Genetic determinants of healthy longevity
Mooijaart, S.P.
Citation
Mooijaart, S. P. (2007, December 4). Genetic determinants of healthy longevity. Retrieved
from https://hdl.handle.net/1887/12477
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the
Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/12477
Note: To cite this publication please use the final published version (if applicable).
VITAMINDRECEPTORANDCOGNITION|39
CHAPTER 3
VDRGENEVARIANTSASSOCIATE
WITHCOGNITIVEFUNCTIONAND
DEPRESSIVESYMPTOMSINOLD
AGE
MarisKuningas
SimonP.Mooijaart
JelleJolles
P.ElineSlagboom
RudiG.J.Westendorp
DianavanHeemst
NeurobiologyofAgeing,inpress
Abstract
VitaminDhasbeenrecentlyimplicatedinbrainfunction.Ourobjectivewastotest
whethergeneticvarianceinthevitaminDreceptor(VDR)geneisassociatedwithcognitive
functioninganddepressivesymptomsinoldage.Thestudywascarriedoutinthe
prospectivepopulationbasedLeiden85plusStudy.All563participantsofthestudywere
genotypedforCdx2,FokI,BsmI,ApaIandTaqIpolymorphismsintheVDRgene.Ourdata
revealedanoverallworseperformanceontestsmeasuringcognitivefunctioningfor
carriersofBsmI(p=0.013)andTaqI(p=0.004)polymorphisms,andofhaplotype2(BAt)
(p=0.004).Incontrast,carriersofApaIvariantalleleandofhaplotype1(baT)hadbetter
cognitivefunctioningtogetherwithlessdepressivesymptoms.Theseassociationscould
notbeexplainedbydifferencesincalciumlevels,andbyselectivesurvival,sinceno
associationsbetweentheVDRgenevariantsandcalciumlevelsandmortalitywere
observed.Inconclusion,ourresultsshowthatgeneticvarianceintheVDRgeneinfluence
thesusceptibilitytoagerelatedchangesincognitivefunctioningandindepressive
symptoms.
VITAMINDRECEPTORANDCOGNITION|41
INTRODUCTION
AmpledataprovideevidencethatvitaminDisinvolvedinbrainfunction.Thereported
biologicalprocessesinfluencedbyvitaminDinthebrainincludeneuroprotection,
immunomodulationanddetoxification(Brownetal.,2003;Garcionetal.,2002).The
neuroprotectiveeffectsofvitaminDappeartobeexertedviatheregulationofcalcium
homeostasis(Breweretal.,2001,2006;deViraghetal.,1989),andsynthesisof
neurotrophins,suchasnervegrowthfactorandneurotrophin3(Naveilhanetal.,1996;
Neveuetal.,1994;Saporitoetal.,1994;Wangetal.,2000).Thesebiologicaleffects
suggestthatvitaminDcouldinfluencecognitivefunctioningandtheprevalenceof
depressivesymptoms.
ThefunctionsofvitaminDaremediatedbyvitaminDreceptor(VDR),whichbelongsto
thenuclearhormonereceptor(NHR)superfamily,andwhichisubiquitouslyexpressedin
theorganism(Kameietal.,1995;Langubetal.,2001).DefectsinthevitaminDsignaling
systemhavebeenassociatedwithmultiplesclerosis,andvariousbehavioralandmood
disordersinanimalsandhumans(Cantornaetal.,1996;Garcionetal.,2002;Lansdowne
andProvost,1998;Mungeretal.,2004).Ithasbeenshownthatanimalsexposedto
prenatalvitaminDdeficiencyhavealterationsinbrainmorphology(Eylesetal.,2003),
locomotion(Burneetal.,2004;Kesbyetal.,2006),andlearningandmemory(Beckeret
al.,2005).Inaddition,micelackingafunctionalVDRgeneappeartosufferfromanxiety
likebehavior(Kalueffetal.,2004,2006).Inhumans,vitaminDdeficiencyhasbeen
associatedwiththepresenceofanactivemooddisorderandwithworsecognitive
functioning(PrzybelskiandBinkley,2007;Wilkinsetal.,2006).Incontrast,littleisknown
aboutwhetherandhowdisturbedfunctionoftheVDRgeneinfluencestheseendpoints.
TheVDRgenecontainsseveralpolymorphismsofwhichfive;Cdx2,FokI,BsmI,ApaIand
TaqI,havebeenmostofteninvestigated,andassociatedwithanumberofphenotypes,
suchasbonemineraldensity,andrisksforfracturesandcancer(Uitterlindenetal.,
2004b).Inaddition,haplotypealleleshavebeenidentifiedthatinfluencetheriskof
osteoporoticfracturesandtheexpressionoftheVDRgene(Fangetal.,2005;Grundberg
etal.,2007).Theriskhaplotypesthathaverecentlyemerged,baTandBAt,arecomposed
oftheBsmI,ApaIandTaqIpolymorphisms,locatedinthe3‘UTR.
TheaimofthisstudywastoassesstheinfluenceofthesefivepolymorphismsintheVDR
geneandtheriskhaplotypesoncognitivefunctioninganddepressivesymptomsinold
age.Furthermore,theassociationwithcalciumlevels,andtheincidenceoffracturesand
mortalitywereassessedfortheVDRpolymorphismsandthehaplotypes.Thestudywas
carriedoutintheLeiden85plusStudy,apopulationbasedprospectivestudyoftheoldest
old.
SUBJECTSANDMETHODS
SUBJECTS
TheLeiden85plusStudyisaprospectivepopulationbasedstudyinwhichall85yearold
inhabitantsofthecityLeiden,inTheNetherlands,wereinvitedtotakepart.Therewere
noselectioncriteriarelatedtohealthordemographiccharacteristics.Thepopulation
understudyconsistsof599subjects,allCaucasiansandmembersofthe19121914birth
cohort,enrolledinthemonthoftheir85thbirthdaybetween1997and1999(Bootsmavan
derWieletal.,2002).Forthepresentstudy,DNAwasavailablefor563participants.All
participantsoftheLeiden85plusStudywerefollowedformortalityuntilAugust1st2005.
PrimarycausesofdeathwereobtainedfromtheDutchCentralBureauofStatisticsand
categorizedaccordingtothe10thInternationalClassificationofDiseases(ICD10).The
MedicalEthicalCommitteeoftheLeidenUniversityMedicalCentreapprovedthestudy
andinformedconsentwasobtainedfromallparticipantsorincaseofseverecognitive
impairment,fromtheirguardian.
CALCIUMLEVELSATBASELINE
Calciumandalbuminconcentrationsweredeterminedinserumusingfullyautomated
analyzers(Hitachi747and911;Hitachi,Ltd,Tokyo,Japan).Totalcalciumlevelswere
adjustedforalbuminusingthefollowingformula:correctedcalcium=uncorrected
calcium[(40–albumin)×0.02](Palmeretal.,1988).
COGNITIVEFUNCTIONANDDEPRESSIVESYMPTOMS
OverallcognitivefunctionwasmeasuredwiththeMiniMentalStateExamination(MMSE)
(Folsteinetal.,1975).Fromthespecificdomainsofcognitivefunctioning,attentionwas
assessedwiththeStroopTest(Kleinetal.,1997),processingspeedwiththeLetterDigit
CodingTest(LDT)(Houxetal.,2002)andmemorywiththe12WordLearningTest,which
assessesimmediaterecall(WLTI)anddelayedrecall(WLTD)(BrandandJolles,1985).The
prevalenceofdepressivesymptomswasassessedwiththe15itemGeriatricDepression
Scale(GDS15)(DeCraenetal.,2003).Thetestsassessingspecificdomainsofcognitive
functioningcouldnotbeadministeredto92participantsbecauseofseverecognitive
impairment
measureme periodof4.2 oftestingar compositec (StroopTest computingt onthetests
THEINCID Theincidenc numberoff Whenthep information institutional participant.
Fracturesin
POSSIBLEC Sociodemo possiblecon includingind yearsofsch
GENOTYPI TheCdx2G (SNPs)were USA),consis PCR(Applied standardco (Eurogentec (rs10735810 genotypedu (Sequenom
(MMSEscore entofcognitiv
2years.Durin redescribede cognitivescore
t,LDT,WLTIa theaverage.A
measuringco
ENCEOFFR ceoffracture fractureswas
articipantwas .Inaddition,
lization,wasi Thecomposit cludedhip,w
CONFOUND ographicchara
nfounders.Ed dividualswith ooling),anda
NG
G/A(rs115688 egenotypedu stingofPCRp dBiosystems, nditionsexce c,Liege,Belgiu 0),ApaIA/C(r usingMassArr Inc.,SanDieg
e 18points).
vefunctioning
ngthestudy,p elsewhere(Ho ewascalculat ndWLTD)int Ahighercomp ognitivefunct
RACTURES
swasassesse obtainedbys sseverelycog thegeneralp nterviewedco teofselfrepo wristandother
ERS
acteristics,suc ucationwasd houtschooling ahighereduca
20)andBsmI usinganAssay
rimersandTa ,FosterCity,C ptforthefoll um)andonet rs7975232)an rayplatforma go,CA,USA).
.Allparticipan anddepressi parallelversio ouxetal.,200 tedbyconver oazscore((i positecognitiv
ioning.
edyearlydurin selfreporting
gnitivelyimpa ractitioner,or oncerningfrac ortedandphy
rfractures.
chassexand
dividedintotw gorwithonly
ationlevel(6y
C/T(rs15444 ybyDesign(A aqManMGBp CA,USA).Amp
owingmodific thirdoftheam ndTaqIA/G(r accordingtot
VITAMIND
ntswerevisite vesymptoms onsofthetest 2).Inaddition rtingthescore ndividualleve vescorereflec
ngthefiveyea usingastand ired,aguardi rthenursing
cturerelated
ysicianreporte
levelofeduca wolevels:alo
primaryscho yearsormore
10)singlenuc AppliedBiosys probes,onan
plificationrea cations.AqPC mountofassa rs731236)poly
heprotocolso
RECEPTORAND
edannuallyfo
duringamea tswereuseda ntothespecif esoftheindiv el–meanleve ctsbetterper
arfollowupp dardizedyes/n anwasasked homephysicia
contactswith edfracturesw
ationwereco owereducatio ooleducation
eofschooling
cleotidepolym tems,FosterC ABI7900HTr ctionswerem CRcorekitwa aymix.FokIG/
ymorphismsw ofthemanufa
COGNITION|43
orre
anfollowup
anddetails
fictests,a
vidualtests
el)/SD),and
rformance
period.The
noformat.
forthe
anincaseof
hthe
wereused.
nsideredas
nlevel,
(lessthan6
).
morphisms
City,CA,
realtime
madeat
asused
/A
were
acturer
STATISTICALANALYSIS
TheprogramHaploview(Barrettetal.,2005)wasusedtoestimateallelefrequencies,test
forHardyWeinbergequilibriumandtoestimatepairwiselinkagedisequilibrium(LD)
betweentheSNPs.Haplotypesandhaplotypefrequencieswerecalculatedusingthe
programSNPHAP(http://wwwgene.cimr.cam.ac.uk/clayton/software).Inordertotake
intoaccounttheuncertaintyinhaplotypeprobabilities,themultipleimputationapproach
wasused(Rubin,1987),andwithSNPHAPtendatasetsweregeneratedbyrandomly
assigningahaplotypetoeachsubjectaccordingtoitshaplotypeprobabilities.Allstatistical
analyseswereperformedwiththetendatasets.Thehaplotypespecificestimateswere
calculatedbyaveragingthetendatasetspecificestimates,andthestandarderrorswere
estimatedusingtheestimatedvariancewithinandacrossthedatasets.Theassociations
betweenbaselinecalciumlevelsandVDRpolymorphismsandhaplotypesweretested
usingsexadjustedlinearregression.Associationsbetweencognitivefunctioning,
depressivesymptomsandVDRpolymorphismsandhaplotypeswereanalyzedusingasex
andeducationadjustedlinearmixedmodel,estimatingtheoverallmeandifferencein
cognitivefunctioningordepressivesymptomsduringfollowup.Coxproportionalhazard
model,measuringtimetoeventwasusedtoestimatetheriskofincidentfractures,and
mortalityduringthefollowupperiod,inrelationtothepolymorphismsorhaplotypes.The
referencegroupcontainedzerocopiesofariskalleleorhaplotype.Allanalyseswere
performedwithSPSS,version12.0(SPSSInc.,Chicago,IL,USA)statisticalsoftware.
RESULTS
Demographiccharacteristicsandbaselinemeasuresofcognitivefunctioningand
depressivesymptomsofthe563participantsoftheLeiden85plusStudyarepresentedin
table3.1.AllstudysubjectsweregenotypedforCdx2,FokI,BsmI,ApaIandTaqI
polymorphismsintheVDRgene.ThegenotypefrequenciesoftheSNPs(table3.1)werein
agreementwithHardyWeinbergequilibriumandsimilartothosereportedinother
Caucasianpopulations(Uitterlindenetal.,2004).TheBsmI,ApaIandTaqIpolymorphisms
wereinstronglinkagedisequilibrium(LD)(D’>0.99)(figure3.1),anddefinedfive
haplotypes,ofwhichthefirstthreehadfrequencies>5%.Thesehaplotypeshave
previouslybeendescribedasbaT,BAtandbAT,respectively.Thehaplotypefrequencies
weresimilartothosereportedinotherCaucasianpopulations(Fangetal.,2005;
Grundbergetal.,2007).
Table3.1.Char
1Dataareprese MMSEMiniM Recall;WLTD
Figure3.1.The 14exons(indic indicatedwith
definefivehap beendescribed
racteristicsofstu Characteristic
Number
Age1 Female(%)
Lowlevelofedu Calcium(mmol/
MMSE(points)1
MMSEt19poin Specificdomain StroopTest(sec LDT(digits)
WLTI(pictures)
WLTD(pictures) GDS15(points) Polymorphisms2 Cdx2(G/A)
FokI(G/A)
BsmI(C/T)
ApaI(A/C)
TaqI(A/G)
entedasmedian MentalStateExam
WordLearningT
eVDRgenestruct catedwithboxes) arrows.TheBsm plotypes.Thefirst dasbaT,BAtand
dyparticipants
ucation(%)
/l)1
nts(%)
sofcognitivefunct conds)
)
1
2
swithinterquart mination;LDTL TestDelayedReca
tureandhaplotyp ).Theapproxima mI,ApaIandTaqIp
tthreehaplotype
bAT,respectivel tioning1
tileranges;2Data LetterDigitCodin all;GDS1515it
pes.TheVDRgen tepositionsofth polymorphismsa es,haplotype1,h
y.
VITAMIND
Value 563 85() 375(67
362(65
2.23(2.16
26(222 471(84
74(609 16(122 25(202 9(711 2(13
0.19 0.34 0.43 0.46 0.42 aarepresenteda gTest;WLTIWo temGeriatricDep
nespansagenom hefivepolymorph areinstronglinka haplotype2andh
RECEPTORAND
e
)
%)
%)
2.29)
28)
%)
97)
21)
28)
1)
3)
sminorallelefre ordLearningTest pressionScale
micregionof100
hismsanalyzedin agedisequilibrium haplotype3have
COGNITION|45
equencies;
tImmediate
kbandcontains
nthisstudyare
m(LD)and
previously
Globalcognitivefunctioning,attention,processingspeed,memoryandtheprevalenceof
depressivesymptomswereassessedatbaseline,age85years,andreexaminedannually
duringameanfollowupperiodof4.2years.Duringfollowup,asignificantdeclinein
cognitivefunctioning,andanincreaseindepressivesymptomswereobservedinall
participants(allp<0.001)(Vinkersetal.,2005).Thesechangeswerenotattributabletothe
Cdx2orFokIpolymorphisms,sinceduringfollowupnodifferencesincognitive
functioninganddepressivesymptomswereobservedforcarriersofthesepolymorphisms
(datanotshown).Ontheotherhand,carriersoftheBsmIandTaqIpolymorphisms
performedworseonalltestsmeasuringcognitivefunctioning(table3.2).Thisworse
performancewasreflectedbyalowercompositecognitivescore(BsmIptrend=0.013;TaqI
ptrend=0.004),butnotbyalowerMMSE,whichmeasuresglobalcognitivefunctioning
(BsmIptrend=0.999;TaqIptrend=0.899).
Table3.2.CognitivefunctioninganddepressivesymptomsduringfollowupdependentontheVDR
polymorphisms
VDRgenotypes
wt/wt wt/var var/var
Mean(SE) Difference(SE) Difference(SE) ptrend
BsmI(C/T)
Compositescore 0.08(0.06) 0.12(0.08) 0.25(0.10)* 0.013*
MMSE(points) 22.6(0.48) 0.47(0.61) 0.15(0.78) 0.999
StroopTest(seconds) 84.3(2.31) 2.00(2.95) 10.4(3.78)* 0.010*
LDT(digits) 16.2(0.51) 1.04(0.65) 0.34(0.83) 0.471
WLTI(pictures) 21.3(0.48) 1.07(0.61) 2.18(0.77)* 0.004*
WLTD(pictures) 7.33(0.23) 0.30(0.29) 0.78(0.39)* 0.037*
GDS15(points) 2.93(0.21) 0.05(0.26) 0.55(0.34) 0.158
ApaI(A/C)
Compositescore 0.22(0.07) 0.00(0.08) 0.16(0.10) 0.135
MMSE(points) 23.1(0.51) 0.32(0.63) 0.56(0.76) 0.456
StroopTest(seconds) 89.9(2.45) 2.65(3.05) 6.80(3.68) 0.068
LDT(digits) 16.4(0.53) 1.40(0.66)* 0.06(0.80) 0.737
WLTI(pictures) 19.8(0.51) 0.49(0.63) 1.61(0.76)* 0.041*
WLTD(pictures) 6.83(0.24) 0.23(0.30) 0.44(0.36) 0.222
GDS15(points) 3.42(0.21) 0.56(0.26)* 0.72(0.32)* 0.019*
TaqI(A/G)
Compositescore 0.07(0.06) 0.13(0.08) 2.94(0.10)* 0.004*
MMSE(points) 22.7(0.48) 0.55(0.60) 0.30(0.78) 0.899
StroopTest(seconds) 84.0(2.29) 2.12(2.92) 11.0(3.80)* 0.008*
LDT(digits) 16.3(0.50) 0.96(0.64) 0.63(0.84) 0.310
WLTI(pictures) 21.4(0.47) 1.15(0.60) 2.51(0.78)* 0.001*
WLTD(pictures) 7.40(0.23) 0.37(0.29) 0.99(0.37)* 0.009*
GDS15(points) 2.93(0.20) 0.06(0.26) 0.60(0.33) 0.135
*p<0.05;SE–standarderror;MMSEMiniMentalStateExamination;LDTLetterDigitCodingTest;WLTI
WordLearningTestImmediateRecall;WLTDWordLearningTestDelayedRecall;GDS1515itemGeriatric
DepressionScale
VITAMINDRECEPTORANDCOGNITION|47
Fromspecificdomainsofcognitivefunctioning,attention,immediateanddelayed
memorywereaffectedmost,whereasfortheprevalenceofdepressivesymptomsno
differenceswereobserved(table3.2).Incontrast,carriersoftheApaIvariantallele
tendedtohavelessdepressivesymptomsthannoncarriersduringfollowup(ptrend=0.019)
(table3.2).Thesedifferenceswereobservedforbothheterozygous(0.56points,95%CI:
1.07to0.04,p=0.036)andhomozygous(0.72points,95%CI:1.35to0.09,p=0.026)
ApaIvariantallelecarriers.Inaddition,theseparticipantsperformedbetter,althoughnot
statisticallysignificant,ontestsmeasuringprocessingspeed,attentionandmemory(table
3.2).
Thehaplotypeanalysesrevealedsimilarresultsasthosewiththeindividual
polymorphisms.Carriersofatleastonecopyofhaplotype1(baT),whichcontainstheApaI
polymorphism,hadlessdepressivesymptoms(ptrend=0.026),andperformedbetter,
althoughstatisticallynotsignificantly,ontestsmeasuringattention,immediatememory
anddelayedmemorycomparedtononcarriers(table3.3).Theoppositewasobservedfor
carriersofhaplotype2(BAt),whichcombinesthevariantallelesofBsmIandTaqI
polymorphisms(table3.3).Foralltheseassociations,analleledosagedependenteffect
wasobserved,whichwasmorepronouncedforhaplotype2(BAt)carriers,whohadmainly
impairmentsinattention(ptrend=0.008),immediatememory(ptrend=0.001)anddelayed
memory(ptrend=0.009)(figure3.2).
Inordertoexplorewhetherdifferencesincalciumlevelsorselectivesurvivalhave
influencedtheassociationsobservedwithcognitivefunctioning,weanalyzedtherelation
betweenthesephenotypesandVDRgenevariants.Incrosssectionalanalysesatage85
years,serumcalciumlevelswerenotassociatedwiththeVDRpolymorphisms(datanot
shown),excepttheCdx2polymorphism.Homozygous(0.05mmol/l,95%CI:0.10to
0.00,p=0.032)butnotheterozygous(0.001mmol/l,95%CI:0.02to0.02,p=0.955)Cdx2
variantallelecarriershadlowerserumcalciumlevelscomparedtononcarriers.No
associationsbetweentheVDRhaplotypesandcalciumlevelswereobserved.Likewise,the
mortalityrisksdidnotdifferbetweenthedifferentVDRpolymorphismandhaplotype
carriersduringmean4.2yearfollowupperiod(datanotshown).
Table3.3.Cogn
Haplotype1(ba
Compositesc
MMSE(points
StroopTest(s
LDT(digits)
WLTI(picture
WLTD(picture
GDS15(point Haplotype2(BA
Compositesc
MMSE(points
StroopTest(s
LDT(digits)
WLTI(picture
WLTD(picture
GDS15(point Haplotype3(bA
Compositesc
MMSE(points
StroopTest(s
LDT(digits)
WLTI(picture
WLTD(picture
GDS15(point
*p<0.05SE–
WordLearning DepressionSca
Figure3.2.Diff 2(BAt).Thepv
nitivefunctioning
aT)
ore
s)
seconds)
es)
es)
ts)
At)
ore
s)
seconds)
es)
es)
ts)
AT)
ore
s)
seconds)
es)
es)
ts)
standarderror;M
TestImmediate
ale
ferencesincognit valuerepresents
ganddepressive
0copies
Mean(SE)
-0.09 (0.07)
22.5(0.55)
84.3(2.72)
15.3(0.59)
22.6(0.56)
7.95(0.27)
2.60(0.24)
0.05(0.07)
22.0(0.53)
79.3(2.70)
15.2(0.59)
24.1(0.55)
8.51(0.26)
2.14(0.24)
0.10(0.05)
22.0(0.40)
83.3(2.05)
14.2(0.44)
22.9(0.42)
8.05(0.20)
2.14(0.18)
MMSEMiniMe Recall;WLTDW
tivefunctioningb theoverallmean
symptomsduring VD 1copy
Difference(SE
0.01(0.08)
0.32(0.62)
1.72(3.01)
1.44(0.66)*
0.52(0.62)
0.26(0.29)
0.48(0.26)
0.13(0.08)
0.57(0.60)
2.12(2.92)
1.01(0.64)
1.15(0.60)
0.37(0.28)
0.03(0.26)
0.12(0.09)
0.76(0.68)
3.17(3.30)
1.25(0.72)
0.76(0.68)
0.32(0.32)
0.39(0.30)
ntalStateExamin WordLearningTe
betweenparticip ndifferenceinco
gfollowupdepe DRhaplotypes
)
2copie Difference
0.13(0.
0.60(0.
5.78(3.
0.23(0.
1.47(0.7 0.39(0.3
0.69(0.3
0.29(0.1
0.29(0.
10.7(3.7 0.63(0.8
2.50(0.7
0.98(0.3 0.56(0.3
0.56(0.2
2.77(2.
13.3(10 3.74(2.
3.84(2.
2.13(1.0 0.48(0.9 nation;LDTLett stDelayedRecall
antscarrying0,
ognitivefunctioni
endentontheVD
es
e(SE)
10)
.76)
.67)
.80)
76)
36)
33)*
10)*
.77)
77)*
83)
77)*
37)*
33)
28)*
25)
0.5)
29)
20)
04)*
94)
terDigitCodingT l;GDS1515ite
1or2copiesof
ngduringfollow
DRhaplotypes
ptrend
0.215
0.424
0.124
0.576
0.056
0.254
0.026*
0.004*
0.912
0.008*
0.298
0.001*
0.009*
0.147
0.032*
0.124
0.153
0.024*
0.074
0.067
0.171
Test;WLTI
emGeriatric
thehaplotype
up.*p<0.05
VITAMINDRECEPTORANDCOGNITION|49
Sinceinotherstudiesthesamepolymorphismsandhaplotypesasanalyzedinthisstudy
havebeenassociatedwiththeriskoffractures,weassessedtheinfluenceofVDR
polymorphismsandhaplotypesontheincidenceoffracturesduringfollowup.Therewere
nodifferencesintheincidenceoffracturesbetweenthedifferentpolymorphismcarriers
(datanotshown).However,theadditionalhaplotypeanalysesrevealedatrendfor
increasedincidenceoffracturesforhaplotype1(baT)carriers(1copyHR:1.46,95%CI:
0.86to2.49;2copiesHR:1.52,95%CI:0.81to2.83).
DISCUSSION
ThemainfindingofthisstudyisthatgeneticvarianceintheVDRgenecontributesto
differencesincognitivefunctioninganddepressivesymptomsinoldage.Anoverallbetter
performanceontestsmeasuringattention,processingspeedandmemory,togetherwith
alowerprevalenceofdepressivesymptomswereobservedforcarriersofApaIvariant
alleleandofhaplotype1(baT),whichcontainstheApaIvariantallele.Incontrast,carriers
oftheBsmIandTaqIpolymorphismshadimpairmentsinattentionandmemory.Similar
associationswereobservedwithhaplotype2(BAt),whichcombinesthevariantallelesof
BsmIandTaqI.
TheresearchontheroleofvitaminDinthehumanbrainhassofarfocusedmainlyonthe
influenceofvitaminDonmooddisorders.VitaminDdeficiencyisconsideredasapossible
contributortoseasonalaffectivedisorder(SAD),sinceSADhasbeenassociatedwith
wintermonthsandsunlightdeprivation(Rosenthaletal.,1984;Schlageretal.,1993;
Spoontetal.,1991).Inaddition,therearestudiesshowingassociationsbetweenlow
vitaminDlevelsandmooddisorders,accompaniedwithworsecognitivefunctioning
(LansdowneandProvost,1998;PrzybelskiandBinkley,2007;Wilkinsetal.,2006).In
accordancewiththelatterstudies,weobservedinthisstudythatgeneticvarianceinthe
VDRgeneinfluencesbothcognitivefunctioninganddepressivesymptoms.Fromthe
specificdomainsofcognitivefunctioning,attentionandmemorywereaffectedmost.
Thesetwocognitivedomainsaremostvulnerableandtendtodeclineconstantlyacross
adultlifespan,incontrasttocognitiveabilitiessuchasautobiographicalmemoryand
emotionalprocessing,whichstayunchangedthroughoutlife(HeddenandGabrieli,2004).
OurdatasuggestthatcarriersoftheBsmIandTaqIpolymorphismsaremoresusceptible
fortheagerelateddeteriorationofcognitivefunctioning,whereastheApaIpolymorphism
contributestoaprotectiveeffect.
IntheassessmentofoverallinfluenceoftheVDRgenepolymorphismsoncognitive
functioning,weobserveddifferenceswithcompositecognitivescorebutnotwithMMSE.
Thecompositecognitivescoreusedinthisstudywascalculatedfromfourindividualtests
thathavebeendesignedtomeasurechangesinspecificdomainsofcognitivefunctioning.
Therefore,thecompositecognitivescoremightbemoresensitiveindetectingcognitive
impairmentsthanMMSE,whichhasbeendesignedtoassessglobalcognitivefunctioning,
andcontainsonlyfewitemsfromthespecificcognitivetests.Italsomightbethatdueto
the‘ceiling’effectoftheMMSE,mildimpairmentsincognitivefunctioningarenot
detectable(Houxetal.,2002).
ThereareseveralmechanismsthroughwhichvitaminDcanaffectmentalperformance.
ThedownregulationoftheexpressionofLtypevoltagesensitivecalciumchannels(L
VSCC)byvitaminDinhippocampalneurons,hasbeenshowntoreducetheinfluxand
excitotoxiceffectsofcalciumtoneurons(Breweretal.,2001).Thedetrimentalroleof
excessivecalciumformemoryformationandoverallcognitivefunctioningiswidely
acknowledged(SattlerandTymianski,2000;Thibaultetal.,2001;Vengetal.,2003).
However,thedifferencesincognitivefunctioningbetweentheVDRpolymorphismand
haplotypecarrierswereunlikelycausedbyincreasedordecreasedcalciumlevels,since
noneofthesepolymorphismsandhaplotypeswereassociatedwithcalciumlevels.
However,itisunknownhowwelltheperipheralcalciumlevelsreflectthoseinthebrain.
ThevitaminDendocrinesystemplaysanessentialroleinoverallcalciumhomeostasisand
thereforeweexpectedtoseedifferencesalsoinperipheralcalciumlevelswithinthe
polymorphismandhaplotypecarriers.Possibly,otherbrainspecificfunctionsofvitaminD
areresponsiblefortheobservedeffects.Ithasbeenshownthatinthebrain,vitaminD
increasestheproductionofneurotrophins,whichsupportthesurvivalofexistingneurons
andencouragethegrowthanddifferentiationofnewneuronsandsynapses(Naveilhanet
al.,1996;Neveuetal.,1994;Saporitoetal.,1994;Wangetal.,2000).Theseeffects
provideprotectionto,anddiminishcognitiveimpairmentunderlyingneurodegenerative
disorders.
Inpreviousstudies,severalpolymorphismsandhaplotypesintheVDRgenehavebeen
associatedwithbonemineraldensityandriskoffractures(Fangetal.,2005;Uitterlinden
etal.,2004).Theriskhaplotypesthathavebeenidentifiedincludehaplotype1(baT)and
haplotype2(BAt),whichwerereportedtocontributetoincreasedanddecreasedriskof
fractures,respectively(Fangetal.,2005;Uitterlindenetal.,2004).Inthisstudy,we
analyzedthesameriskhaplotypes,andobservedatrendforincreasedriskoffracturesfor
haplotype1(baT)carriers,whichisinaccordancewiththeotherstudies.However,forthe
samehaplotypecarrierswealsoobservedbetterperformanceontestmeasuringcognitive
functioning.Anexplanationfortheapparentcontradictioncouldbethatpeoplewith
bettercognitivefunctioningaremoreactiveandthereforemayfaceahigherriskforan
VITAMINDRECEPTORANDCOGNITION|51
incidentfallandfracture.ThelackofassociationsbetweentheVDRpolymorphismsand
haplotypesandmortalitysuggestthatselectivesurvivalhasnotinfluencedthe
associationsobservedwithcognitivefunctioning.Wespeculatethatthebettercognitive
functioninginthehaplotype1(baT)carriersisduetoahigherexpressionofthe
haplotype,sinceincreasedvitaminDlevelshavepreviouslybeenassociatedwithbetter
cognitivefunctioning.Theevidenceforthefunctionalityofhaplotype1(baT),however,is
contradictory.Inonestudyitwasreportedthatthehaplotype1(baT)resultsinlowerVDR
geneexpressionandincreasedmRNAdecay(Fangetal.,2005),whereasinanotherstudy
itwasshownthatthehaplotype1(baT)isoverexpressedinhumantrabecularbone
samples(Grundbergetal.,2007).Itmightalsobethattheassociationsobservedinthis
studyareresultedbyotherpolymorphismsthatareinLDwiththoseanalyzedinthis
study.Recently,severalnewpolymorphismsintheVDRgenehavebeenidentified
togetherwithacompletedescriptionoftheLDandhaploblockstructure(Fangetal.,
2005).
Thestrengthsofthepresentstudyincludethepopulationbasedsampleoftheoldestold
withahighincidenceofdepressionandcognitivedecline,andtheannualrepeated
assessmentofdepressivesymptomsandthefunctioningofvariouscognitivedomains.The
limitationsofthestudyincludethelackofvitaminDlevels,andinformationon
environmentalfactors,suchas(dietary)calciumandvitaminDintake,whichcouldhave
influencedtheassociations.Anotherlimitationistheascertainmentofincidentfractures
throughselfreportingbyaquestionnaire.Thismighthaveledtoascertainmenterrors,
butrandomerrorswouldonlyunderestimateassociations.Toourknowledge,thisisthe
firstandonlyreportofarelationshipbetweengeneticvarianceintheVDRgeneand
cognitivefunctioninganddepressivesymptoms,andthereforeuntilfurtherreplication,
thepossibilityofachancefindingcannotbeexcluded.
Inconclusion,ourresultsshowthatgeneticvarianceintheVDRgeneinfluencescognitive
functioningandtheprevalenceofdepressivesymptomsinoldage.