–biasesincludetherequirementof

havingsiblingsandbeingabletocontactbothbiological

parents. The latter conditions are, however, unlikely to

reducerepresentativenessofthesample,asthemean

num-berofchildrenperfamilyinthegeneralpopulationis1.5

(Quebec,2014),andthetwo-parentrequirementdidnot

demandcohabitation.Recentreplicationsofthe

relation-ship betweenexternalizing behavior and substance use

duringadolescencein twogeographicallyand culturally

(12)

Table8

Anexampleofa60-minMRprotocolenablingonetocharacterizeanumberofstructuralandfunctionalpropertiesofthehumanbrain.FromPaus,2013.

MRIsequence Time(min) Structureandphysiology

T1-weighted 10 Volumes,thickness,folding,shape,tissuedensity

T2-weighted 4 Whitematterhyperintensities(number,volume,location)

Diffusiontensorimaging 12 Fractionalanisotropy,meandiffusivity,trackdelineation

Magnetizationtransfer 8 Myelinationindex

Arterialspinlabeling 5 Perfusion

Restingstatefunctional 8 Spontaneouscerebralnetworks;functionalconnectivity

Paradigm-basedfunctional 6–10 Brainresponseassociatedwithspeciﬁcstimuli/tasks;functionalconnectivity

namely the Saguenay Youth Study and the Northern

FinlandBirthCohort1986,suggestthatﬁndingsobtained

inoursamplearegeneralizable(Lotﬁpouretal.,2014).

Brainimagingrepresentsauniquetoolallowingoneto

obtainawidearrayofquantitativephenotypes(Table8).

A number of considerationsare at play when choosing

speciﬁc MR sequences. For example, studies of

typi-callydevelopingchildrenandadolescentsaremorelikely

to include scans sensitive to changes in myelination

(e.g.,magnetizationtransferratio ormyelin water

frac-tion, Dean et al., 2014) rather than those sensitive to

white-matterhyperintensities,which aremorecommon

in the aging brain (e.g., T2 ﬂuid attenuated inversion

recovery [FLAIR]). But perhaps the most relevant

con-siderations relate to the trait (vs. state) qualities of a

given measure;afterall,we base mostof our

develop-mental work on the assumption that genes and early

environments shape brain function and structure in a

stableandlong-termmanner.Therefore,test–retest

reli-ability of imaging-derived measures is paramount. Not

surprisingly,variousmetricsderivedfrom(multi-modal)

structural images show high test–retest reliability. For

example,Wonderlick andcolleagues (Wonderlick etal.,

2009)haveevaluatedtest–retestreliability(twosessions,

2weeks apart),andtheinﬂuence ofseveralacquisition

parameters(same3Tscanner),foranumberof

morpho-metric measures derived from T1-weighted images by

FreeSurfer. They found that the reliability – estimated

withintra-classcorrelationcoefﬁcients(ICCs)–was

“excel-lent”formostmeasures;withtheexception theglobus

pallidus,allICCsvalues wereabove0.95.Thetest-retest

reliabilityofDTI-basedmeasuresappearstovaryacross

themeasuresand ﬁbertracts.For example,Wangetal.

(2012)foundexcellentreliability(ICCs>0.75)forthemean

lengthofthecorpuscallosumandtheuncinatefasciculus,

andfairreliabilities(ICCsbetween0.4and0.75)for

frac-tionalanisotropyinmostﬁbertracts.Ontheotherhand,

test–retestreliabilityofdataobtainedwithfMRIhasbeen

characterizedas“fair”(ICC:0.4–0.75)inadultsand

ado-lescents,and“poor” (ICC<0.4)inchildren;itislowerin

regionswithweak“activation”, as revealedbygroup

t-maps(Caceresetal.,2009;Koolschijnetal.,2011;Plichta

etal., 2012).The relativelylow test–retestreliability of

functional datais likely related to a number of factors,

includingthefactthatthefMRIsignalisanindirectmeasure

ofbrainactivity,itsmeasurementisaffectedbyanumber

ofnoise-generatingfactors(e.g.,headmotion,

physiolog-ical“noise”relatedtorespirationandcardiaccycle)and,

most importantly, by the state of the participant

dur-ingscanning.Thelatter factors,suchasinter-individual

and session-by-sessionvariations intask-related

behav-ior(performance,attention)and generalstateofarousal

(anxiety,sleepiness)areverydifﬁculttoassessand

con-trol,therebyaddingsigniﬁcanterrortothemeasurementof

thefunctionalphenotype.Fortheabovereasons,we

advo-cateimagingprotocolsthatputemphasisonmulti-modal

imagingofbrainstructure.Weconsiderbrainstructure“a

windowintotheindividual’s lifehistory”,a notion

sup-portedbythewealthofimagingdataonexperience-related

brainplasticity(Lovdenetal.,2013).

Brainimaging(andgenetic)dataaloneareinsufﬁcient

for trackingtrajectoriesinbraindevelopment.A

signiﬁ-canttimecommitmentmustbemadetotheassessmentof

cognitionandmentalhealth.IntheSaguenayYouthStudy,

forexample,wehavespent∼8hoftheparticipant’stime

inthesedomains.Givenalargevarietyofapproachesand

availabletools,onlybroadrecommendationscanbemade.

Whenassessing cognition,wehave useda combination

ofstandardized(e.g.,WISC-III)andcomputer-based(e.g.,

auditoryprocessing)toolscombinedintwo3-hsessions

(adolescents).Inadults,wehavedecidedtousea1-h

(stan-dardized)batteryofcognitivetests.Inthemental-health

domain, we have focused on self-reported symptoms

ratherthanusingadiagnosis-driven(psychiatric-interview

based)approach.Conceptually,thisstrategyisconsistent

withamoveawayfromcategoricaldeﬁnitionsof

psychi-atricdisorders,asdeﬁnedintheDiagnosticandStatistical

ManualofMentalDisorders(DSM),andtowardsymptoms

as a preferredlevel ofanalysis (Borsboom etal., 2011).

Fromamethodologicalstandpoint,self-reportsarearich

sourceofreliableinformation,especiallyinthecontextof

substanceuse(SobellandSobell,1990).

Finally, any developmental study must consider the

caregivers. Asillustratedin Fig.4,parentsare themain

source of geneticand environmental inﬂuences onthe

developing brain. For this reason, we have obtained

parental DNA and basic information about the mental

healthoftheparents(includinganti-socialbehavior

dur-ingtheiradolescence)inWave1oftheSaguenayYouth

Study, andhave embarkedon deepphenotyping ofthe

parentsinWave2.Parentsarealsoakeysourceof

infor-mationaboutthelifeeventsencounteredbytheirchildren

atdifferentstagesofdevelopment;prospective

longitudi-nalbirthcohorts,suchastheAvonLongitudinalStudyof

ParentsandChildren(Boydetal.,2013),NorthernFinland

BirthCohort1986(Taanila etal.,2004)orGenerationR

(Jaddoeetal.,2010)arein anadvantageouspositionto

usesuchinformationtopredictthestateofbrain

devel-opment(andmentalhealth)atlaterpointinthelivesof

(13)

6. Conclusions

Aspointedoutin Introduction(Section1.1), the

life-spanperspectiveonhealthtrajectoriesreﬂectstheconcept

of developmental cascades: transactions occurring over

time, as wellasacrosssystems and organs. The

Sague-nayYouthStudyanditsparentarmattemptstointegrate

detailed information about brain and cardio-metabolic

health acquiredatthesystemand molecularlevelsina

family-basedmulti-generationalcontext.Wehopethatthe

richnessofthedatasetwillallowustocontributetoward

currenteffortsaimedatdistinguishingthekeyprocessesof

healthytrajectoriesfromthoseleadingtocommonchronic

disordersofthebrainandbody.

Conﬂictofinterest

Noneoftheauthorshasanyactualorpotentialconﬂict

ofinterestincludinganyﬁnancial,personalorother

rela-tionshipswithotherpeopleororganizationswithinthree

years ofbeginningthesubmittedworkthatcould

inap-propriatelyinﬂuence,orbeperceivedtoinﬂuence,their

work.

Acknowledgements

TheSaguenayYouthStudyanditsparentarmarefunded

bytheCanadianInstitutesofHealthResearch(TP,ZP),Heart

andStrokeFoundationofQuebec(ZP),andtheCanadian

FoundationforInnovation(ZP).Wethankallfamilieswho

tookpartintheSaguenayYouthStudyandthefollowing

individualsfortheircontributionsindesigningthe

proto-colandacquiringthedataintheParentarmofthestudy:

Dr.SoljaNiemela,Prof.JuhaVeijola, Dr.VesaKiviniemi,

Dr.RosanneAleong,CourtneyGray,HélèneSimard,Annie

GauthierandtheECOGENE-21/CMGCstaff.WethankDr.

Michel Berube for the radiological review of MR scans

withincidentalﬁndings,ManonBernardfordesigningand

managingouronlinedatabase,andDeborahSchwartzand

AngelitaWongfortheirhelpwithpreparingthetablesand

reviewingtheﬁrstdraftofthemanuscript.

References

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

AlMamun,A.,Lawlor,D.A.,Alati,R.,O’Callaghan,M.J.,Williams,G.M., Najman,J.M.,2006.Doesmaternalsmokingduringpregnancyhavea directeffectonfutureoffspringobesity?Evidencefromaprospective birthcohortstudy.Am.J.Epidemiol.164(4),317–325.

Anblagan,D.,Jones,N.W.,Costigan,C.,Parker,A.J.,Allcock,K.,Aleong,R., Gowland,P.A.,2013.Maternalsmokingduringpregnancyandfetal organgrowth:amagneticresonanceimagingstudy.PLOSONE8(7), e67223,http://dx.doi.org/10.1371/journal.pone.0067223.

Anon,2013.Substanceabuseandmentalhealthservices administra-tion,resultsfromthe2012NationalSurveyonDrugUseandhealth: summary of national ﬁndings. (HHS Publication No. (SMA) 13-4795.).SubstanceAbuseandMentalHealthServicesAdministration, Rockville,MD,Retrievedfromhttp://store.samhsa.gov/home Barbor,T.F.,delaFuente,J.R.,Saunders,J.,Grant,M.,1992.TheAlcohol

UseDisorderIdentiﬁcationTest:Guidelinesforuseinprimaryhealth care.WorldHealthOrganization,Geneva,Switzerland.

Barker,D.J.,Winter,P.D.,Osmond,C.,Margetts,B.,Simmonds,S.J.,1989. Weightininfancyanddeathfromischaemicheartdisease.Lancet2 (8663),577–580.

Barnes, D.E., Yaffe, K., 2011. The projected effect of risk fac-tor reduction on Alzheimer’s disease prevalence. Lancet Neurol. 10(9),819–828,http://dx.doi.org/10.1016/S1474-4422(11)70072-2, pii:S1474-4422(11)70072-2.

Borsboom,D.,Cramer,A.O.,Schmittmann,V.D.,Epskamp,S.,Waldorp,L.J., 2011.Thesmallworldofpsychopathology.PLoSONE6(11),e27407, http://dx.doi.org/10.1371/journal.pone.0027407.

Boyd,A.,Golding,J.,Macleod,J.,Lawlor,D.A.,Fraser,A.,Henderson,J., DaveySmith,G.,2013.CohortProﬁle:the‘childrenofthe90s’–the indexoffspringoftheAvonLongitudinalStudyofParentsand Chil-dren.Int.J.Epidemiol.42(1),111–127,http://dx.doi.org/10.1093/ ije/dys064.

Buysse,D.J.,Reynolds3rd,C.F.,Monk,T.H.,Berman,S.R.,Kupfer,D.J.,1989. ThePittsburghSleepQualityIndex:anewinstrumentforpsychiatric practiceandresearch.PsychiatryRes.28(2),193–213.

Buzzard, I.M., Faucett, C.L., Jeffery, R.W., McBane, L.,McGovern, P., Baxter, J.S., Wynder, E.L., 1996. Monitoring dietary change in a low-fat diet intervention study: advantagesof using 24-hour dietaryrecallsvsfoodrecords.J.Am.Diet.Assoc.96(6),574–579, http://dx.doi.org/10.1016/S0002-8223(96)00158-7.

Caceres,A.,Hall,D.L.,Zelaya,F.O., Williams,S.C., Mehta,M.A.,2009. MeasuringfMRI reliability with the intra-classcorrelation coef-ﬁcient. Neuroimage 45 (3), 758–768, http://dx.doi.org/10.1016/j. neuroimage.2008.12.035,pii:S1053-8119(08)01327-X.

Cornelius,M.D.,Day,N.L.,2009.Developmentalconsequencesofprenatal tobaccoexposure.Curr.Opin.Neurol.22(2),121–125.

Costa,P.T.,McCrae,R.R.,1992.RevisedNEOPersonalityInventory (NEO-PI-R)andNEOFiveFactoryInventory(NEO-FFI)professionalmanual. PsychologicalAssessmentResources,Odesa,FL.

Crovitz,H.F.,Zener,K.,1962.Agroup-testforassessinghand-and eye-dominance.Am.J.Psychol.75,271–276.

D’Onofrio, B.M., Rickert, M.E., Langstrom, N., Donahue, K.L., Coyne, C.A.,Larsson, H., Lichtenstein, P., 2012. Familial confounding of theassociationbetweenmaternalsmokingduringpregnancyand offspring substance useand problems.Arch.Gen. Psychiatry69 (11),1140–1150,http://dx.doi.org/10.1001/archgenpsychiatry.2011. 2107.

DeBraekeleer,M.,1991.HereditarydisordersinSaguenay-Lac-St-Jean (Quebec,Canada).Hum.Hered.41(3),141–146.

DeBraekeleer,M.,Mari,C.,Verlingue,C.,Allard,C.,Leblanc,J.P.,Simard, F.,Ferec,C.,1998.Completeidentiﬁcationofcysticﬁbrosis trans-membraneconductanceregulatormutationsintheCFpopulation ofSaguenayLac-Saint-Jean(Quebec,Canada).Clin.Genet.53(1), 44–46.

Dean3rd,D.C.,O’Muircheartaigh,J.,Dirks,H.,Waskiewicz,N.,Lehman, K.,Walker,L.,Deoni,S.C.,2014.Modelinghealthymalewhitematter andmyelindevelopment:3through60monthsofage.Neuroimage 84,742–752,http://dx.doi.org/10.1016/j.neuroimage.2013.09.058. Falk,E.B.,Hyde,L.W.,Mitchell,C.,Faul,J.,Gonzalez,R.,Heitzeg,M.M.,

Schu-lenberg,J.,2013.Whatisarepresentativebrain?Neurosciencemeets populationscience.Proc.Natl.Acad.Sci.U.S.A.110(44),17615–17622, http://dx.doi.org/10.1073/pnas.1310134110.

Fjell,A.M.,Walhovd,K.B.,Brown,T.T.,Kuperman,J.M.,Chung,Y.,Hagler Jr., D.J., 2012. Multimodal imaging of the self-regulating devel-opingbrain. Proc.Natl.Acad.Sci. U.S.A.109(48), 19620–19625, http://dx.doi.org/10.1073/pnas.1208243109.

Thefourthreportonthediagnosis,evaluation,andtreatmentofhighblood pressureinchildrenandadolescents.,2004.Pediatrics114(2Suppl 4thReport),555–576.

Franke,R.M.,Park,M.,Belluzzi,J.D.,Leslie,F.M.,2008.Prenatal nico-tine exposure changes natural and drug-induced reinforcement in adolescent male rats. Eur. J. Neurosci. 27 (11), 2952–2961, http://dx.doi.org/10.1111/j.1460-9568.2008.06253.x,EJN6253. Gaysina,D.,Fergusson,D.M.,Leve,L.D.,Horwood,J.,Reiss,D.,Shaw,

D.S.,Harold,G.T.,2013.Maternalsmokingduringpregnancyand offspringconductproblems:evidencefrom3independent geneti-callysensitiveresearchdesigns.JAMAPsychiatry70(9),956–963, http://dx.doi.org/10.1001/jamapsychiatry.2013.127.

Gearhardt,A.N.,Corbin,W.R.,Brownell,K.D.,2009. Preliminary vali-dationoftheYaleFoodAddictionScale.Appetite52(2),430–436, http://dx.doi.org/10.1016/j.appet.2008.12.003.

Goodwin, K., Syme, C., Abrahamowicz, M., Leonard, G.T.,Richer, L., Perron,M.,Pausova,Z.,2013.Routineclinicalmeasuresof adipos-ityaspredictorsofvisceralfatinadolescence:apopulation-based magnetic resonance imaging study. PLOS ONE 8 (11), e79896, http://dx.doi.org/10.1371/journal.pone.0079896.

Gradie, M.I., Jorde, L.B., Bouchard, G., 1988. Genetic structure of the Saguenay, 1852–1911: evidence from migration and isonymy matrices. Am. J. Phys. Anthropol. 77 (3), 321–333, http://dx.doi.org/10.1002/ajpa.1330770305.

(14)

Grompe,M.,St-Louis,M.,Demers,S.I.,al-Dhalimy,M.,Leclerc,B.,Tanguay, R.M.,1994.Asinglemutationofthefumarylacetoacetatehydrolase geneinFrenchCanadianswithhereditarytyrosinemiatypeI.N.Engl. J.Med.331(6),353–357.

Guelen,I.,Westerhof,B.E.,vanderSar,G.L.,vanMontfrans,G.A.,Kiemeneij, F.,Wesseling,K.H.,Bos,W.J.,2008.Validationofbrachialartery pres-surereconstructionfromﬁngerarterialpressure.J.Hypertens.26(7), 1321–1327.

Guyer,B.,Freedman,M.A.,Strobino,D.M.,Sondik,E.J.,2000.Annual sum-maryofvitalstatistics:trendsinthehealthofAmericansduringthe 20thcentury.Pediatrics106(6),1307–1317.

Haghighi,A., Melka,M.G., Bernard, M.,Abrahamowicz, M., Leonard, G.T.,Richer,L.,Pausova,Z.,2014.Opioidreceptormu1gene,fat intakeandobesityinadolescence.Mol.Psychiatry19(1),63–68, http://dx.doi.org/10.1038/mp.2012.179.

Haghighi,A.,Schwartz,D.H.,Abrahamowicz,M.,Leonard,G.T.,Perron, M.,Richer, L.,Pausova, Z., 2013. Prenatal exposure to maternal cigarette smoking, amygdala volume, and fat intake in adoles-cence.JAMAPsychiatry70(1),98–105,http://dx.doi.org/10.1001/ archgenpsychiatry.2012.1101,pii:1356544.

Hampshire, A., Highﬁeld, R.R., Parkin, B.L., Owen, A.M., 2012. Fractionating human intelligence. Neuron 76 (6), 1225–1237, http://dx.doi.org/10.1016/j.neuron.2012.06.022.

Heatherton,T.F.,Kozlowski,L.T.,Frecker,R.C.,Fagerstrom,K.O.,1991.The Fagerstromtestfornicotinedependence:arevisionoftheFagerstrom tolerancequestionnaire.Br.J.Addict.86(9),1119–1127.

Herve,P.Y.,Leonard,G.,Perron,M.,Pike,B.,Pitiot,A.,Richer,L.,Paus,T., 2009.Handedness,motorskillsandmaturationofthecorticospinal tractintheadolescentbrain.Hum.BrainMapp.30(10),3151–3162. Hesselbrock,M.,Easton,C.,Bucholz,K.K.,Schuckit,M.,Hesselbrock,V.,

1999.AvaliditystudyoftheSSAGA–acomparisonwiththeSCAN. Addiction94(9),1361–1370.

Hibell,B.,Guttormsson,U.,Ahlstrom,S.,Balakireva,O.,Bjarnason,T., Kokkevi,A.,Kraus,L.,2012.The2011ESPADReport.SubstanceUse AmongStudentsin36EuropeanCountries.Stockholm,Sweden. Hofman,A.,vanDuijn,C.M.,Franco,O.H.,Ikram,M.A.,Janssen,H.L.,Klaver,

C.C.,Witteman,J.C.,2011.TheRotterdamStudy:2012objectivesand designupdate.Eur.J.Epidemiol.26(8),657–686,http://dx.doi.org/10. 1007/s10654-011-9610-5.

Ino,T.,2010.Maternalsmokingduringpregnancyandoffspringobesity: meta-analysis.Pediatr.Int.52(1),94–99,http://dx.doi.org/10.1111/ j.1442-200X.2009.02883.x.

Jaddoe,V.W.,vanDuijn,C.M.,vanderHeijden,A.J.,Mackenbach,J.P., Moll,H.A.,Steegers,E.A.,Hofman,A.,2010.ThegenerationRstudy: designandcohortupdate2010.Eur.J.Epidemiol.25(11),823–841, http://dx.doi.org/10.1007/s10654-010-9516-7.

Johns,M.W.,1991.Anewmethodformeasuringdaytimesleepiness:the Epworthsleepinessscale.Sleep14(6),540–545.

Kandel,D.B.,Griesler,P.C.,Schaffran,C.,2009.Educationalattainmentand smokingamongwomen:riskfactorsandconsequencesforoffspring. DrugAlcoholDepend.104(Suppl.1),S24–S33,http://dx.doi.org/10. 1016/j.drugalcdep.2008.12.005.

Koolschijn,P.C.,Schel,M.A.,deRooij,M.,Rombouts,S.A.,Crone,E.A., 2011.Athree-yearlongitudinalfunctionalmagneticresonance imag-ingstudyofperformancemonitoringandtest-retestreliabilityfrom childhood to early adulthood. J. Neurosci. 31 (11), 4204–4212, http://dx.doi.org/10.1523/JNEUROSCI.6415-10.2011.

Leary,S.D.,Smith,G.D.,Rogers,I.S.,Reilly,J.J.,Wells,J.C.,Ness,A.R.,2006. Smokingduringpregnancyandoffspringfatandleanmassin child-hood.Obesity(SilverSpring)14(12),2284–2293.

Lee,K.W.K.,Richmond,R.,Hu,P.,French,L.,Shin,J.,Bourdon,C.,Reischl, E.,Waldenberger,M.,Zeilinger,S.,Gaunt,T.,McArdle,W.,Ring,S., Woodward,G.,Bouchard,L.,Gaudet,G.,Davey-Smith,G.,Relton,C., Paus,T.,Pausova,Z.,2014a.Prenatalexposuretomaternalcigarette smokingisassociatedwithlastingmodulationsofDNAmethylation intheexposedoffspring.Environ.HealthPerspect.

Lee,K.W.K.,Abrahamowicz,M.,Leonard,G.T.,Richer,L.,Perron,M., Veil-lette,S.,Bouchard,L.,Gaudet,D.,Paus,T.,Pausova,Z.,2014b.Prenatal exposureto cigarette smokeinteracts withOPRM1 to modulate dietaryintakeoffat:roleofDNAmethylation.J.PsychiatryNeurosc. Lesieur,H.R.,Blume,S.B.,1987.TheSouthOaksGamblingScreen(SOGS):

anewinstrumentfortheidentiﬁcationofpathologicalgamblers.Am. J.Psychiatry144(9),1184–1188.

Liu, D.,Diorio, J.,Tannenbaum, B.,Caldji, C., Francis, D.,Freedman, A.,Meaney,M.J.,1997.Maternalcare,hippocampalglucocorticoid receptors,andhypothalamic-pituitary-adrenalresponsestostress. Science277(5332),1659–1662.

Lotﬁpour,S.,Ferguson,E.,Leonard,G.,Miettunen,J.,Perron,M.,Pike, G.B.,Paus,T.,2014.Maternalcigarettesmokingduringpregnancy

predictsdrug useviaexternalizingbehavior intwo community-based samples of adolescents. Addiction 109 (10), 1718–1729, http://dx.doi.org/10.1111/add.12665.

Lotﬁpour,S.,Ferguson,E.,Leonard,G.,Perron,M.,Pike,B.,Richer,L., Paus,T.,2009.Orbitofrontalcortexanddruguseduringadolescence: roleofprenatalexposuretomaternalsmokingandBDNFgenotype. Arch.Gen.Psychiatry66(11),1244–1252,http://dx.doi.org/10.1001/ archgenpsychiatry.2009.124,pii:66/11/1244.

Lotﬁpour,S.,Leonard,G.,Perron,M.,Pike,B.,Richer,L.,Seguin,J.R.,Paus, T.,2010.Prenatalexposuretomaternalcigarettesmokinginteracts withapolymorphisminthealpha6nicotinicacetylcholinereceptor genetoinﬂuencedruguseandstriatumvolumeinadolescence.Mol. Psychiatry15(1),6–8.

Lovden, M., Wenger, E., Martensson, J., Lindenberger, U., Back-man, L., 2013. Structural brain plasticity in adult learning and development. Neurosci. Biobehav.Rev. 37(9 Pt B), 2296–2310, http://dx.doi.org/10.1016/j.neubiorev.2013.02.014.

Lowe,C.R.,1959.Effectofmothers’smokinghabitsonbirthweightoftheir children.Br.Med.J.2(5153),673–676.

Mareckova,K.,Chakravarty, M.M.,Huang, M.,Lawrence, C.,Leonard, G.,Perron,M.,Paus,T.,2013.Doesskullshapemediatethe rela-tionship between objective features and subjective impressions abouttheface?Neuroimage79,234–240,http://dx.doi.org/10.1016/ j.neuroimage.2013.04.110.

Mareckova, K., Weinbrand, Z., Chakravarty, M.M., Lawrence, C., Aleong, R., Leonard, G., Paus, T., 2011. Testosterone-mediated sex differences in the face shape during adolescence: subjec-tive impressions and objective features. Horm. Behav. 60 (5), 681–690,http://dx.doi.org/10.1016/j.yhbeh.2011.09.004, pii:S0018-506X(11)00215-7.

Masten, A.S., Cicchetti, D., 2010. Developmental cascades. Dev. Psychopathol. 22 (3), 491–495, http://dx.doi.org/10.1017/ S0954579410000222.

Mathers,C.D.,Iburg,K.M.,Salomon,J.A.,Tandon,A.,Chatterji,S.,Ustun, B.,Murray, C.J.,2004. Globalpatternsof healthylife expectancy in theyear 2002. BMC public health 4, 66,http://dx.doi.org/10. 1186/1471-2458-4-66.

Melka,M.G.,Abrahamowicz,M.,Leonard,G.T.,Perron,M.,Richer,L., Veil-lette,S.,Pausova,Z.,2013a.Clusteringofthemetabolicsyndrome componentsinadolescence:roleofvisceralfat.PLOSONE8(12), e82368,http://dx.doi.org/10.1371/journal.pone.0082368.

Melka,M.G.,Bernard,M.,Mahboubi,A.,Abrahamowicz,M.,Paterson,A.D., Syme,C.,Pausova,Z.,2012.Genome-widescanforlociofadolescent obesityandtheirrelationshipwithbloodpressure.J.Clin.Endocrinol. Metab. 97 (1), E145–E150, http://dx.doi.org/10.1210/jc.2011-1801.

Melka,M.G.,Gillis,J.,Bernard,M.,Abrahamowicz,M.,Chakravarty,M.M., Leonard,G.T.,Pausova,Z.,2013b.FTO,obesityandtheadolescent brain.Hum.Mol.Genet.22(5),1050–1058,http://dx.doi.org/10.1093/ hmg/dds504.

Miller,D.B.,O’Callaghan,J.P.,2008.Doearly-lifeinsultscontributeto the late-lifedevelopment of Parkinson and Alzheimer diseases? Metabolism 57 (Suppl. 2), S44–S49, http://dx.doi.org/10.1016/j. metabol.2008.07.011.

Oken,E.,Huh,S.Y.,Taveras,E.M.,Rich-Edwards,J.W.,Gillman,M.W.,2005. Associationsofmaternalprenatalsmokingwithchildadiposityand bloodpressure.Obes.Res.13(11),2021–2028.

Oken,E.,Levitan,E.B.,Gillman,M.W.,2008.Maternalsmokingduring pregnancy and child overweight: systematic review and meta-analysis.Int.J.Obes.(Lond.)32(2),201–210,http://dx.doi.org/10. 1038/sj.ijo.0803760.

Parati,G.,Casadei,R.,Groppelli,A.,DiRienzo,M.,Mancia,G.,1989. Com-parisonofﬁngerandintra-arterialbloodpressuremonitoringatrest andduringlaboratorytesting.Hypertension13(6(Pt1)),647–655. Parker,G.,Tupling,H.,Brown,L.,1979.Aparentalbondinginstrument.Br.

J.Med.Psychol.52,1–10.

Paus,T.,2010.Populationneuroscience:whyandhow.Hum.BrainMapp. 31(6),891–903,http://dx.doi.org/10.1002/hbm.21069.

Paus,T.,2013.PopulationNeuroscience.Springer-Verlag,Berlin, Heidel-berg.

Paus,T.,Bernard,M.,Chakravarty,M.M.,DaveySmith,G.,Gillis,J., Lour-dusamy,A.,Pausova,Z.,2012.KCTD8geneandbraingrowthinadverse intrauterineenvironment:agenome-wideassociationstudy.Cereb. Cortex22(11),2634–2642,http://dx.doi.org/10.1093/cercor/bhr350. Paus,T.,Keshavan,M.,Giedd,J.N.,2008a.Whydomanypsychiatric disor-dersemergeduringadolescence?Nat.Rev.Neurosci.9(12),947–957, http://dx.doi.org/10.1038/nrn2513.

Paus, T., Nawazkhan, I., Leonard, G., Perron, M., Pike, G.B., Pitiot, A., Pausova,Z., 2008b. Corpus callosum in adolescent offspring