edinburgh research explorer - university of edinburgh finkel1, carol e. franz2, briana horwitz3,...

Edinburgh Research Explorer

Gender differences in marital status moderation of genetic andenvironmental influences on subjective health

Citation for published version:Finkel, D, Franz, CE, Horwitz, B, Christensen, K, Gatz, M, Johnson, W, Kaprio, J, Korhonen, T, Niederhiser,J, Petersen, I, Rose, RJ & Silventoinen, K 2016, 'Gender differences in marital status moderation of geneticand environmental influences on subjective health' Behavior Genetics, vol 46, no. 1, pp. 114-123. DOI:10.1007/s10519-015-9758-y

Digital Object Identifier (DOI):10.1007/s10519-015-9758-y

Link:Link to publication record in Edinburgh Research Explorer

Document Version:Peer reviewed version

Published In:Behavior Genetics

General rightsCopyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s)and / or other copyright owners and it is a condition of accessing these publications that users recognise andabide by the legal requirements associated with these rights.

Take down policyThe University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorercontent complies with UK legislation. If you believe that the public display of this file breaches copyright pleasecontact [email protected] providing details, and we will remove access to the work immediately andinvestigate your claim.

Download date: 24. Jun. 2018

http://dx.doi.org/10.1007/s10519-015-9758-y

https://www.research.ed.ac.uk/portal/en/publications/gender-differences-in-marital-status-moderation-of-genetic-and-environmental-influences-on-subjective-health(513a2310-1874-4620-bc07-5f2eab3d0a00).html

GenderDifferencesinMaritalStatusModerationofGeneticandEnvironmentalInfluencesonSubjectiveHealth

DeborahFinkel1,CarolE.Franz2,BrianaHorwitz3,KaareChristensen4,MargaretGatz5,WendyJohnson6,JaakoKaprio7,8,TellervoKorhonen7,8,9,JenaeNiederheiser10,IngePetersen4,RichardJ.Rose11,andKarriSilventoinen7,fortheIGEMSconsortium121. IndianaUniversitySoutheast2. UniversityofCalifornia,SanDiego3. CaliforniaStateUniversity,Fullerton4. UniversityofSouthernDenmark5. UniversityofSouthernCalifornia6. UniversityofEdinburgh7. UniversityofHelsinki8. NationalInstituteforHealthandWelfare,Helsinki,Finland9. UniversityofEasternFinland,Kuopio,Finland10. PennsylvaniaStateUniversity11. IndianaUniversityBloomington12. MembersoftheconsortiumonInterplayofGenesandEnvironmentacrossMultiple

Studies(IGEMS)include:NancyL.Pedersen(DepartmentofMedicalEpidemiologyandBiostatistics,KarolinskaInstitutet,Stockholm,Sweden,andDepartmentofPsychology,UniversityofSouthernCalifornia,LosAngeles,CA),KaareChristensen(DepartmentofEpidemiology,UniversityofSouthernDenmark,Odense,Denmark),AnnaDahl(InstituteofGerontology,SchoolofHealthSciences,JönköpingUniversity,Jönköping,Sweden),DeborahFinkel(DepartmentofPsychology,IndianaUniversitySoutheast,NewAlbany,IN),CarolE.Franz(DepartmentofPsychiatry,UniversityofCalifornia,SanDiego,LaJolla,CA),MargaretGatz(DepartmentofMedicalEpidemiologyandBiostatistics,KarolinskaInstitutet,Stockholm,Sweden,andDepartmentofPsychology,UniversityofSouthernCalifornia,LosAngeles,CA),BrianaN.Horwitz(DepartmentofPsychology,CaliforniaStateUniversity,Fullerton,CA),BooJohansson(DepartmentofPsychology,UniversityofGothenburg,Gothenburg,Sweden),WendyJohnson(DepartmentofPsychologyandCentreforCognitiveAgeingandCognitiveEpidemiology,UniversityofEdinburgh,Edinburgh,UK),JaakkoKaprio,DepartmentofPublicHealth,UniversityofHelsinki,Helsinki,Finland),WilliamS.Kremen(CenterofExcellenceforStressandMentalHealth,VASanDiegoHealthcareCenter,LaJolla,CA,andDepartmentofPsychiatry,UniversityofCalifornia,SanDiego,LaJolla,CA),RobertKrueger(DepartmentofPsychology,UniversityofMinnesota,Minneapolis,MN),MichaelJ.Lyons(DepartmentofPsychologicalandBrainSciences,BostonUniversity,Boston,MA),MattMcGue(DepartmentofPsychology,UniversityofMinnesota,Minneapolis,MN),JenaeM.Neiderhiser(DepartmentofPsychology,ThePennsylvaniaStateUniversity,UniversityPark,PA),IngePetersen(DepartmentofEpidemiology,UniversityofSouthernDenmark,Odense,Denmark),andChandraA.Reynolds(DepartmentofPsychology,UniversityofCalifornia-Riverside,Riverside,CA).

Maritalstatusandsubjectivehealth 2

Abstract

FromtheIGEMSConsortium,datawereavailablefrom26,579individualsaged23to102

yearson3subjectivehealthitems:self-ratedhealth(SRH),healthcomparedtoothers

(COMP),andimpactofhealthonactivities(ACT).Maritalstatuswasamarkerof

environmentalresourcesthatmaymoderategeneticandenvironmentalinfluenceson

subjectivehealth.Resultsdifferedforthe3subjectivehealthitems,indicatingthattheydo

nottapthesameconstruct.Althoughtherewaslittleimpactofmaritalstatusonvariance

componentsforwomen,maritalstatuswasasignificantmodifierofvarianceinall3

subjectivehealthmeasuresformen.ForbothSRHandACT,singlemendemonstrated

greatersharedandnonsharedenvironmentalvariancethanmarriedmen.FortheCOMP

variable,geneticvariancewasgreaterforsinglemenvs.marriedmen.Resultssuggest

genderdifferencesintheroleofmarriageasasourceofresourcesthatareassociatedwith

subjectivehealth.

Keywords:subjectivehealth,maritalstatus,agedifferences,genderdifferences,GxE

interaction,moderationmodel


Subjectivehealthisthefocusofgreatresearchinterestbecauseoftheroleitplaysin

predictingobjectivehealthandmortality.Infact,measuresofsubjectivehealthpredict

mortalityaboveandbeyondobjectivehealthmeasures(Idler&Benyamini,1997;Latham&

Peek,2013;McFaddenetal.,2009).Manyresearchershavepositedexplanationsforthe

paradoxicalobservationthatputativelysimplequestionsabouthealthperceptionscan

provideinformationaboutobjectivehealth-relatedoutcomesdistinctfrommultiple

objectivemeasuresofhealth(Benyamini,2011).Arecentanalysistestedfour

conceptualizationsofsubjectivehealth(Franzetal.,inrevision);testsofageandgender

moderationofgeneticandenvironmentalvarianceinsubjectivehealthmeasures

supportedtheideathatsubjectivehealthtapspersonalintuitionsabouthealthandthat

thesepersonalintuitionsreflectculturaldefinitionsandpersonalconceptsofhealth(Bailis,

Segall,&Chipperfield,2003;Jylhä,2009,2010).Theseconceptionsofsubjectivehealthrely

primarilyonmechanismswithintheindividual–intuitionsandperceptionsabouthealth.

Inthecurrentanalysis,weshiftedthefocustoanexternalmechanism,marriage,which

mayinfluencesubjectivehealthandthusmaymoderatethegeneticandenvironmental

contributionstosubjectivehealth.

Researchhasdemonstratedthatsubjectivehealthisrelatedtoexternalfactorssuch

aseducation,financialstatus,socialsupport,maritalstatus,andneighborhood

characteristicsthatindicatetheextentofresourcesindividualshavetosupportand

maintaintheirhealth(Benyamini,2011;Subramanian,Kubzansky,Berkman,Fay,&

Kawachi,2006).Althoughmarriagehasmanymeanings,atitsmostbasiclevelmarital

statuscanreflectsocioeconomicstatusaswellassocialandphysicalsupport(Benyamini,

2011;Zheng&Thomas,2013).Thebeneficialassociationbetweenmarriageandphysical


healthhasbeenamplydemonstrated(Carr&Springer,2010;Robles&Kiecolt-Glaser,

2003)andarecentmeta-analysissupportedthelowerrelativeriskformortalityamong

marriedpeoplecomparedwithnon-marriedgroups(Manzolo,Villari,Pirone,&Boccia,

2007).Researchsuggeststhatmarriagesupportsmaintenanceofhealthbehaviors,thus

affectingdiseasepreventionratherthantreatmentorrecoveryfromsevereillnesses

(Zheng&Thomas,2013).Forexample,inasampleoftwinsdiscordantformaritalstatus,

theunmarriedtwinwasmorelikelytosmokeandlesslikelytoexercise(Osler,McGue,

Lund,&Christensen,2008).Epidemiologicalstudiesofthisnaturecannotdeterminecause

andeffect,however;thusthereisongoingdiscussionaboutwhethertheassociation

betweenmarriageandhealthreflectsselectionorcausation(Silventoinen,Moustgaarid,

Peltonen,&Martikainen,2013).

Theassociationofmaritalstatuswithsubjectivehealthisnearlyaswellestablished

astheassociationwithphysicalhealth(Liu&Umberson,2008;Waite&Gallagher,2000).

Evidencesuggeststhattherelationshipbetweenmaritalstatusandsubjectivehealth

reflectsatendencyformarriedadultstobesomewhatoverconfidentabouttheirhealth

status.Infact,ZhengandThomas(2013)concludethatadultsperceivemarriageasa

sourceofresourcestosupporthealth,whichresultsinbothoverestimationofhealthand

delayinseekingmedicalcare.Historicaltrendsindicatethatasgenderrolesandthe

meaningofmarriagehavechangedoverthelastseveraldecades,therelationshipbetween

maritalstatusandsubjectivehealthhasalsochanged(Liu&Umberson,2008).

Theassociationbetweenmaritalstatusandphysicalandsubjectivehealthmaynot

bethesameformenandwomen(Liu&Umberson,2008).Researchsuggeststhatmenand

womenhavediverseexperiencesofphysicalaging.Mentendtohaveearlierandmore


compressedhistoriesofmajorillnessesanddisabilitypriortodeath,whilewomenlive

longer,havemorehealthcomplaintsacrossthelifecourse,andhigherprevalenceof

chronicdisablingbutnotfataldiseaseslaterinlife(Sainioetal.,2006).Asaresult,men

mayfocusmoreonlife-threateningconditionswhenjudgingtheirownhealth,whereas

womenmayfocusonchronicconditionsthatareagreaterpartoftheirexperienceofaging

(Deeg&Kriegsman,2003).Consistentwiththis,womentendtoreportpoorersubjective

health,andsubjectivehealthappearstobeaweakerpredictorofmortalityinwomenthan

inmen(Benyamini,2011;Benyamini,Blumstein,Lusky,&Modan,2003;Deeg&Kriegsman,

2003).Evidenceforagenderdifferenceintheassociationbetweenmaritalstatusand

subjectivehealthismixed,withsomeresearchersfindingastrongerprotectiveeffectof

marriageformenthanwomen(Liu&Umberson,2008;Williams&Umberson,2004),while

othersreportnogenderdifferences(Zheng&Thomas,2013).

Whereaspreviousstudiesfocusedprimarilyongenderdifferencesinmeansandthe

predictivepowerofsubjectivehealthmeasures,weexaminedhowgeneticand

environmentalcomponentsofvarianceinsubjectivehealtharemoderatedbymarital

status,andwhetherthatmoderationeffectdiffersformenandwomen.Multiplestudies

havereportedheritabilityestimatesforsubjectivehealth;however,toourknowledge,no

otherstudyhasexaminedmaritalstatusmoderationoftheseestimates.Studiesofadult

twinsinAustralia,Denmark,Finland,Sweden,andtheU.S.havereportedheritability

estimatesforsubjectivehealthprimarilyintherangeof25%to30%(forareviewsee

(Franzetal.,inrevision).Arecenttwinanalysisthatincluded12,900individualsaged25

to102fromtheInterplayofGenesandEnvironmentacrossMultipleStudiesconsortium

(IGEMS;(Pedersenetal.,2013),whichisalsothebasisforthepresentstudy,provideda


morenuancedunderstandingofgeneticandenvironmentalinfluencesonsubjectivehealth.

Resultsindicatedthatheritabilityvariedsignificantlybyage,gender,andsubjectivehealth

measure.Here,weexpandedthoseanalysestoexaminehowage,sex,andmaritalstatus

moderatedgeneticandenvironmentalinfluencesonsubjectivehealth.Althoughmarital

statusisnotpurelyanenvironmentalmeasure(Trumbetta,Markowitz,&Gottesman,

2007),wefocusedonrelationshipstatusasamarkerforresourcestosupporthealthby

differentiatingindividualswhowerelivingwithpartners(marriedorcohabitating)from

thoselivingalone(single,divorced,orwidowed).Wepredictthatlivingwithapartner

providesaprotectiveorstabilizinginfluencethattosomedegreebuffersindividuals

againstagedifferencesingeneticandenvironmentalinfluencesonsubjectivehealth

identifiedbyFranzandcolleagues(inrevision).Furthermore,giventhepossibilityof

genderdifferencesintheroleofmaritalstatusinsubjectivehealth,wepredictthatliving

withapartnerwillmodulatetheheritabilityofsubjectivehealthdifferentlyformenand

women.Finally,basedonpreviousresults,wealsopredictthatthemoderationeffectof

marriagewillvaryacrossdifferentmeasuresofsubjectivehealth.

METHOD

Participants

IGEMSisaninternationalconsortiumoftwinstudiesfromtheNordiccountriesand

theU.S.coveringtheadultlifespan(Pedersenetal.,2013).Thesamplesizesandage

rangesfromtheIGEMSstudiesincludedherearepresentedinTable1:atotalof26,579

individualscontributedrelevantdatatothecurrentstudy.Agerangedfrom23-102years,

withameanof55.2(sd=16.6).Forreportingofsamplesizesandmeans,thesamplewas

dividedintofourapproximatelyequalagegroups:agelessthan50,50-59,60-69,and


greaterthan70years.Forthemoderatoranalyses,bothmembersofatwinpairwere

needed:thesame-sextwinpairsavailableforeachsubjectivehealthmeasureineachage

grouparepresentedinTable2.Althoughsamplesizeispresentedseparatelybyagegroup

toindicatecoverageacrossthelifespan,agewasincludedasacontinuousmoderatorinthe

biometricmodels.

Measures

MaritalStatus.MaritalstatuswasrecordedinvariouscategoriesintheIGEMS

studies.Becausethefocusofthecurrentstudywasonpartnerpresenceasamarkerof

resourcestosupporthealth,wecreatedadichotomousvariable,combiningmarriedand

cohabitatinginonecategory,andwidowed,divorced,andsingleintheothercategory.For

simplicity,thetwocategorieswerelabeled“married”and“single”.Thedistributionsof

maritalstatusacrossthefouragegroupsformenandwomenarepresentedinFigure1.

Percentsingleincreasedmodestlybutsignificantlyfrom8%to12%formenacrossthe

fouragegroups(χ2(df=3,N=12201)=202.2,p<.01).Itincreasedmoredramaticallyfrom

9%to41%forwomen,followingpopulationtrends(χ2(df=3,N=14378)=2046.9,p<.01).

SubjectiveHealth.Threedifferenttypesofquestionswereusedtoassesssubjective

healthintheIGEMSstudies(seeTable1).Nineofthestudiesincludedthemostcommon

questionusedtoassesssubjectivehealth:“Howwouldyourateyouroverallhealth?”Inthe

literature,theacronymSRHistypicallyusedtoidentifythisquestion.EightIGEMSstudies

askedparticipantstocomparetheirhealthwithothers(COMP)usingtwoslightlydifferent

forms:“comparedtoothersyourage,howwouldyourateyouroverallhealth?”wasused

bysixstudiesand“IamashealthyasanyoneIknow”fromtheSF-36version1(Ware,

Kosinski,&Keller,1994)usedbytwo.Participantsineightstudiesalsoindicatedhowtheir


healthaffectedtheirdailyactivities(ACT);fivestudiesincludedasinglequestion,“Isyour

healthconditionpreventingyoufromdoingthingsyouliketodo?”Threeindicated

whethertheirhealthaffectedtheirphysicalfunctioninginalistofmultiplebehaviorsfrom

theSF-36.ResponsestoactivitieswereaveragedtocreateasingleACTscoreforthese

threestudies.

Althoughthesubjectivehealthquestionsadministeredacrossthestudieswere

similaroridentical,theresponsescalesvariedfromdichotomousoptionsto7-pointLikert

scales.Toexamineandreconciledifferencesamongtheseputativelysimilarmeasures,we

engagedinaharmonizationprocess,collectingnewdataonallcombinationsofquestions

andanswerschemesfromanindependentinternationalsampleof1065participantsaged

30to98(Gatzetal.,inpress).Theharmonizationsampleallowedustoverifythatsimilarly

wordedquestionscorrelatedsubstantially,regardlessofexactwordingorresponsescales.

Averagecorrelationsacrossresponsescaleswere.77forSRH,.78forACT,and.63for

COMP.Averagecorrelationsacrossthethreedifferentsubjectivehealthquestionswere

rSRH·COMP=.63,rSRH·ACT=.57,andrCOMP·ACT=.46.Comparisonofthreetypesofharmonization

methodsindicatedthattheoptimalapproachinvolvedstandardizingscoreswithinsamples

toachieveacommonmetric,thenpoolingdataacrossstudies.Tothatend,thethree

subjectivehealthquestionswerestandardizedseparatelywithineachsampleand

convertedtoT-scores(mean50,SD10).Forallmeasures,highscoresindicatedbetter

subjectivehealth.MeansacrossagegroupsinthecombinedIGEMSsamplearepresentedin

Figure2,indicatingage,gender,andmaritalstatuseffects.Differenttrendsareevidentfor

eachsubjectivehealthitem,withthesmallestgroupmeandifferencesseenfortheCOMP


variable.Continuousagetrendsinmeansubjectivehealthestimatedbytheagemoderation

modelresultedinthesamepatternofresults.

StatisticalMethods

Toevaluatewhetherthegeneticandenvironmentalinfluencesonsubjectivehealth

ratingsdifferedasafunctionofmaritalstatus,weutilizedamodifiedversionofthe

univariatetwinmodelinwhichageandmaritalstatuswereincludedasmoderating

variables(Purcell,2002;VanderSluis,Dolan,Neale,&Posthuma,2008).Thestandard

univariatetwinmodelincorporatesmonozygotic(MZ)twinsanddizygotic(DZ)twinsto

decomposethevarianceofanyphenotypeintotheproportionattributedtoadditive

geneticinfluences(A),commonorsharedenvironmentalinfluences(C),andunique

environmentalinfluences(E).Themodelusedinthepresentstudyallowsfordifferences

intheA,C,andEparametersasafunctionoftwocontinuousmoderatorvariables(ageand

age-squared)andonecategoricalmoderatorvariable(maritalstatus).Moderationofthe

geneticand/orenvironmentalvariancecomponentsindicatesthatthecontributionsof

thesefactorstothevarianceofsubjectivehealthvarybyageandmaritalstatus.Allmodels

weretestedusingthestructuralequation-modelingpackageClassicMx1.68(Neale,Boker,

Xie,&Maes,2003).Evaluationofrelativemodelfitwasperformedusingthelikelihood-

ratio-test(LRT).SignificantLRTvaluesindicatethatthereductioninparametersresulted

inasignificantreductioninmodelfit.

RESULTS

Previousanalyseshavefocusedonageandsexmoderationofsubjectivehealth

(Franzetal.,inrevision);thefocusherewasprimarilyonmaritalstatusmoderationof

subjectivehealthformenandwomen.Therefore,modelcomparisonfocusedontesting


maritalstatusmoderationparameters.Thefirstphaseofmodeltestingexaminedgender

differencesintheseparametersandtwomodelswerecompared:oneinwhichall19model

parameterswereallowedtovaryacrossgenderversusamodelinwhichthe3marital

statusmoderationparameters(forA,C,andE)weresetequalacrossgenders.Comparison

ofthesetwomodelsindicatedsignificantgenderdifferencesinmaritalstatusmoderation

ofsubjectivehealthforSRH(LRT=8.13,df=3,p<.05)andforACT(LRT=24.64,df=3,p

<.01),butnotforCOMP(LRT=2.01,df=3,ns).

Inthenextphaseofmodelfitting,fivemodelsweretestedseparatelyforeach

gender(seeTable3).Firstthefullmodelestimatedallvariancecomponentsand

moderatorparameters.Inmodel2,allmaritalstatusmoderationparametersweredropped.

Inmodels3through5,maritalstatusmoderationofeachvariancecomponent(A,C,andE)

wastestedindependently.Foreachmodel,allotherparameterswereretained:theprimary

A,C,andEvariancecomponentsandtheageandage-squaredmoderationofthese

components.

Comparingmodel2tomodel1indicatedsignificantmaritalstatusmoderationof

onlyonesubjectivehealthmeasureinwomen:SRH.Testingeachmaritalstatusmoderation

parameterseparately(models3through5)failedtoidentifythesourceofthemarital

statusmoderationofvariancecomponentsofSRHinwomen.MinimizationofAkaike’s

InformationCriterion(log-likelihood–2*degreesoffreedom)canbeusedtoidentifythe

best-fittingmodel.Inthiscase,AICwassmallestformodel3forSRHinwomen,suggesting

modestmaritalstatusmoderationofCandEcomponents.TheestimatesfromtheACE

modelwithfullmoderation(model1)wereusedtodepictmaritalstatusmoderationof

subjectivehealthmeasuresforwomenacrossage(seeFigure3).A,C,andEcomponentsof


variance,alongwithtotalvariance,arepresentedforsingleandmarriedwomenforthe

threesubjectivehealthmeasures.ForSRH,slightlymoreCandEvariancewasevidentfor

singlewomenthanformarriedwomen,resultingingreatertotalvarianceinSRHforsingle

womenthanmarriedwomen.TheestimatesfortheAvariancecomponentfromsingleand

marriedwomenwerenearlyidentical,sothelinesonthegraphoverlap.ResultsforACT

suggestsomewhatmoreAvarianceforsinglewomenthanmarriedwomen;however,the

moderationparameterdidnotachievesignificance(model3vs.model1=3.20,df=1,n.s.).

LittledistinctioncanbedetectedintheA,C,andEvariancecomponentsforCOMPin

women;thelinesforAandEvariancecomponentsoverlap.Thegeneralpatternofvariance

componentsacrossagematchestheresultsreportedbyFranzandcolleagues(inrevision).

HeritabilityofSRHwasestimatedat28%forwomenacrossmostoftheagerange,witha

somewhatlowerheritabilityestimatedinlateadulthood(17%).HeritabilityforACT

showedacurvilineartrendoverage,withhighestestimatesforyoungerwomen(17%and

25%)andlowerestimatesforolderwomen(2%).HeritabilityforCOMPincreasedacross

agefrom7%foryoungerwomento19%forolderwomen.

Incontrast,evidenceforsignificantmaritalstatusmoderationofvariancewasfound

forallthreesubjectivehealthmeasuresinmen,althoughthepatternofresultsdiffers

acrossmeasures.ForbothSRHandACT,modelfittingindicatedsignificantmaritalstatus

moderationofCandEcomponentsofvariance;whereasforCOMP,modelcomparison

indicatedsignificantmaritalstatusmoderationoftheAvariancecomponent.Theimpactof

maritalstatusonvariancecomponentsofthethreesubjectivehealthmeasuresinmenis

presentedinFigure4.ForSRH,singlemendemonstratesignificantlymoreCandEvariance

thanmarriedmen,resultinginmoretotalvariance.ThedifferenceinCvariancebetween


singleandmarriedmendeclineswithage,asdoestotalCvariance,whichreachesnearly

zeroforbothgroupsofmenatage75.GeneticvarianceinSRHwasbasicallyidenticalfor

singleandmarriedmen,sothelinesonthegraphoverlap.Heritabilityislowerinyounger

menandthehighestestimateisatage70(23%).

FortheACTvariable,A,C,andEcomponentsofvariancearehigherforsinglemen

thenformarriedmen,butonlythedifferencesinCandEcomponentsachievesignificance

(seeFigure4).WhereasthedifferencesinAandEvariancesareconstantacrosstheage

range,maritalstatusdifferencesinCvariancepeakinmidlife,aroundage55;group

differencesareminimizedearlierandlaterinthemeasuredagerange(age30and75).

Acrossmostoftheagerange,totalvarianceisabout30%higherforsinglemencompared

withmarriedmen.SimilartoSRHinmen,heritabilityforACTishighestatage65(26%).

Amarkedlydifferentpatternofmaritalstatusmoderationofvariancewasfoundfor

theCOMPvariable.Inthisinstance,maritalstatussignificantlymoderatedAvariance,only

(seeFigure4).Avarianceishighestinmidlifeforbothsingleandmarriedmen,butgenetic

varianceestimatedforsinglemenisnearlythreetimeshigherthenthegeneticvariance

estimatedformarriedmen.Asaresult,heritabilityestimatesinmidlifefortheCOMP

variableare24%versus10%forsingleandmarriedmen,respectively.Note,however,that

regardlessofthedifferentmoderationpatternssuggestedforCOMPformenandwomen,

modelcomparisonsindicatedthatthepatternofmaritalstatusmoderationofCOMPdid

notdiffersignificantlybetweenmenandwomen.Forallthreesubjectivehealthmeasures,

totalvariancewasgreaterforsinglementhanmarriedmen.

DISCUSSION


Ourexaminationofmaritalstatusmoderationofgeneticandenvironmental

influencesonsubjectivehealthacrossadulthoodrevealedvariedpatternsofmoderation

thatdifferedformenandwomenandforthethreedistinctmeasuresofsubjectivehealth.

Inaddition,wereplicatedthepatternofagemoderationofgeneticandenvironmental

influencesonsubjectivehealthreportedbyFranzandcolleagues(inrevision)inasmaller

sampleof12,900individualsfromtheIGEMSconsortium.

Formen,maritalstatusmoderatedsharedandnonsharedenvironmental

componentsofvarianceforSRHandACTandthegeneticcomponentofvarianceforCOMP.

Asaresult,sharedandnonsharedenvironmentalcomponentsofvariancewere

significantlygreaterforsinglementhanformarriedmenfortheSRHandACTmeasures.

Whereasthedifferencesinnonsharedenvironmentalvarianceweregenerallyconsistent

acrosstheagerange,differencesinsharedenvironmentalvariancewerehigherfor

youngermenthanforoldermen.Infact,forACT,estimatesofsharedenvironmentwere

nearzeroformarriedmenbutpeakedat19%atage50forsinglemen.Thus,formen,

marriageapparentlyprovidedabufferthatresultedinmorestabilityincomponentsof

varianceforsubjectivehealthacrosstheagerange.Themaritalstatuscategorythat

differedthemostacrosstheagegroupsformenwaswidowhood:10.2%ofsinglemenin

the50s,23.1%ofsinglemenintheir60s,and57.6%ofsinglemenover70werewidowed.

Parallelingthispattern,nonsharedenvironmentalvarianceofsubjectivehealthwasgreater

insinglemenacrossthesameagerange.Thepercentageofsinglemenreportingthatthey

weredivorcedpeakedinthe50-60agerange,approximatelythesamepointintheage

rangethatsharedenvironmentalvariancepeakedforSRHandACTforsinglemenand

geneticvariancepeakedforCOMPinsinglemen.Itwouldappear,then,thatwithout


marriageasaprotectivefactor,thefluctuationsingeneticandenvironmentalcomponents

ofvarianceareamplified.

Thepatternofresultsforwomenwasfarlesscomplex:maritalstatusplayedonlya

modestroleingreaterenvironmentalvarianceinsinglewomenforSRH.Thus,marriage

providedatmostalimitedbufferagainsttheenvironmentalimpactofhealthperceptions

withage.However,acrossthelifecourse—butespeciallyafterage50—womenareless

likelytoremarryafterdivorceorbereavement(Waite,Laumann,Das,&Schumm,2009);

thusmaritalstatusduringthistimeperiodmaybemorestableforwomenthanformen,

resultinginlessmaritalstatusmoderationofvariationinperceptionsofhealthcompared

withmen.Previousevidenceforgenderdifferencesintheinfluenceofmaritalstatuson

meansubjectivehealthhasbeenmixed(Liu&Umberson,2008;Williams&Umberson,

2004;Zheng&Thomas,2013).Inthecurrentanalyses,despitegenderdifferencesin(a)the

experiencesofphysicalaging(Sainioetal.,2006),(b)theimpactofmaritalstatuson

environmentalresourcesinthesecohorts(Weaver,2010),and(c)theincidenceof

bereavement,maritalstatushadverylittleimpactonsourcesofvarianceinsubjective

healthinwomen.Itmaybethatwomeninthesecohortsarebetterablethanmento

maintainsourcesofsocialsupportindependentofmaritalstatus,withtheresultthat

variancecompositionofsubjectivehealthisfairlyconsistentforsingleandpartnered

women.Someevidencesuggeststhatmenhavesmallersupportnetworksthanwomen

andthusmarriageconstitutesalargerportionofmen’ssocialsupportnetworks(Dykstra&

Fokkema,2007).Moreover,especiallyintheseoldercohorts,womenareoftenresponsible

formaintainingandfosteringthesocialinteractionsofbothmembersofthepair(Dykstra

&deJongGierveld,1994;Rosenthal,1985).Thusbereavementformencanmeanlossof


emotionalandinstrumentalsupportformaintaininghealth(Chipperfield&Havens,2001).

Consequently,althoughbothmenandwomenexperienceincreasedmortalityrates

immediatelyfollowingbereavement,mortalityratestendtoremainelevatedformen,only

(Kaprio,Koskenvu,&Hell,1987).Regardlessofmaritalstatus,then,womenaremorelikely

thanmentobeabletotaptheirlargersupportnetworksfortheemotionaland

instrumentalresourcesthatresultinstableheritabilityestimatesforsubjectivehealth.

Finally,therewerestrikingdifferencesingeneticandenvironmentalcomponentsof

variance,andthemaritalstatusmoderationofvariance,acrossmeasuresofsubjective

health.Infact,genderdifferencesinmaritalstatusmoderationoftheCOMPvariabledid

notachievesignificance.Differentsubjectivehealthitemstapdifferentframesofreference

(Manderbacka,Kåreholt,Martikainen,&Lundberg,2003;Vuorisalmi,Lintonen,&Jylhä,

2006),reflectingdiversecombinationsofpsychologicaldispositions,situationalfactors,

sharedculturalvalues,andcharacteristicssuchasage,gender,class,orethnicity(Jylhä,

2009,2010;Sprangers&Schwartz,1999).Somesubjectivehealthquestionstriggermore

internalframesofreference(e.g.,rateyouroverallhealth),whereasinotherstheframeof

referencemaybemoreexternal(e.g.,rateyourhealthcomparedtoothersyourage;does

healthpreventyoufromdoingthingsyouliketodo?)andmaytriggermoreconsciousor

unconsciousconsiderationofenvironmentalsupportfactors.Thecurrentresultssuggest

thatthedifferentframesofreferencetriggeredbythesubjectivehealthitemswere

differentiallyaffectedbymaritalstatus(atleastformen).

Ourconclusionsaresubjecttomethodologicallimitations.First,combiningdata

acrossstudieswasbothastrengthandaweaknessofourapproach.Combiningstudies

providedsufficientpowertoexamineeffectssimultaneouslyacrossagegroups,gender,


andmaritalstatus,whichisimpossiblewithsmallercohorts.However,italsonecessitated

harmonizingsomewhatdifferentmeasuresofsubjectivehealth.Theindependent

crosswalkstudyofourmeasures(Gatzetal.,inpress)supportedourapproach.Moreover,

consistentwithourdata,across-nationalcomparisonofself-ratedhealthfoundthat

relationshipsamongSRHandcovariates,includingmaritalstatusandgender,were

homogeneousacrosscountries(Bardageetal.,2005).Second,weinterpretedmarital

statusasameasureofenvironmentalresourcestosupportphysicalandsubjectivehealth.

Althoughalternativeinterpretationsoftherelationshipbetweenmaritalstatusand

subjectivehealthexist,maritalstatusasamarkerforhealthresourceshasbeensupported

bytheliterature(Benyamini,2011;Zheng&Thomas,2013).Still,maritalstatusdoesnot

taponlyenvironmentalvariance,butgeneticvarianceaswell.However,theheritable

influencesonmaritalstatusappeartodeclinefrom40%inearlyadulthoodto0%byage50

andbeyond(Trumbettaetal.,2007).Third,theparticipantswereallfromtheU.S.andthe

Nordiccountries,andinfacttheFinnishtwinstudiescontributednearlyhalftheavailable

dataforSRH.ResultsforACT,whichwasnotincludedintheFinnishdata,aresimilartothe

resultsforSRH.Moreover,previousreportsoftheseanalysesthatdidnotincludethe

Finnishtwinstudiesproducedverysimilarresults(Finkel,Horwitz,&Gatz,2014):addition

oftheFinnishtwinsampleprovidedmorepowerbutdidnotchangetheoverall

conclusions.Finally,asinanystudyincludingolderadults,thesamplewassubjectto

survivoreffects,particularintheoldestagegroups.Theslightreductionsintotalvariance

generallyevidentinlateadulthood(particularlyformen)likelyresultedfromabsenceof

individualsinpooresthealthfromthesample.


Overall,weobservedthatexternalfactors,suchasthosetappedbymaritalstatus,

wereassociatedwithgeneticandenvironmentalcontributionstosubjectivehealth,

indicativeofgenebyenvironmentinteraction.Genderandagedifferences,combinedwith

maritalstatusdifferencesthatmayimpactaccesstoresourcestosupporthealth,were

associatedwithfluctuationsingeneticandenvironmentalcomponentsofvariancein

healthperceptions.Thisrelationshipwasfarmorepronouncedformenthanforwomen,

likelyasaresultofdifferentrolesthatmarriageandpartnersplayinsocialnetworksfor

menandwomeninthesecohorts.Finally,theseresultsjoinagrowingbodyofevidence

thatnotallmeasuresofsubjectivehealthareequal.Themannerinwhichthequestionis

posedtriggersaframeofreferencethatwillimpacttheinterplayofgeneticand

environmentalinfluencesonhealthperceptions.


AuthorNote/Acknowledgements

IGEMSissupportedbytheNationalInstitutesofHealthgrantno.R01AG037985.SATSAwassupportedbygrantsR01AG04563,R01AG10175,theMacArthurFoundationResearchNetworkonSuccessfulAging,theSwedishCouncilForWorkingLifeandSocialResearch(FAS)(97:0147:1B,2009-0795)andSwedishResearchCouncil(825-2007-7460,825-2009-6141).OCTO-TwinwassupportedbygrantR01AG08861.TOSSwassupportedbygrantR01MH54610fromtheNationalInstituteofHealth.TheDanishTwinRegistryissupportedbygrantsfromTheNationalProgramforResearchInfrastructure2007fromtheDanishAgencyforScienceandInnovation,theVeluxFoundationandtheUSNationalInstituteofHealth(P01AG08761).TheMinnesotaTwinStudyofAdultDevelopmentandAgingwassupportedbyNIAgrantR01AG06886.VETSAwassupportedbyNationalInstituteofHealthgrantsNIAR01AG018384,R01AG018386,R01AG022381,andR01AG022982,and,inpart,withresourcesoftheVASanDiegoCenterofExcellenceforStressandMentalHealth.TheCooperativeStudiesProgramoftheOfficeofResearch&DevelopmentoftheUnitedStatesDepartmentofVeteransAffairshasprovidedfinancialsupportforthedevelopmentandmaintenanceoftheVietnamEraTwinRegistry.MIDUStwinstudywassupportedbytheJohnD.andCatherineT.MacArthurFoundationResearchNetworkonSuccessfulMidlifeDevelopmentandbyNationalInstituteonAgingGrantAG20166.DatacollectionandanalysesintheFinnishtwincohorthavebeensupportedbyENGAGE–EuropeanNetworkforGeneticandGenomicEpidemiology,FP7-HEALTH-F4-2007,grantagreementnumber201413,NationalInstituteofAlcoholAbuseandAlcoholism(grantsAA-12502,AA-00145,andAA-09203toRJRose),andtheAcademyofFinland(grants100499,205585,118555,141054,265240,263278and264146toJ.Kaprio).ThecontentofthismanuscriptissolelytheresponsibilityoftheauthorsanddoesnotnecessarilyrepresenttheofficialviewsoftheNIA/NIH,ortheVA.


References

Bailis,D.S.,Segall,A.,&Chipperfield,J.G.(2003).Twoviewofself-ratedgeneralhealthstatus.SocialScience&Medicine,56,203-217.

Bardage,C.,Pluijm,S.M.F.,Pedersen,N.L.,Deeg,D.J.H.,Jylhä,M.,Noale,M.,...Otero,Á.(2005).Self-ratedhealthamongolderadults:Across-nationalcomparison.EuropeanJournalofAgeing,,2,149-158.

Benyamini,Y.(2011).Whydoesself-ratedhealthpredictmortality?Anupdateoncurrentknowledgeandaresearchagendaforpsychologists.PsychologyandHealth,26,1407-1413.

Benyamini,Y.,Blumstein,T.,Lusky,A.,&Modan,B.(2003).Genderdifferencesintheself-ratedhealth--mortalityassociation:Isitpoorself-ratedhealththatpredictsmortalityorexcellentself-ratedhealththatpredictssurvival?TheGerontologist,43,396-405.

Carr,D.,&Springer,K.W.(2010).Advancesinfamiliesandhealthresearchinthe21stcentury.JournalofMarriageandFamily,72,743-761.

Chipperfield,J.G.,&Havens,B.(2001).Genderdifferencesintherelationshipbetweenmaritalstatustransitionsandlifesatisfactioninlaterlife.JournalofGerontology:PsychologicalSciences,56B,P176–P186.

Christensen,K.,Holm,N.V.,McGue,M.,Corder,L.,&Vaupel,J.W.(1999).ADanishpopulation-basedtwinstudyongeneralhealthintheelderly.JournalofAgingandHealth,11,49-64.

Deeg,D.J.H.,&Kriegsman,D.M.W.(2003).Conceptsofself-ratedhealth:Specifyingthegenderdifferenceinmortalityrisk.TheGerontologist,43,376-386.

Dykstra,P.A.,&deJongGierveld,J.(1994).Thetheoryofmentalincongruity,withaspecificapplicationtolonelinessamongwidowedmenandwomen.InR.Erber&R.Gilmour(Eds.),Theoreticalframeworksinpersonalrelationships(pp.235–259).Hillsdale,NJ:ErlbaumAs-sociates,Inc.

Dykstra,P.A.,&Fokkema,T.(2007).Socialandemotionallonelinessamongdivorcedandmarriedmenandwomen:Comparingdeficitandcognitiveperspectives.BasicandAppliedSocialPsychology,29,1-12.

Finkel,D.,Horwitz,B.N.,&Gatz,M.(2014).Maritalstatusmoderatesgenderdifferencesingeneticandenvironmentalinfluencesonsubjectivehealth[Abstract].TheGerontologist,54.

Finkel,D.,&McGue,M.(1993).Theoriginsofindividualdifferencesinmemoryamongtheelderly:Abehaviorgeneticanalysis.PsychologyandAging,8,527-537.

Finkel,D.,&Pedersen,N.L.(2004).Processingspeedandlongitudinaltrajectoriesofchangeforcognitiveabilities:TheSwedishAdoption/TwinStudyofAging.Aging,Neuropsychology,andCognition,11,325-345.

Franz,C.E.,Finkel,D.,Panizzon,M.S.,Spoon,K.,Christensen,K.,Gatz,M.,...Pedersen,N.L.(inrevision).GeneticandEnvironmentalInfluencesonTheManyFacetsofSubjectiveHealthfromEarlyAdulthoodtoOldAge.JournalofAgingandHealth.

Gatz,M.,Reynolds,C.A.,Finkel,D.,Hahn,C.,Zhou,Y.,&Zavala,C.(inpress).Dataharmonizationinagingresearch:Notsofast.ExperimentalAgingResearch.

Idler,E.L.,&Benyamini,Y.(1997).Self-ratedhealthandmortality:areviewoftwenty-sevencommunitystudies.JHealthSocBehav,38,21-37.


Jylhä,M.(2009).Whatisself-ratedhealthandwhydoesitpredictmortality?Towardsaunifiedconceptualmodel.SocialScience&Medicine,69,307-316.

Jylhä,M.(2010).Self-ratedhealthbetweenpsychologyandbiology.AresponsetoHuismanandDeeg.SocialScience&Medicine,70,655-657.

Kaprio,J.,Koskenvu,M.,&Hell,R.(1987).Mortalityafterbereavement:Aprospectivestudyof95,647widowedpersons.AmericanJournalofPublicHealth,77,283-287.

Kaprio,J.,&Koskenvuo,M.(2002).Geneticandenvironmentalfactorsincomplexdiseases:theolderFinnishTwinCohort.TwinResearch,5,358-365.

Kaprio,J.,Pulkkinen,L.,&Rose,R.J.(2002).Geneticandenvironmentalfactorsinhealth-relatedbehaviors:studiesonFinnishtwinsandtwinfamilies.TwinResearch,5,366-371.

Kremen,W.S.,Thompson-Brenner,H.,Leung,Y.J.,Grant,M.D.,Franz,C.E.,Eisen,S.A.,&Lyons,M.J.(2006).Genes,environment,andtime:TheVietnamEraTwinStudyofAging(VETSA).TwinResearchandHumanGenetics,9,1009-1022.

Latham,K.,&Peek,C.W.(2013).Self-ratedhealthandmorbidityonsetamonglatemidlifeU.S.adults.JournalsofGerontologySeriesB:PsychologicalSciences&SocialSciences,68,107-116.

Liu,H.,&Umberson,D.(2008).Thetimestheyareachangin':Maritalstatusandhealthdifferentialsfrom1972to2003.JournalofHealthandSocialBehavior,49,239-253.

Manderbacka,K.,Kåreholt,I.,Martikainen,P.,&Lundberg,O.(2003).Theeffectofpointofreferenceontheassociationbetweenself-ratedhealthandmortality.SocialScience&Medicine,56,1447-1452.

Manzolo,L.,Villari,P.,Pirone,G.M.,&Boccia,A.(2007).Maritalstatusandmoralitintheelderly:Asystematicreviewandmeta-analysis.SocialScience&Medicine,6,77-94.

McClearn,G.E.,Johansson,B.,Berg,S.,Pedersen,N.L.,Ahern,F.,Petrill,S.A.,&Plomin,R.(1997).Substantialgeneticinfluenceoncognitiveabilitiesintwins80ormoreyearsold.Science,276,1560-1563.

McFadden,E.,Luben,R.,Bingham,S.,Wareham,N.,Kinmonth,A.-L.,&Khaw,K.-T.(2009).Doestheassociationbetweenself-ratedhealthandmortalityvarybysocialclass?SocialScience&Medicine,68,275-280.

Neale,M.C.,Boker,S.M.,Xie,G.,&Maes,H.H.(2003).Mx:StatisticalModeling.VCUBox900126,Richmond,VA23298:DepartmentofPsychiatry.

Neiderheiser,J.M.,&Lichtenstein,P.(2008).TheTwinandOffspringStudyinSweden:Advancingourunderstandingofgenotype-environmentinterplaybystudyingtwinsandtheirfamilies.ActaPsychologicaSinica,40,1116-1123.

Osler,M.,McGue,M.,Lund,R.,&Christensen,K.(2008).Maritalstatusandtwins'healthandbehavior:ananalysisofmiddle-agedDanishtwins.PsychosomMed,70,482-487.

Pedersen,N.L.,Christensen,K.,Dahl,A.,Finkel,D.,Franz,C.E.,Gatz,M.,...Reynolds,C.A.(2013).IGEMS:TheConsortiumonInterplayofGenesandEnvironmentacrossMultipleStudies.TwinResearchandHumanGenetics,16,481-489.

Purcell,S.(2002).Variancecomponentsmodelsforgene-environmentinteractionintwinanalysis.TwinResearchandHumanGenetics,5,554-571.

Robles,T.F.,&Kiecolt-Glaser,J.K.(2003).Thephysiologyofmarriage:Pathwaystohealth.Physiol.Behav,79,409-416.

Rosenthal,C.(1985).Kinkeepinginthefamilialdivisionoflabor.JournalofMarriageandtheFamily,47,965-974.


Sainio,P.,Koskinen,S.,Heliovaara,M.,Martelin,T.,Harkanen,T.,Hurri,H.,&Aromaa,A.(2006).Self-reportedandtest-basedmobilitylimitationsinarepresentativesampleofFinnsaged30.ScandinavianJournalofPublicHealth,34,378-286.

Silventoinen,K.,Moustgaarid,H.,Peltonen,R.,&Martikainen,P.(2013).Changingassociationsbetweenpartnershiphistoryandriskofaccidents,violenceandsuicides.JournalofEpidemiologyandCommunityHealth,67,265-270.

South,S.C.,&Krueger,R.F.(2012).Geneticstrategiesforprobingconscientiousnessanditsrelationshiptoaging.DevelopmentalPsychology,50,1362-1376.

Sprangers,M.A.G.,&Schwartz,C.E.(1999).Integratingresponseshiftintohealth-relatedqualtyofliferesearch:Atheoreticalmodel.SocialScience&Medicine,48,1507-1515.

Subramanian,S.V.,Kubzansky,L.,Berkman,L.,Fay,M.,&Kawachi,I.(2006).Neighborhoodeffectsontheself-ratedhealthofelders:Uncoveringtherelativeimportanceofstructuralandservice-relatedneighborhoodenvironments.JournalofGerontology:SocialSciences,61B,S153-S160.

Trumbetta,S.L.,Markowitz,E.M.,&Gottesman,I.I.(2007).Marriageandgeneticvariationacrossthelifespan:Notasteadyrelationship?BehaviorGenetics,37,362-375.

VanderSluis,S.,Dolan,C.V.,Neale,M.C.,&Posthuma,D.(2008).AGeneralTestforGene–EnvironmentInteractioninSibPair-basedAssociationAnalysisofQuantitativeTraits.BehaviorGenetics,38,372-389.

Vuorisalmi,M.,Lintonen,T.,&Jylhä,M.(2006).Comparativevsglobalself-ratedhealth:associationswithageandfunctionalability.AgingClinicalandExperimentalResearch,18(211-217).

Waite,L.,&Gallagher,M.(2000).TheCaseforMarriage:WhyMarriedPeopleAreHappier,Healthier,andBetterOffFinancially.NewYork:Doubleday.

Waite,L.,Laumann,E.O.,Das,A.,&Schumm,L.P.(2009).Sexuality:measuresofpartnerships,practices,attitudes,andproblemsintheNationalSocialLife,Health,andAgingStudy.JournalofGerontology:SocialSciences,64,56-66.

Ware,J.E.,Kosinski,M.,&Keller,S.D.(1994).SF-36PhysicalandMentalHealthSummaryScales:AUsers'Manual.Boston:TheHealthInstitute.

Weaver,D.A.(2010).Widowsandsocialsecurity.SocialSecurityBulletin,70,89-109.Williams,K.,&Umberson,D.(2004).Maritalstatus,maritaltransitions,andhealth:A

genderedlifecourseperspective.JournalofHealthandSocialBehavior,45,81-98.Zheng,H.,&Thomas,P.A.(2013).Maritalstatus,self-ratedhealth,andmortality:

Overestimationofhealthordiminishingprotectionofmarriage?JournalofHealthandSocialBehavior,54,128-143.


Table1:IGEMSstudiesStudy Label Reference N

subjectsAge

rangeVars

FinnishTwinCohort FTC KaprioandKoskenvuo(2002)

7870 53-67 SRH

Finntwin16 FT16 Kaprio,Pulkkinen,andRose(2002)

4246 21-29 SRH

LongitudinalStudyofAgingDanishTwins

LSADT Christensen,Holm,McGue,Corder,andVaupel(1999)

3311 70-102 SRHCOMPACT

Middle-AgeDanishTwins MADT Osleretal.(2008) 4037 45-68 SRH

COMPACT

MidlifeintheUnitedStates MIDUS Southand

Krueger(2012)1764 25-74 SRH

COMPACT

MinnesotaTwinStudyofAdultDevelopmentandAging

MTSADA FinkelandMcGue(1993)

835 25-92 COMPACT

OriginsofVarianceintheOldest-Old

OCTO-Twin

McClearnetal.(1997)


SwedishAdoptionTwinStudyofAging

SATSA FinkelandPedersen(2004)


TwinandOffspringStudyinSweden

TOSS NeiderheiserandLichtenstein(2008)


VietnamEraTwinStudyofAging

VETSA Kremenetal.(2006)


Note:SRH–self-ratedhealth;COMP=healthcomparedwithothers;ACT=healthinfluencesactivities


Table2:Numberoftwinpairs

Age

Group

SRH ACT COMP

Men Women Men Women Men Women

<50

MZ 525 674 303 359 316 361

DZ 584 648 329 340 332 350

50-59

MZ 778 548 590 237 587 234

DZ 858 801 517 256 516 258

60-69

MZ 380 464 202 247 201 248

DZ 549 706 209 223 212 222

70+

MZ 248 390 251 394 246 393

DZ 315 597 319 593 316 581

TOTAL 9065 5373 5369


Table3:Model-fitstatistics

Model SRH ACT COMP

-2LL df -2LL df -2LL df

Women

1.Fullmodel 75425 10198 38845 5214 38843 5205

2.DropallMSmoderation 75435* 10201 38849 5217 38847 5208

3.DropMSmoderationonA 75425 10199 38848 5215 38843 5206

4.DropMSmoderationonC 75427 10199 38845 5215 38843 5206

5.DropMSmoderationonE 75427 10199 38845 5215 38844 5206

Men

1.Fullmodel 68817 9310 39369 5368 39822 5377

2.DropallMSmoderation 68844** 9313 39399** 5371 39831* 5380

3.DropMSmoderationonA 68817 9311 39369 5369 39826* 5378

4.DropMSmoderationonC 68821* 9311 39373* 5369 39824 5378

5.DropMSmoderationonE 68821* 9311 39526** 5369 39823 5378

*Modelfitdifferssignificantlyfrommodel1atp<.05

**Modelfitdifferssignificantlyfrommodel1atp<.01


FigureCaptions

1. Distributionofmaritalstatusacrossagegroupsandgender

2. Meansforsubjectivehealthvariablesacrossagegroups,gender,andmaritalstatus

3. Geneticandenvironementalcomponentsofvarianceacrossageandmaritalstatusfor

women.SRH:self-ratedhealth;ACT:healthimpactsactivites;COMP:healthcompared

withother.

4. Geneticandenvironementalcomponentsofvarianceacrossageandmaritalstatusfor

men.SRH:self-ratedhealth;ACT:healthimpactsactivites;COMP:healthcomparedwith

other.


0%

10%

20%

30%

40%

50%

<50 50-59 60-69 70+

PercentSingle

AgeGroup

MaritalStatus

Men

Women


45

47

49

51

53

55

<50 50-60 60-70 70+

Means

AGE

SRH

45

47

49

51

53

55

<50 50-60 60-70 70+

Means

AGE

ACT

45

47

49

51

53

55

<50 50-60 60-70 70+

Means

AGE

COMP

SingleWomen MarriedWomen SingleMen MarriedMen


0

30

60

90

120

150

30 35 40 45 50 55 60 65 70 75 80 85

Variance

AGE

SRHWomen Totalsingle

Totalmarried

Esingle

Emarried

Asingle

Amarried

Csingle

Cmarried

0

30

60

90

120

150

30 35 40 45 50 55 60 65 70 75 80 85

Variance

AGE

ACTWomenTotalsingle

Totalmarried

Esingle

Emarried

Asingle

Amarried

Csingle

Cmarried

0

30

60

90

120

150

30 35 40 45 50 55 60 65 70 75 80 85

Variance

AGE

COMPWomenTotalsingle

Totalmarried

Esingle

Emarried

Asingle

Amarried

Csingle

Cmarried


0

30

60

90

120

150

30 35 40 45 50 55 60 65 70 75 80 85

Variance

AGE

SRHMenTotalsingle

Totalmarried

Esingle

Emarried

Asingle

Amarried

Csingle

Cmarried

0

30

60

90

120

150

30 35 40 45 50 55 60 65 70 75 80 85

Variance

AGE

ACTMenTotalsingle

Totalmarried

Esingle

Emarried

Asingle

Amarried

Csingle

Cmarried

0

30

60

90

120

150

30 35 40 45 50 55 60 65 70 75 80 85

Variance

AGE

COMPMenTotalsingle

Totalmarried

Esingle

Emarried

Asingle

Amarried

Csingle

Cmarried

edinburgh research explorer - university of edinburgh finkel1, carol e. franz2, briana horwitz3,...

Documents