the use of animal models to study the effects of aging on cognition

0084-6570/97/0201-0339$08.00 339

GALLAGHER & RAPPNEUROBIOLOGY OFCOGNITIVE AGINGAnnu.Rev. Psychol. 1997.48:339–70Copyright© 1997by AnnualReviewsInc. All rightsreserved

THE USE OF ANIMAL MODELSTOSTUDY THE EFFECTSOF AGING ONCOGNITION

MichelaGallagherDepartment of Psychology, University of North Carolina at Chapel Hil l, Chapel Hill ,North Carolina27599

PeterR. RappCenter for Behavioral Neuroscience, State University of New York at Stony Brook,StonyBrook,NewYork 11794

KEY WORDS: memory, hippocampus,medial temporal lobe,frontal lobe,attention

ABSTRACT

This review addresses the importance of animal models for understanding theeffects of normal aging on the brain and cognitive functions.First, studies oflaboratoryanimalscan help todistinguishbetweenhealthyagingandpathologi-cal conditions that may contribute to cognitive decline late in lif e. Second,research on individual differences in aging, a theme of interest in studies ofelderly humanbeings, can beadvancedby theexperimental control afforded intheuseof animal models. Thereview offers a neuropsychological frameworkto compare the effects of aging in human beings, monkeys, androdents. Weconsideraging in relation to theroleof themedial temporal lobein memory, theinformation processing functionsof theprefrontal cortex in thestrategic useofmemory, and theregulation of attentionby distributedneural circuitry. Wealsoprovideanoverview of the neurobiological effects of aging that mayaccountfor alterations inpsychological functions.

CONTENTSINTRODUCTION..................................................................................................................... 340NORMAL AGING AND ANIMAL MODELS....................................................................... 340MEDIAL TEMPORAL LOBE SYSTEM................................................................................. 343

Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

Psychological Functions ofthe Medial Temporal Lobe...................................................... 343Psychological Functions ofthe Medial Temporal Lobein Normal Aging: Human

Beings......................................................................................................................... 345Psychological Functions ofthe Medial Temporal Lobein Normal Aging: Animal

Models........................................................................................................................ 346Neurobiology ofAging in the Medial Temporal Lobe......................................................... 348

FRONTAL LOBE SYSTEMS.................................................................................................. 355Psychological Functions ofFrontal LobeSystems............................................................. 355Psychological Functions ofFrontal LobeSystems inNormal Aging: Human Beings....... 356Psychological Functions ofFrontal LobeSystems inNormal Aging: Animal Models....... 358Neurobiology ofAging in Frontal Lobe Systems................................................................ 360

AGING AND ATTENTION IN HUMAN BEINGSAND ANIMAL MODELS................... 361CONCLUSION ......................................................................................................................... 363

INTRODUCTION

From a life-span perspective,the study of aging seeksto understandthechangingcapacitiesof theelderlyasa normaldevelopmental process.Withinthis framework, the biology of aging is an important determinant.Just asfunctionsandadaptivecapacitiesdependon thebiological developmentof theyoungchild or adolescent,later life providesa distinctive biologicalsettinginwhich familiar tasksareperformedandnew challengesaremet. In addition,environmentalfactorslate in life combinewith many decadesof a person’shistory to influencethe capacitiesof an individual. We offer the view thatimportantinsights into agingasa developmentalprocesscanbe providedbythestudyof animalmodels.Thereviewcoversareasof researchthat illustrateand supportthis premise.

NORMAL AGING AND ANIMAL MODELS

A majorchallengein thestudyof agingis to definetheboundariesof normalchangeas distinct from pathological conditions. Suchboundariesare recog-nized for developmentin early life. For example,landmarksaredefinedforphysicalgrowthandcognitivefunctions;departurefrom the normis identifiedby a failure to manifest certainchanges that areexpectedin the usual courseofdevelopment.The boundariesof normal aging,as distinct from pathologicalconditions, arelessclearlydefinedbecausecertainexpectedchangesin cogni-tion during agingdo not differ from theearliestmanifestations of age-relatedpathologicalconditions. For example,the incidenceof Alzheimer’s disease(AD) increaseswith advancingage.Although the occurrenceof AD is rela-tively rare before the age of 60, it becomesincreasinglyprevalentin thedecadesthatfollow. Impairedmemory,which is thehallmarksymptom of ADin its earlieststages,is also one of the most commoncognitive featuresofnondementedelderly individuals(Craik & Jennings1992,Crook et al 1986).Currently no definitive test can diagnoseAD in its earlieststagesso that

340 GALLAGHER & RAPP

Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

memoryimpairmentassociatedwith this pathological conditioncanbe sepa-ratedfrom memorydeclinethat might be attributableto nonpathological ag-ing.

Age-relatedneurodegenerativediseasessuchas Alzheimer’s havea slowprogressivecourse.If suchdiseasesremainundetectedfor manyyearsbeforeaclinically significantphaseof decline,relativelysubtlechangesin presumablyhealthyolder individuals that areascribedto normalagingmight, at leastinpart,bedueto occultpathologicalprocesses.Effectsof Alzheimer’s diseaseonthe brain, such as plaques and neurofi brillary tangles, that meet neuro-pathologicalcriteriafor diagnosis have been detectedin autopsymaterialfromindividuals not judgedto havebeenimpairedby clinical assessmentbeforedeath(Crystal et al 1988,Morris et al 1991).On this basis,it is likely thatstudiesof elderlyhumansubjectswill includesomeindividualswith unrecog-nized neurologicaldisease.

Additional findings from neurologicalassessments combinedwith func-tional testingsupportthis conjecture.Elderly individuals who haveradiologi-cal evidenceof medial temporallobe atrophy in the brain along with mildcognitive impairmentdeterminedby clinical screeningwere found to be athigh risk for developingdementia(de Leon et al 1993). At follow-up fouryearsafter the original assessment,25 of 86 such subjectshad receivedadiagnosisof AD. Brain atrophicchangesarealsofound in someelderly indi-viduals who show no evidenceof cognitive impairment in clinical assess-ments.In a sampleof approximately150 elderly subjects(ages55–88years,mean70.0 years)who were presumedto be healthy,Golomb et al (1993)foundatrophyof themedialtemporallobein certainindividuals.Althoughnoparticipantsin this studyshowedevidenceof cognitiveimpairmentin clinicalassessments,differencesbetweenthesubjectswith andwithout medialtempo-ral lobe atrophywere evident in more sensitivetestsof cognitive function.Subjectswith atrophyperformedmorepoorly on testsof recentmemorythantheneurologically normalindividuals.Muchcurrentinterestis alsofocusedona biological marker,the e4 allele of apolipoprotein E, associatedwith higherrisk for dementia (Corderet al 1993).Recentreportsindicatethatpresumablyhealthyindividuals bearingthis markerperform lesswell on assessmentsofmemorythantheir agedcohorts(Bondi et al 1995,Helkalaet al 1996).Suchfindings raisethe questionof whetherevenmild memoryimpairmentmightrepresent a very early indicatorof a pathologicalprocess.

Notwithstanding thefindings discussed above,otherevidenceindicatesthatcognitivealterationsin agingoccurapartfrom degenerativeneurologicaldis-ease.Age-associatedmemoryimpairment(AAMI) wasdefineda decadeago(Crook et al 1986) to identify elderly individuals who complainof memoryimpairmentby self-reportand have memory test performanceat least one

NEUROBIOLOGYOFCOGNITIVE AGING 341

Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

standarddeviationbelowthemeanestablishedfor youngadults.Accordingtothesecriteria it is estimated that the occurrenceof AAMI substantially exceedswhat would be expectedbasedon the prevalenceand incidenceof probableAlzheimer’s disease(prevalenceandincidenceof AAMI about35%and6.6%per year comparedwith 13% and 3% per year for probableAD) (Lane &Snowdon 1989). Furthermore, Youngjohn & Crook (1993) reported thatAAMI is stable in elderly individuals basedon follow-up assessmentfouryearslater. Theseinvestigatorsconcludedthat AAMI is a relatively benignconditionthatdoesnot follow a progressivecourse.Severallimitationsof thisstudy,however,areworth noting.First, a considerablenumberof participantsthat did not return for follow-up testing might havehadgreaterdeclinethanthosethat did, a phenomenonthat is problematicfor researchof this type.Second,the subjectsin this study,who were on average intheir early sixties atthe first assessment,werefollowed for a relativelybrief intervalof four yearswhentheincidenceof Alzheimer’s diseasecontinuesto below. Furtherexami-nation of such individuals into later decadesmight be neededto provide abetter indication of thecourse of AAMI.

A relatedissuein thedefinitionof normalagingconcernstheobservationofindividual differencesin theelderlypopulation. As indicatedabove,cognitivedeclineis evidentin someindividuals in theelderly population (e.g.AAMI),whereas functionis betterpreservedin otheragedindividuals.In contrastwiththe usual individual differencesthat exist at earlier points in development,variability is often describedas markedly increasedwith advancingage.Inaccordancewith this description, a recentsurvey of publisheddata foundgreatervariability (coefficientsof variation)amongtheelderlycomparedwithyoungadultsonanumberof measureswidely reportedto besensitiveto aging,e.g. reactiontime,memory,andfluid intelligence(Morse1993). Although thisphenomenonis often observed,the origin of increasedvariability in aging isnot well defined. Perhapsvariability in aging reflects differencesthat areexpressedonly in thelaterdecadesof life. Alternatively, individual differencesmight becomemagnified late in life becauseof the cumulative impact ofdifferent biological and experientialbackgroundsover many decades.Moreinformationaboutfactorsunderlyingindividual differenceswill be importantfor understandingnormalaging.

Thediscussionaboveprovidesa background for consideringthe usefulnessof animalmodelsin thestudyof aging.Thelikelihoodthatthesamepathologi-cal processes inhumandisease occur andaremanifest in identicalwaysacrossseveralspeciescouldbeconsideredquite low. Many progressiveneurologicaldiseasessuchasAlzheimer’s do not afflict speciescommonlyusedin labora-tory researchon aging,e.g.rats,mice,monkeys.However,it is reasonabletoexpect that at least somefeaturesthat characterizebiological aging of the


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

mammalianbrainwould be evidentin differentspecies.Thus,commonalitiesacrossspeciesmight helpto identify normalneurobiological aging,asdistinctfrom pathologicalconditions,andthosepsychological functionsmostaffectedby aging.

The useof laboratoryanimalscan addressother aspectsof humanagingthat haveprovendifficult to study in a systematic way. If normal aging ischaracterizedby increasedvariability, this phenomenonof individual differ-ences mightbeevident in otherspecies. Because cohorts oflaboratory animalscan be maintained under relatively controlled conditions it should also bepossibleto isolatefactors contributing tosuchindividualdifferences.

The relevanceof researchwith laboratoryanimalsfor anunderstandingofhumanaging,however,dependsonwhetherthespecificfunctions andbiologi-cal systemstargetedfor studyareappropriatemodelsfor humanaging.Scien-tific advancesover the past few decadeshave provideda foundation fordevelopingusefulanimalmodelsof aging.We outlinea frameworkcurrentlyemployedfor investigating theneurobiologicalbasisof functionalchangesinaging.Theapproachis built on a backgroundof researchin neuropsychology,cognitivepsychology, and neuroscience.

The field of neuropsychology originally led to clinical descriptionsandpsychometricprofilesfor certaintypesof braindamageasanaid to diagnosis.Becausetheconsequencesof certainformsof damagewereremarkablyselec-tive, neuropsychological studiesalso cameto serveas a basis for makinginferencesabout the normal functionof specific brainregions. Alongsidedevelopmentsthat camefrom neuropsychological research, cognitivepsychol-ogy hasgreatlycontributed to our understandingof psychological processes.Cognitive psychology studiesthe components and organizationof functionssuchas memorywithin the frameworkof informationprocessingand repre-sentation.As a relatedburgeoningfield, cognitive neuroscienceis building onadvances in psychology using new technologies, such as functional neuro-imagingandmethodsfor recordingtheensembleencodingof information byneurons,to studyinformationprocessingin the brain.Researchderivedfromthesetraditionsof neuropsychologyand cognitive neurosciencehascontrib-utedto thedevelopmentof animalmodelsin thestudyof aging,asexemplifiedby theareascovered inthefollowing sections.

MEDIAL TEMPORAL LOBE SYSTEM

PsychologicalFunctionsof theMedial Temporal LobePatientswith medial temporal lobe damagehave circumscribeddeficits inmemory; the syndromeincludesan anterogradeamnesiaand sparesremotememoryandgeneralintellectualcapacities(Corkin 1984,Scoville & Milner1957).Anterogradeamnesiarefersto the inability to remembernew informa-


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

tion and episodesof life that occurafter medial temporallobe damage.Theanterogradememoryimpairmentis consideredto representa defectin mecha-nismsthat allow long-termretentionof new material.In supportof this con-cept many domainsof information processingand immediate memory(e.g.digit span)arepreservedin theseamnesicpatients.Moreover,therecognitionthatpatientswith suchamnesiahaveareasof preservedmemory,e.g.primingand skill learning, advancedthe conceptthat the brain possessesmultiplememorysystems(Cohen& Squire1980).The domainof memoryin medialtemporallobe amnesiais variouslydescribedasdeclarativeor explicit mem-ory, referring to representationsin memorythat providea basisfor the con-sciousrecollectionof facts andevents.

Researchusinganimalmodelshassoughtto definethecomponentsof themedial temporallobe that contribute to the amnesicsyndrome.Considerableevidencefor deficitsin declarative memory has beenobtained inotherspecies,but agreementhasnot yet beenachievedabouttheunderlyingstructuressub-servingmemorywithin the medial temporalsystem.The most widely usedanimal model under study inthis areaof researchis arecognition memory taskperformed by rhesus (or cynomolgous) monkeys(Squire & Zola-Morgan1991).In this delayednonmatch-to-sample task,monkeysarepresentedwithanobjecton an information trial. After a variabledelay,theoriginal objectispresentedwith a novel object,andselectionof the novel object is rewarded.Considerable consensussurrounds theobservation thatdamageto corticalregionsof the medial temporallobe (perirhinal,entorhinal,parahippocampalcortex)producesa significant delay-dependentdeficit in this object-recogni-tion task (Meunier et al 1993,Suzuki et al 1993).Lessconsensushasbeenachievedconcerning the effectsof damageconfinedto thehippocampus (Mur-ray1996).However,it is clearthatdamageto thehippocampusaloneproduceslesssevereimpairmentthan damagerestrictedto the cortical regionsof themedial temporallobe. It is interestingto note that relatively similar findingshave been reportedin studies ofrodentsdesignedto parallel the delayednonmatch-to-sample task usedwith monkeys.Damageto cortical structuresassociatedwith thehippocampal formationproducesdelay-dependentimpair-mentsthatarenotobservedafterselectivedamageto thehippocampus(Otto&Eichenbaum1992,Wilner et al 1993).

Different perspectivesare offered to accountfor findings in this line ofresearch.By oneview, acommon function inmemoryis servedby thecompo-nent medial temporallobe cortical regionsand the hippocampus, with moresevere impairment resulting from more extensive damage to this system(Squire& Zola-Morgan1991).Another view is that the componentsof thissystemmay servesomewhatdifferent functionsin declarativememory.Forexample,the relative insensitivity of delayednonmatch-to-sampletasks to


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

damage ofthe hippocampusalone may indicate that themedial temporalcortical regions cansubservememory representationfor individual items inde-pendent ofthehippocampus, while thehippocampusis essential fortheforma-tion of morecomplexrepresentationsin memory(Eichenbaumet al 1994).Inagreementwith this distinction, certainrepresentationsthatprovidea basisforthe flexible useof information in memoryappearto be highly sensitive toselectivehippocampaldamagein laboratoryanimals.For example,it hasbeenarguedthat spatialinformation is an exemplarof this type of representation,andit is well documented that severedeficitsin spatialtasks areobservedafterlesionsof thehippocampus.Anotherinstanceof memoryrepresentationsensi-tive to disruption of the hippocampuscomesfrom a study using probesformemoryafter animalswere trainedon a set of nonspatialstimulus-stimulusassociations.Normal rats demonstrated two forms of flexible memory thatwere not shownby rats with selectivehippocampallesions,i.e. transitivity,reflectedin theability to compareacrossstimulus pairsthat sharea commonelement,and symmetry,referring to the ability to associatepairedelementspresented inthe reverse of thetrainingorder (Bunsey& Eichenbaum1996).

Additional researchis neededto establishmore firmly whetherthe brainregionsthatcomprisethemedialtemporallobeare functionallyheterogeneouswith respectto their rolesin declarativememory.Furtheradvances inthis areaof cognitiveneurosciencewill continueto providean importantbackgroundfor understanding theneurobiologicalbasisof alteredmemoryprocessesin theelderly.

PsychologicalFunctionsof theMedial Temporal Lobe inNormal Aging:HumanBeingsThe characteristicsof memory impairment in presumablyhealthy elderlyadultsappearto parallelthegeneralfeaturesof medialtemporallobeamnesia(Craik & Jennings1992).Remotememoryandimmediatememory(e.g.digitspan)arespared.Elderly subjects,however,performmorepoorly on typicaltestsof declarativememory(e.g.pairedassociates,delayedparagraphrecall).Suchdeficits point to involvementof medial temporallobe structures.Evi-dencethatalterationsin themedialtemporallobemayunderlieage-associatedmemoryimpairmentis notedabove;Golomb et al (1993) found that elderlyindividualswith hippocampalatrophyperformedlesswell on testsof delayedrecall.

Functionalneuroimagingstudiesarenow providing newinformation aboutthe relativeactivationof this systemin youngandelderly individualsduringperformanceof memory tasks.Grady et al (1995) measuredcerebralbloodflow during encodingand recognitionof faces.They reportedthat poorermemory performancein healthyelderlyindividualsrelativeto youngindividu-als was associatedwith a reductionin hippocampaland prefrontal cortical


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

activationduringencoding.In thecontextof this observation, it is noteworthythathypoactivity in themedialtemporallobeis not invariablyobservedin thebrainsof elderly subjects.In contrastwith the resultsof Gradyet al, anotherrecentstudyfound comparablemedialtemporallobe activationin youngandelderly subjectsduring successfulrecall (Schacteret al 1996).In that experi-ment,a word-stemcompletion taskwas administeredto produce eitherhigh orlow recallof studywords.When neuroimaging wasdoneduringretrievaltests,equivalenthippocampalactivation in young and agedgroupswas observedduring successful recollection (high recallversus either low recallor baseline),which indicatesthatanage-relateddeficiencylocalizedto themedialtemporallobe doesnot occur in all conditions where hippocampalactivation is ob-served.Theseresultsalsosuggestthat memorydeficits in the elderly arenotdueto deficientmedialtemporalfunction during the retrievalof information.Rather,deficitsmaybeattributable,at leastin part,to a functionalimpairmentin medial temporallobe processingof new information, i.e. encoding,thatserves as a basisfor later recognitionor recall.

Beforeturning to theexaminationof memoryperformancein agedlabora-tory animals,we reiteratethat age-associatedmemoryimpairmentis not evi-dent inall individualsin the elderly human population. Thus, it is ofinterestinstudiesof laboratoryanimalsto assesswhethersimilarities can be foundin theeffectsof agingonmemory thatresemblethosefeaturesencounteredin humanbeingsand to determinewhether individual differencesalso exist in otherspecies over thecourse of aging.

PsychologicalFunctionsof theMedial Temporal Lobe inNormal Aging:AnimalModels

The delayednonmatching-to-sampletaskusedto assessrecognitionmemoryin youngmonkeyswith medial temporallobe damagehasalsobeenusedtotestagedmonkeys(Mosset al 1988,Prestyet al 1987,Rapp& Amaral1989).Althoughagedmonkeyshavedifficulty in learningthis taskwith a very briefretentioninterval,givensufficient trainingvirtually all oldersubjectsareableto reacha criterionequivalentto youngmonkeys.Whenthememorydemandsarethenmanipulatedby increasingdelays,monkeysapproximately25yearsorolderareimpaired.Individual differencesin recognitionmemoryamongagedmonkeysarealsoobserved,with an impairmentin a subsetof agedmonkeysthat qualitatively resemblesthe effect of medial temporal lobe damageinyoung monkeys(Rapp & Amaral 1991). Furthermore,agedmonkeyswithsuchdeficits alsoperformmorepoorly on rapidly learnedtwo-choiceobjectdiscrimination problems, another assessmentthatis sensitive to medial tempo-ral lobe damage(Rapp1993).An additionalparallelin thepatternof impair-ment acrossagedmonkeysand young monkeyswith medial temporallobedamageis found in a task that increasesthe load of information in memory.


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

Subjectsare testedfor identification of eachnew item addedto an array ofpreviouslypresenteditems(Killiany et al 1995).Thus,studiesof nonhumanprimateshave demonstratedmemory impairments on testssensitiveto theintegrity of the medialtemporallobe.Moreover,thepresenceand/orseverityof suchdeficitsvariesconsiderablyin theagedpopulation.Becausetheneuralsubstratewithin themedialtemporallobesystemfor themostcommonly usedtask in this research,delayednonmatching-to-sample,is not clearly defined,age-relatedimpairmenton this assessmentmight reflect alterationsin medialtemporallobe cortical systemseitheraloneor togetherwith alterationsin thehippocampus.

Otherevidencefor alterationsin thehippocampusin agedmonkeyscomesfrom a reportof impairmentin the flexible useof informationin memory.Asnotedin the prior section,a testfor the useof information in memorythat issensitiveto selectivedamageof the hippocampus hasrecentlybeendemon-stratedby probesfor transitive inferencein rodents.After learninga set ofstimulus-stimulus associations,young intact rats infer relationsamong theitems,anability that is lacking in ratswith lesionsconfinedto thehippocam-pus(Bunsey& Eichenbaum1996).In a recentstudy,monkeysthat learnedahierarchyof object-objectdiscriminationswere tested for theiruse ofinforma-tion in memory.The performanceof agedmonkeysduring probesfailed toshowresponselatencyeffectsthataretakento reflectthe relationalprocessingof informationthatunderliestransitiveinference (Rappet al 1996).

Deficits that are widely studiedin agedrodentsin spatialtasksmay alsoreflect a declarativememory impairment. Aged rats, like young rats withdamageto thehippocampus,havedeficits in a varietyof spatialtasks(for anoverview,seeGallagheret al 1995).Moreover,impairedspatial learninginagedratsor in youngratswith hippocampaldamagecanbe demonstratedtooccur independentof decline in sensorimotor/motivational functionsand inlearningthat is guidedby a stimulus or objectusedasa local cue(Gageet al1989, Gallagheret al 1993). In addition, individual differenceshave beenparticularlywell documentedin this line of research.Amongcertainstrainsofrats, a proportionof agedanimalsexhibit highly preservedperformanceonsuch tasks,while other agedcohortsperform entirely outsidethe rangeofyoungrats (Gallagher etal 1993).

With respectto the neurologicalbasisfor impairmentsof agedrodentsonspatialtasks,notethat thesetestsdo not providean entirely selectiveassess-ment of the function of the hippocampus (Gageet al 1984). For example,youngratswith lesionsof corticalsystemsinterconnectedwith hippocampus,e.g.entorhinal/perirhinal cortex,alsoexhibit deficitsin spatial tasks(Nagaharaet al 1995). Thus additional assessment is needed to distinguish betweenimpairmentthat may havea basisin the alteredstatusof the hippocampus


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

versusthesecorticalregions.As notedearlier,a versionof delayednonmatch-to-sampledevelopedfor studiesof rodentshasrevealeda sensitivity to en-torhinal/perirhinal cortexdamagethat wasnot seenafter selectivedamagetothe hippocampus in young rats (Otto & Eichenbaum1992). A recentstudyreportedthat agedrats that learnedthis task at short retentionintervalsper-formedno differently thanyoungratswhenincreasingdelayswere introduced(Zyzaket al 1995).Thesameratsusedin this assessmentwerealsoevaluatedin aspatial task whereanage-relateddeficit wasobserved.Thus,agedratsthatareimpairedin a spatialtaskthat is sensitive to theintegrity of thehippocam-puscandisplayintact performanceon an assessmentthat is moreselectivelysensitiveto the integrity of relatedcortical regionsin the temporallobe.Suchresultssupportthe conceptthat impairment inspatialtasksreflects an effect ofagingon thehippocampus.

Behavioralstudiesdemonstratinga decline in functions associatedwithmedial temporallobe structuresin laboratoryanimalsprovide evidenceforage-associatedmemory impairment independentof pathological conditionsthat affect the elderly humanpopulation.Moreover,individual differencesinaging are often documentedin studiesof nonhuman primatesand rodents,which providesanadditionalparallelwith observationsin humanbeings.

Neurobiology ofAging in the MedialTemporalLobeMuch currentresearchin laboratoryanimalsis directedat theneurobiologicalbasisof decline in cognitive functionsassociatedwith the medial temporallobe. Two conceptsabout the basis for this decline in aging, which haveprevailedfor severaldecades,haverecentlycomeundernew scrutiny.In thefirst case,neurodegeneration withinthehippocampushadbeenthought toplaya significantrole (Meaneyet al 1988).A secondinfluential conceptheld thatdegenerationin thebasalforebraincholinergicsystem, a componentof whichinnervates thehippocampusand relatedcortical structures, provides abasisformemory deficits in aging (Bartus et al 1982). In eachcase,recentstudiesindicatethat theselong-standing conceptsmay be incorrect.After discussingthe researchdealingwith thosetopics,we considerotherpotentialsubstratesfor age-related lossof functionin the medialtemporallobe system.

The conclusion that neuron loss occurs inthe aged hippocampus wasreachedin earlier studiesof human,nonhuman primate,and rodent brains(Brizzee etal 1980, Dam 1979, Issaet al1990, Meaney etal 1988). Insome ofthesereports,theconceptthatneurodegenerationcausesage-relatedcognitivedeclinewas further bolsteredby evidencethat the presenceand severityofbehavioralimpairmentwere correlatedwith loss of principal neuronsin thehippocampus(Issaet al 1990).It is importantto note,however,thatstudiesofneurodegenerationwere originally basedon methodsfor estimating neurondensity.Neurodegenerationis now beingstudiedusingnew methodsthat are


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

unbiasedfor many factorsthat could influencemeasuresof neurondensity,suchasthesizeof neuronsandthesizeor composition of nonneuralcells,andso forth. Recentresults usingthesenew methods formeasuring thetotalnumberof neuronsin a brainstructureindicatethatno neuronlossappearstooccur in thehippocampusduringnormalagingacross a variety ofspecies.

Studiesof aging in human,nonhuman primate,and rat brain have nowshown equivalent numbersof theprincipal neurons(Figure1) in thehippocampus(Rapp& Gallagher1996,Rasmussenet al 1996,West1993,Westet al 1993).Moreover, therecent anatomicalstudiesof ratsshowed that even agedanimalswith substantial deficits indicating hippocampal dysfunctionexhibitedno lossof neurons in this structure. Forexample, Rapp& Gallagher(1996)observedawide range of spatial ability in the aged ratsincluded in their study butrelatively little variability in numbers of hippocampal neurons and no suggestionof a declinein neuronnumberassociated withage orbehavioral impairment.

Additionalevidenceindicates thatneuronnumberis preservedin thehippo-campalformationof agednonhumanprimates,including thosemonkeyswithidentified memoryimpairment(West et al 1993). In addition to comparablenumbersof principalneuronsin thehippocampusproper,this studyfoundnosignificantlossof neuronsin corticalregionsassociatedwith thehippocampalformation(e.g.entorhinalcortexandsubiculum). This finding is particularlynoteworthybecausea subsetof theagedmonkeyshaddeficits in recognitionmemory assessed indelayednonmatch-to-sample, a taskthatis sensitive totheintegrityof cortical areas inthe medialtemporallobe.

Neurodegenerationin thehippocampuswasformerly viewedasaninevita-ble consequenceof normal aging.Reportsthat the principal neuronsof thehippocampusarepreservedacrossa varietyof species,evenin thepresenceofsubstantialbehavioralimpairment,mayprompta shift in theview thatneuro-nal loss in this structureservesas a basisfor age-relatedcognitive decline.Neurodegenerationin the hippocampus also appearsto distinguish certainpathologicalconditions from normal aging. In contrastwith normal aging,stereologicalmethodsdetectsignificantreductionsin thenumberof principalneuronsin the hippocampusin individualswith diagnosedAlzheimer’s dis-ease (Westet al 1994).

A secondlong-standingconceptaboutthe neurobiological basisof cogni-tive decline in aging hasfocusedon the basalforebraincholinergicsystem(Bartuset al 1982).Cholinergicneuronswithin this systemthatarelocatedinthe medialseptumandvertical limb of the diagonalband(MS/vDB) provideinnervationof thehippocampusandrelatedmedialtemporalcortex(Koliatsoset al 1990). Atrophy and degenerationof theseneuronsis detectedin agedbrains (Fischeret al 1989, Smith & Booze 1995, Stroessner-Johnsonet al1992), and marked pathology affects this system in Alzheimer’s disease


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

(Coyle et al 1983,Davies& Maloney1976).It is furthernoteworthythat thenumberand size of theseneuronsare affectedby age acrossa variety ofspecies, including rats, monkeys, and human beings.The possibility thatneurodegenerationwithin this population of neuronscontributesto functionalimpairmenthasfound supportin numerousstudies showingthat theamount ofcholinergicneurondeteriorationis relatedto theseverityof behavioraldeficitin agedsubjects(for anoverview,seeGallagheretal 1995).Notethatthevastmajority of thesecorrelationalstudies haveusedrat performanceon spatialtasksasthebehavioralassessment. Recentevidence,however,challengestheconclusionthat deteriorationof cholinergicneuronscan accountfor age-re-lated impairmentsin spatialtasks(Gallagher& Colombo 1995).At issueinthis line of researchis not whetherdegenerationoccurswithin thebasalfore-brain cholinergicsystem,but whetherthat effect of aging causesbehavioraldecline.

One way to assessthe contribution of the septohippocampalcholinergicsystemto age-relatedimpairments is to examinewhether removingthese

Figure 1 The schematic shows the principal neurons of the hippocampal formation, the granulecells of the dentate gyrus (DG), and the pyramidal neurons of the hippocampus proper (areasCA1–CA3). Twomajor input pathwaysto thehippocampal formationareil lustrated.Input fromthelateral andmedial entorhinal cortex (Lat Ent and MedEnt) provideshighly processed informationandterminatesprimarily on the dendritesof the granule cells in the dentategyrus. A subcorticalpathway(Septal Inputs) entersthrough another route andprovideswidespreadinnervation of thehippocampal formation. This latter input includesthe cholinergic innervation of the hippocampalformation. A trisynaptic circuit through the hippocampus (not shown) begins with the synapsesformedby theentorhinal cortexprojection onto thedentategranulecells.Thosecells in turn projectto the CA3 area.The CA3 pyramidal neurons form a third setof synapseswithin this structure,projecting onto thepyramidal cells inthe CA1area.


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

neuronsin younganimalsreproducesthedeficitsobservedin aging.A newlydevelopedimmunotoxin 192 IgG-saporincan be usedto target selectivelybasalforebraincholinergic neurons.After injection of the immunotoxin intotheMS/vDB, a nearlycompleteremovalof thecholinergicinnervationof thehippocampuscanbeachieved.It hascomeasa surprise,giventhecholinergichypothesisof age-relatedimpairment, that immunotoxin-inducedlesionsoftheseptohippocampal cholinergicsystemfail to producereliablespatiallearn-ing deficits(Baxteret al 1995a,1996;Berger-Sweeneyet al 1994,Torreset al1994).Young ratswith over 90% depletion of thecholinergic-specific enzymecholineacetyltransferase(ChAT) in hippocampusperformnormallyon a spa-tial learningprotocolthat is highly sensitiveto deficits in agedrats(Baxteretal 1995a),andno impairmentis even detectedafter removalof the entirebasalforebraincholinergic system,including theinput to thehippocampalformationand the widespreadinnervationof the cerebralcortex (Baxter et al 1996).Moreover,comparablecholinergiclesionsin agedratsdonotappearto exacer-bateor induceimpairmentin spatialtasks(Baxter& Gallagher1996).Thusitis unlikely that deteriorationin the septohippocampalcholinergicsystembyitself providesa sufficient basisfor age-relateddeficits that are commonlyobservedin spatial tasks.We return in a later sectionof this review to thefunction of the cholinergic innervationof the hippocampus and its possiblecontribution tobehavioral deficitsin aging.

A current themein researchon normal aging is that a reductionin thenumberof synaptic connections, ratherthanfrankneurodegeneration,providesa basis for age-relatedalterationsin cognition. For example, atrophyanddegenerationin the basalforebraincholinergicsystemis likely to result insomelossof hippocampal innervation by theseneurons.In agreementwith thisexpectation,thecholinergicresponsemediatedby stimulationof theseptohip-pocampalinput is reducedin all areasof theagedrat hippocampalformation(Shen& Barnes1996).However,the failure of specificcholinergic lesionstoreproducememorydeficits that havebeenwell documentedin agedrodentsmakesit unlikely thatthis alteration,by itself, servesa broadbasisfor impair-mentsin aged rats.

Apart from the subcorticalinput to the hippocampus that originatesin theseptalregion,anadditional lossof synapticinput from anothersourceis welldocumentedto occur in aging.That input, which providesthe primary routefor transferof highly processedcortical informationto the hippocampus,de-rives from neuronsin theentorhinalcortex(refer to Figure1). Ultrastructuralstudiesof rat brain havedemonstrateda significant loss of synapticconnec-tions in thehippocampalformationthatareformedby entorhinalcortexinput(Geinismanet al1992). Inaddition,individualdifferencesin cognitive declineamongagedratsare reportedto coincidewith differencesin the lossof this


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

innervation(Geinismanetal 1986).Reducednumbersof synapticconnectionsmay also occur in other areasof the hippocampalformation (Barneset al1994).

It might bereasonableto expectthat reducednumbersof synapticconnec-tions in thehippocampal formationcouldprovidea basisfor behavioraldefi-cits that dependon the integrity of this system.However,it shouldbe recog-nized that a numberof factorsmight argueagainstthat outcome.Neurobi-ologicalsystemspossessa numberof mechanismsthataregearedto maintainfunction.Considerableevidenceindicatesthatsuchmechanisms arerecruitedin theagedbrain.For example,althoughfewersynapticconnectionsaremadein thedentategyrusof thehippocampalformationin agedrats,the responsetoinput at the remainingsynapsesis greatly increased(Fosteret al 1991). Inaddition,a lossof synapses from corticalinput induces asproutingresponse inwhich newconnectionsareformedby inputsfrom othersources.Evidenceforsuchsproutingin the hippocampaldentategyrus hascomefrom studiesofagedbrainsand is alsoobservedin Alzheimer’s disease(Geddeset al 1992,Nicolle et al 1996).This typeof sproutingmaybecompensatoryin nature,asnewconnectionsaremadeto replacethosethatarelost. However,suchreac-tive growth may not alwaysbe beneficial.New connectionsmight addto theadverseeffectsof agingbecausethey comefrom a different sourcethan theoriginal input. Studiesof agingoften lack thefunctionalanalysisnecessarytodistinguishbetweenthesepossibilities. Recentstudiesusingbehavioralassess-ment along with neurobiologicalanalysisare being conductedto evaluatewhethersuch alterationsin the brain provide protectionor impose furtheradverseeffects onoutcome(Nicolle etal 1996,Stack et al 1995,Stenverset al1996).If a neurobiological responsein the agedbrain is compensatory,thenthedegreeto which this responseoccursshouldpredicta betteroutcome,i.e.lessbehavioralimpairment. However,if a reactiveprocessonly addsto dys-function thenthe presenceof sucha changein the brain might be associatedwith greaterimpairment.Researchof this type shouldleadto a betterunder-standingof the consequencesof reactiveandreorganizationalprocessesthatoccur in theaged brain.

Studiesof neurodegenerationandsynaptic connectivity provideimportantinformation aboutstructuralfeaturesof the brain during aging. Apart fromsuchstructuralfeatures,effectsof agingareevidentin thefunctionalintegrityof existingneuronsandconnections.Suchchangesmaybeimportantfactorsindiminishingtheoverall performanceof themedialtemporallobesystemdur-ing aging.

Neuronsin thehippocampalsystempossessa complexarrayof biologicalmechanismsfor information processing.In addition to specializedreceptorsfor specificinputs,receptorsarecoupledto a varietyof transductionsystems


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

thatarenecessaryto producephysiological responsesto thoseinputs(refer toFigure 2). Furthermore,transduction mechanisms are not only usedfor theprocessingof informationtransmittedbetweenneuronsbut alsoplay a role inalteringthefunctionalpropertiesof synapticconnections.Long-termpotentia-tion (LTP), referringto thelong-lasting increasein theeffectivenessof synap-tic connectionsthat can be readily inducedin the hippocampalsystem, hasattractedwidespreadinterestas a possible physiological mechanismfor thestorage of information in the mammalian brain.

Studieson transduction mechanismsmayprovideinsight into thebasisforchangesin informationprocessingandinformation storageduringaging.Thefindings from such researchare especiallyinformative in a systemwhereneurodegeneration is not a prominentfeature of aging. In such a setting,measurable decreases in componentsof transductionsystems,includingrecep-tors, couplingmechanisms, andsecondandthird messengers,do not merelyreflect a loss of neuronsbut indicatea changein the functional integrity ofexisting neurons.For this reason,recentevidencefor preservednumbersof

Figure 2 Schematic il lustratesinterneuroncommunication. Transmission from aninput producesa postsynaptic responsethrough a variety of transduction mechanisms.The components of thiscommunicationsystemincludereceptorsthateitherdirectly regulateaneuron’s excitabil ity orworkthrough biochemicalcascades (2nd and3rd messengersystems).


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

neuronsin the hippocampus is important for interpretingthe effectsof agingon otherneurobiological measures.

In additionto anabsenceof frank neurodegenerationwithin thehippocam-pus during aging, at least sometypesof receptorsthat serveas targetsforneurotransmittersarealsorelativelyunaffectedby agingin this structure.Forexample, substantial preservationof postsynaptic receptorsfor acetylcholine isreportedin a numberof studies(Chouinard et al 1995, Quirion et al 1995,Smith et al 1995).To the extentthat postsynaptictargetsarepreserved,thenstrategiesto addressthe effects of aging might be developedto increasefunction at thesesites.Sucha rationalehasservedasa basisfor developingdrugs to compensate for age-related deterioration of the basal forebrain choliner-gic neurons.Acetylcholinesterase inhibitors (e.g.Tacrine) areintendedto aug-menttheactionof acetylcholineat its receptorsin theforebrainby preventingthedegradationof this transmitter (Davisetal 1992,Thal etal 1989).Becausesuch drug treatmentscan improve cognitive function in young rats, somebenefitmight beachievedin agedanimalsevenif thebasisfor impairmentisnot solely due to a cholinergic defect (Gallagher& Colombo 1995). Theeffectivenessof augmentingcholinergicfunction would depend,however,onan intact physiological responseat cholinergic receptors. In thecaseof those recep-tors,however,a bluntedresponseto cholinergicstimulationhasrecentlybeendocumentedin theagedrodenthippocampus(Chouinardet al 1995,Undieetal 1995). In one of these studies, a greater reduction in the postsynaptic re-sponseto cholinergicstimulation wasseenin agedratsthatwerefoundto havemorepronouncedcognitive impairment(Chouinard etal 1995).Moreover,theneurobiological defect in this caseappearedto resideat a point in the bio-chemical machinery that would potentially affectthephysiologicalresponsetoother transmitter/receptor inputsthat usethe sametransductionpathway.

In addition to the role that transductionmechanismsservein informationprocessing,long-lasting changesin the propertiesof synapsesdependon thefunctionalintegrity of neurons.It haslong beenrecognizedthatalterationsinthemechanismsrequiredfor neuralplasticity couldprovidea basisfor cogni-tive declineduringaging.Oneof theearlieststudiesof individual differencesin spatiallearningin agedratsshowedthat this impairmentcorrelatedwith adeficiencyin neuralplasticity in hippocampus(Barnes1979).In thoseexperi-mentsanin vivo studyof LTP at perforant/dentatesynapseswasconductedinthesameratsthatwerebehaviorallytestedin a spatiallearningtask.Althoughasymptotic (saturated)LTP did not differ betweenthe agegroups,this LTPwasachievedlessreadily anddecayedmore rapidly in the agedrats than inyoung rats.Furthermore,the impairmentin behaviorallearningwas signifi-cantlycorrelatedwith theeffectof agingon LTP. Additional studiescontinueto documentthatLTP is adverselyaffectedduringtheagingprocess(Barnes&


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

McNaughton1985,Mooreet al 1993).A modestlossof the receptorsrequiredfor inductionof LTP mayoccur inthe aged hippocampus inbothprimatesandrodents(Clark et al 1992,Gazzaleyet al 1996,Nicolle et al 1996,Wanget al1996). A highly active areaof researchaimedat defining the mechanismsunderlyingLTP in adultanimalswill providea routefor betterunderstandingthe basisof thedeteriorationseen inaging.

Consistentwith a varietyof evidencefor functionalalterationsin theagedhippocampalformation,recordingthe activity of hippocampal neuronswhileanimalsperformcertaintaskshasprovidedevidencethattherepresentationofinformation inthissystem isalteredin olderanimals.Suchstudieshaveshownless reliability and specificity of the information encodedby hippocampalneurons(Barneset al 1983,Mizumori et al 1996;but seeMarkuset al 1994).In a recent studyof aged rats,individual differenceswerealsoevident.In agedrats that were impairedin a cognitive assessmentof spatial learning,repre-sentationsof relationships amongstimuli in a spatialenvironment wererela-tively impoverishedand inflexible comparedwith either young rats or agedcohortswith preservedbehavioralfunctions(Tanila et al 1996).Furtherre-search of thistypewill help toelucidatethe computationalcost ofthe neurobi-ological effectsof aging within this brain system.Identification of mecha-nismsunderlyingdiminishedfunction within that circuitry will undoubtedlyprovidean impetusto the development of new therapeuticstrategiesto treatage-related impairment.

In conclusion,cognitive impairments that resemblethoseseenin elderlyhumanbeingscanbe observedin the studyof agedlaboratoryanimals.Indi-vidual differencesin the effectsof agingon taskssensitive to the integrity ofthe medial temporallobe are also mirrored in the presenceand severityofsome neurobiological changes found in this system. Beyond theeffort toconstructa descriptionof normalaging,researchusinganimalmodelsprom-ises toprovide a setting forproductive researchon mechanismsof brainaging.This may includea betterunderstandingof how the rateor severityof agingprovidesa basisfor individualdifferences in cognitive abilitieslate inlife.

FRONTAL LOBE SYSTEMS

PsychologicalFunctionsof FrontalLobe Systems

In contrastwith amnesiaresultingfrom medialtemporallobedamage,humanbeingswith frontal lobe lesionsperformaccuratelyon manystandardtestsofdeclarativememory(reviewedin Moscovitch & Ulmita 1991).Currentper-spectivesinsteademphasizethat the prefrontalcortex supportsa variety oforganizationalprocessesthat importantly influencethe strategicuseof mem-ory. In addition, compelling evidencehasrevealedfunctional heterogeneity


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

acrossthe regionscomprising the prefrontalcortex. A popularview is thattheseareasfunction in a “central executive”capacity,mediatingthe on-linemanipulation of memory,particularly undercircumstancesemphasizingthespatial,temporal,or other contextualattributesof acquiredinformation (re-viewedin Moscovitch & Ulmita 1991).Recentevidenceconsistentwith thisview comesfrom neuroimagingstudiesin normal humansubjects.Cerebralblood flow is selectivelyincreasedin a region of the dorsolateralprefrontalcortexwhenmemoryfor temporalorderis necessaryfor successfulperform-ance relativeto conditionsinvolving thesamesensoryandmotordemandsbutlackinga temporalordercomponent(Petrideset al 1993a,b).A slightly moreposteriorprefrontalregion(area8), by comparison,is activatedduringacondi-tional discrimination procedure placingrelatively greateremphasison theenvironmentalcontingencies governing ongoing behavior (Petrides et al1993a).Suchdatasupportthe conceptthat the prefrontalcortexcomprisesavariety of functionally distinct subsystems. In addition, this backgroundofinformationhelpsto accountfor the patternof impairmentsobservedfollow-ing frontal lobe damage,which includesprominent deficits in memory fortemporalorder, impaired recall for the sourceof acquiredinformation (i.e.sourceamnesia),anddifficulties modifying behaviorappropriately inresponseto changingenvironmentalcontingencies(i.e. perseveration)(Janowskyet al1989a,b;Shimamuraet al1990).

Another featureof specializationwithin the prefrontalcortexis suggestedby studiesfocusingon the componentprocessesof declarativememory,e.g.encoding,retrieval,andsoforth. As notedpreviously,encodingprocessesandthesuccessfulconsciousrecollectionof eventsareassociatedwith hippocam-pal activation (Gradyet al 1995,Schacteret al 1996).Lateralizedprefrontalcorticalactivationis particularlyassociatedwith effortful retrievalof informa-tion from memory (Schacteretal 1996).This observationsupports theconceptthat theactivity of prefrontalcortexis engagedby specificretrievalstrategiesin support ofdeclarative memory (Buckner & Petersen1996).By this account,successfulperformanceon testsof declarativememoryin patientswith frontaldamagepresumablyreflectsthe utilization of alternateretrievalmechanismsmediatedby intact structures.Findingsfrom studiesof normalhumanaging,reviewed inthe nextsection,are consistentwith this proposal.

PsychologicalFunctionsof FrontalLobe Systemsin NormalAging: HumanBeings

Althoughdeficitsin declarativememoryarefrequentlyobservedin theelderly,oldersubjectsexhibit a varietyof impairments thatwould notbeanticipatedasa consequenceof dysfunctionrestrictedto themedialtemporallobe.Many ofthe most prominentand consistentsignsof age-relatedcognitivedeclinein-


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

steadoccur in the information-processingcapacitiestraditionally associatedwith the prefrontal cortex (for recent reviews, seeRapp & Heindel 1994,Shimamura1994). For example,normal elderly individuals have difficultyrememberingthesourceof acquiredinformation(Janowskyetal 1989b,McIn-tyre& Craik1987, Naveh-Benjamin &Craik1995),evenundercircumstanceswhere explicit recollectionof target items is relatively intact (Dywan et al1994, Glisky etal 1995).Sourcememory deficitscanalsopredictperformanceon othertestsof frontal lobefunction,which suggeststhata commonneurobi-ological basismay underlietheseimpairments(Craik et al 1990,Glisky et al1995;but see Spencer& Raz 1994). Afurtherparallel with the effectsof frankfrontal lobe damageis that memoryfor temporalorder appearsparticularlysusceptibleto declineashumanbeingsage(Daigneault& Braun1993,Parkinet al 1995).

It is noteworthythat certainimpairmentsassociatedwith the function ofprefrontalcortexemergerelatively early in the life span,during middle-age,andareunrelatedto thestatusof encodingandretrievalprocessesthatsupportnormal declarativememory. Moreover, when age-relatedimpairments thatresemblebothmedialtemporalandprefrontaldysfunctioncoexistin thesamepersons,thesecognitive deficitsmaybesomewhatdissociable.A particularlyinterestingreport relevantto this point studied healthyagedindividuals be-tween the agesof 65 and 87 years(Glisky et al 1995). Two factors wereobtainedin a factor analysisof the neuropsychologicaltestdata.Teststradi-tionally viewed as assessingthe statusof prefrontalcortex (e.g. WisconsinCard Sorting) loaded onto one factor,whereasassessmentsof declarativememorysensitive to medial temporallobe status(e.g. pairedassociates,de-layedcuedrecall)loadedstronglyontoa secondfactor.A subsequentstudyofitem and sourcememoryshoweda doubledissociation amongtheseelderlyindividualsthatcorrespondedwith their relativefunctioning on medialtempo-ral lobe andprefrontal assessments,respectively. In addition to suggesting thatthereis not necessarilyanobligatoryrelationshipin theeffectsof agingacrossdifferent information processingdomains, thesedatasuggestthat theunderly-ing biological alterations that causedecline in medial temporal lobe-andprefrontal cortex–dependentfunctionsmay occur somewhatindependently.

In agreementwith theneuropsychological assessments of agedhumanbe-ings notedabove,assessmentof the neurobiologicalstatusof the prefrontalcortexindicatesits susceptibility to age-relateddecline.Cortical atrophydur-ing normal humanagingis especially pronouncedin thefrontal lobe,progress-ing at a rate greatlyexceeding atrophyobserved in the cerebralhemispheres asa whole (Coffey et al 1992).Measurementsof regionalcerebralblood flowundera varietyof testingconditionsalsoprovidean indicationof diminishedfunction. Oneof the principal findings to emergefrom this approachis that


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

task demandssufficient to produceprefrontal cortical activation in youngsubjectsfail to increase activityin this sameregion inolder adults(Grady et al1995).Neuroimaginghasalsolocalizeddifferentpatternsof activation coinci-dent with efforts to retrieveinformation, relative to activity inducedby theconsciousrecollectionof targetitems(Schacteret al 1996).In youngsubjects,retrievalefforts areaccompaniedby significant activationin anterioraspectsof the frontal lobe, but a more posteriorfrontal region is activatedin agedsubjectsunderthesametestingconditions.No agedifference,in contrast,wasobservedduring the successfulrecollectionof targetinformation, which pre-dominantly engagesmedialtemporallobestructures.The interestingimplica-tion of theseresultsis that whenrecollectionis not readily achieved,youngandagedsubjectsmay usedifferent retrievalstrategies,mediatedby distinctprefrontalprocessingsystems.Independentof the validity of this particularhypothesis, itis evident thatabnormalitiesin theactivationof prefrontalcortexoccur in relationto cognitive aging.

PsychologicalFunctionsof FrontalLobe Systemsin NormalAging: AnimalModels

The developmentof a nonhumanprimatemodelof normal cognitiveaging hasrevealeda numberof interesting parallelswith findings in humanbeings.Deficits on delayedresponsetestsof short-termmemoryareamongthemostconspicuousand well characterizedsigns of behavioraldecline in the agedmonkey(Bartuset al 1978, Dean& Bartus1988). In the standarddelayedresponsetask, a reward,placedin one of two locations,is retrievedby themonkeyafter avaryingdelay.There iscompelling evidence,however, thatthedelayedresponsedeficit is not necessarilysymptomatic of a generalmemoryimpairmentof the type that resultsfrom damageto themedialtemporallobe.For example,agedmonkeyswith pronounceddelayedresponsedeficits oftenperformnormally on standardtestsof recognitionmemory(i.e. delayednon-match-to-sample) and on a variety of other proceduresthat are sensitive tomedialtemporallobe lesions(Bachevalieret al 1991,Rapp& Amaral1989).In addition, delayedresponseimpairmentsemergerelatively early in the lifespan,precedingthe decline in memory abilities that require the functionalintegrity of themedialtemporallobe(Bachevalieret al 1991).Consistentwithconclusionsfrom humanresearchdiscussedin the prior section(Glisky et al1995),thesefindingsemphasizethatageeffectsarenot uniform acrossdiffer-ent information-processingdomains,andthat medial temporallobe dysfunc-tion alonemay fail to accountfor certainkey featuresof cognitiveaging innonhumanprimates(for recent reviews,see Dean & Bartus1988,Rapp 1995).

A numberof thebehavioralimpairmentsobservedin agedmonkeysappearto reflect a declinein memory-relatedprocessesmediatedby the prefrontal


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

cortex(Dean& Bartus1988,Rapp1995).In this context,a noteworthyaspectof standarddelayedresponsetestingis that it makessubstantial demandsonmemoryfor temporalorder.This is a consequenceof theproceduralarrange-mentin which a rewardis hiddenrandomly, acrosstrials, amonga relativelysmall numberof possible locations.Accurateperformancethereforerequiresmemoryfor the locationbaitedmost recently,andthe ability to discriminatethecurrenttrial from information presentedearlierin testing. Standarddelayedresponse testingalso incorporatesan explicitspatialcomponentthat is thoughtto specificallyengageprocessingfunctionsof the dorsolateralprefrontalcor-tex(Wilsonetal 1993).Consistentwith theview thatdelayedresponsedeficitsreflectprefrontalcorticaldecline,agedmonkeysexhibit deficitson other tasksas a function of demandson temporalordering(Rapp& Amaral 1989). In-creasedperseverationis alsoobservedin agednonhumanprimates(Andersonet al 1993,Bartuset al 1979),similar to effectsseenin agedhumanbeings(Janowskyet al 1989b),and qualitatively resemblingthe difficulties youngsubjectswith frontal lobedamagedisplayin modifying behaviorundercondi-tionsof shifting taskcontingencies(Janowskyet al1989a).

A unifiedperspectiveon thefunctional organizationof theprefrontalcortexthataccommodatesresultsfrom bothratsandprimateshasyet to beachieved.Nonetheless,studiesof agedratshavenoteda numberof qualitative similari-tieswith thebehavioraleffects offrontal lobedamagein youngadultrats.In adirect comparisonof this type,Winocur (1992)evaluateddelayednonmatch-to-sampleperformancein young and agedgroups,and in young rats withlesionsof either the prefrontal cortex or dorsal hippocampus.The samplestimulus in this operantprocedureconsistedof a panellight that was illumi-natedat one of two intensities. During the recognitionphaseof eachtrial,rewardwascontingenton the rat’s committing orwithholdinga leverresponse(i.e. “go,” “no-go”) dependingon whethera matchingor nonmatching lightintensitywaspresented.Similar to delayedresponsetestingin monkeys,theopportunity for intertrial interference issubstantial in this procedure, andsuccessfulperformancerequiresanimalsto distinguish betweenthe currentsampleandthesameitems presentedin earliertrials. Agedrats,andyoungratswith prefrontalcortical lesions,displayed substantial acquisitiondeficits underconditionswhereno delaywasimposedbetweenthesamplepresentationandrecognitiontest.In contrast,youngratswith hippocampal lesionsacquiredthetaskat a normalrate.Thesefindings broadlyparallelresultsin monkeysandhumanbeings,consistent withthe view that the temporal organizationofmemory is significantly disrupted intheagedrat.Qualitativesimilaritiesin theeffectsof aginganddirectprefrontalcorticaldamagehavealsobeennotedinstudiesusingotherbehavioraltestingprocedures(Winocur1991,Zyzaket al1995).


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

Neurobiology ofAging in Frontal LobeSystems

Comparedwith the researchon the medial temporallobe systemreviewedabove,only limi tedexperimentalattentionhasfocusedon defining theneuro-biological consequencesof prefrontal cortical aging. Recentneuroimaging,however,hasrevealedthatmetabolic activity in themonkeyprefrontalcortexdeclineswith age,and interestingly, that variability amongagedsubjectsissubstantially greaterthanamongyounganimals(Eberlinget al 1995).On thisbasis,it is temptingto speculatethat individual differencesin metabolicactiv-ity might predictthe statusof cognitiveprocessesmediatedby the prefrontalcortex.Preliminaryfindingsfrom a studycombining functionalneuroimaginganddelayedresponseassessmentin thesame subjectssuggestthatthis maybethe case (Robertset al1996).

Changesin the structural integrity of the prefrontal cortex are currentlyunderexamination asa possible basisfor age-relatedcognitivedecline.Con-sistentwith a growing body of evidenceindicating that neuronnumber isgenerallypreservedduring normalagingin the medialtemporallobe corticalstructures(Rapp& Gallagher1996,Rasmussenetal 1996, West1993,Westetal 1993;andseesectionon “Neurobiologyof Aging in the Medial-TemporalLobe”), Peterset al (1994)failed to observeanyage-relateddeclinein neurondensity in the dorsolateralprefrontal cortex. A subjective scoring of whitematterpathology in the samesubjects,however,revealedprominentageef-fects, with the greatestdegreeof changeapparentlyobservedamongagedmonkeysthat weremost impairedon standardtestsof learningandmemory.Subtle age-relatedalterationsin other morphological parametershave alsobeennoted, including adecline in the dendritic arborizationof prefrontalcorticalneurons(Cupp& Uemura1980). Providingan additionalparallel withstudiesof themedialtemporallobe,structuralfeaturesin theprefrontalcortexarerelatively intact,with possiblygreateralterationsin neuropilandconnec-tivity as opposedto frank neurodegenerationof cortical neurons.

Comparedwith the relatively preservedstructural featuresof prefrontalcortex,considerable evidence pointsto asubstantial impact ofageonsubcorti-cal systemsthatprojectto cortex.In additionto thecholinergicneuronsin thebasalforebrain system, several collectionsof monoamineneuronsin thebrain-stemappearto undergosignificantdegenerationand/oratrophyduring aging(DeKeyseret al 1990,Irwin et al 1994).In youngsubjects,systemicpharma-cologicalmanipulations of noradrenergicanddopaminergic function signifi-cantly influencespatialandtemporalaspectsof memory,andat leastsomeoftheseeffectsappearto bemediatedat thelevel of corticalprojectiontargetsinthe frontal lobe (Murphy et al 1996).Thesenormative findings, then,leadtotheexpectationthatage-relatedalterationsin neurochemicallydefinedsubcor-


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

tical projection systemsmight significantly disrupt information processingfunctionsdependenton theprefrontal cortex.

Concerningthe function of prefrontalcortex,considerableevidenceindi-catesthat age-relatedalterationsin its dopaminergic innervationmay be par-ticularly important. This interpretationis consistentwith electrophysiologicalresultsdemonstrating that applicationof dopaminereceptorantagonistscanmodulatethe memory-relatedfiring propertiesof single prefrontal corticalneurons(Will iams& Goldman-Rakic 1995).Researchaddressingthe neuro-chemistryof agingin themonkeyindicatesthatendogenousdopamineconcen-trationsaremarkedlyreducedin the prefrontalcortexandthat this declineissubstantially greaterthan that observedin other cortical regions(Goldman-Rakic & Brown 1981, Wenk et al 1989). In addition, Luine et al (1990)observedthat during aging in the rat, a dopaminedeficiency in the frontalcortexwassignificantly correlatedwith impairedworking memoryperform-anceon a radial maze.Although dopamineagonistadministrationin youngsubjectscanaffecta variety of behavioraldomains including motor function,effectsof dopaminergicagentsin agedmonkeysareselectivelyattenuatedontasksthat requirethe functionalintegrity of theprefrontalcortex,suchasthestandarddelayedresponsetask(Arnstenetal 1995).Combinedwith asubstan-tial body of earlierresearch(reviewedin Arnsten1993),thesefindings raisethepossibility thatalterationsin subcorticalsystems,suchasthemesocorticaldopaminergicneurons,might contributeto certainaspectsof cognitive agingby disrupting theinformationprocessingfunctionsof corticaltargetregionsinthe frontal lobe.This areaof researchalsosupportsthebroadertheme,devel-opedthroughout this review, that cortical and subcorticalbrain systemsaredifferentially sensitive to the neurobiological consequencesof normalaging.The neurodegenerationoften associatedwith brain agingappearsto be morecharacteristicof certainsubcorticalsystemsthat innervateforebrainstructuresthan ofcortical neuronsthemselves.

AGING AND ATTENTION IN HUMAN BEINGSANDANIMAL MODELS

As notedin the preceding sections,certain effects ofaging on cognitionresemble,in mild form, damageto systemsin the forebrain, including themedial temporal lobe and prefrontal cortex, for which a substantial back-ground of neuropsychological researchexists.Furthermore,neurobiologicalstudiesarebeginningto providean understandingof alterationsin the brainthat are most likely to serveas a basis for cognitive decline in functionsassociatedwith thosesystems.Alongside theseareasof research,interestinthe studyof attentionin aginghas grownin recent years.


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

Attention refersto multiple componentfunctionsthat areimportantin theselectionand processingof information. The study of attention in aging iscurrentlybenefitingfrom advancesin cognitiveneurosciencethatareprovid-ing a betterdefinition of the neuralsystemsthat are critical for the normalregulationof attention.For example,theseincludesystemsthatregulateover-all levels of sustainedattention (arousalor vigilance) and systemsthat areimportantfor the selective processing ofinformation among competing inputs.Sustainedattentioncanbe assessedin settingsthat requireperformanceof asimple taskwithout the subjectlosing track of the taskobjective, a functionthat appearsto be little affectedby aging(Albert & Moss1996).In contrast,other evidencepoints to an effect of aging on the selectiveprocessingofinformation,particularlyunderconditionsof competition amongmanyitemsfor processingresources(Greenwoodet al 1993, Mouloua & Parasuraman1995).

As recountedin thechapterin thisvolumeon“CentralCholinergicSystemsandCognition” (Everitt & Robbins 1997),a role in theregulationof attentionmayrepresenttheprimary functionof thebasalforebraincholinergic neuronsthat innervate cortex. Furthermore, a growing consensus now views theneurodegenerationwithin this systemthat occurs in aging, and to a moreseveredegreein Alzheimer’s disease,as providing a basis for deficits inattentionratherthanunderlyinga declinein memoryprocesses(Parasuraman& Haxby 1993).

The cholinergic neuronsin the basal forebrain that provide widespreadinnervationof the cortex in rats, monkeys, and humanbeingsare locatedposterior to the cholinergic neuronsin the basal forebrain thattarget thehippocampalformation(Koliatsos et al 1990).Previousstudieshaverevealeddeficits in attention asa consequenceof lesioning the areaof the basalfore-brain that suppliescortical cholinergicinnervation.In onewell-studied para-digm, suchlesionsinterferewith theability of ratsto detectandrespondto abriefly presentedtargetstimulus that can appearin any of severallocations(five-choicereactiontime task)(Muir et al 1994,Robbinset al 1989).Thoselesionsdecreasetheaccuracyof performance,aneffect that canbeovercomeby increasingthetargetduration,which suggeststhat the impairmentis atten-tional in nature.Impairmentsin theability of agedratsto detecttargetsin thefive-choice reaction time task that resemblethe effects of basal forebrainlesionsin youngratshaverecentlybeenreported(Joneset al 1995).Anotherparadigm,a spatialcueingtaskoriginally designedfor studiesof attentioninhumanbeings,hasalsoshownsensitivity to lesionsof the basalforebraininmonkeys(Voytko etal 1994).Althoughthe lesionmethods used inthis line ofresearchwith laboratory animals, untilrecently, havebeenrelativelynonselec-tive, removingbothcholinergicandnoncholinergic neuronsin thebasalfore-


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

brain, studiesusing a selectiveimmunotoxin for cholinergic neuronshavesuccessfullyproduceddeficits in attentionwhenthe cortical cholinergicpro-jectionsare removed (Chibaetal 1995a,b).

The role of thebasalforebraincholinergic systemin attentionmay extendto thecomponentof this systemthatprovidesinnervationof thehippocampalformation. As noted in an earlier sectionof this review, removal of thoseneuronswith the selectiveimmunotoxin fails to reproducedeficits in spatialtasksthatarereadilyobservedin agedrats(Baxteret al 1995a).In thechapterby Everitt& Robbins,theeffects oflessselectivelesionsare citedas evidencethat the septohippocampalcholinergicsystemplaysa role in memory.How-ever, whetherany substantial deficit in memory is observedwith selectiveremovalof thesecholinergicneuronshasyet to be demonstrated.In contrast,youngratswith selectiveimmunotoxic lesionsof thecholinergicneuronsthatprojectto thehippocampalformationdohaveamarkedimpairmentin ataskinwhich attentional processingis modified in intact young rats (Baxter et al1995b).The taskinvolvesrepeated exposureto acuethat is subsequentlyusedas a conditioned stimulus in associativelearning. Preexposureto the cueusuallyretardssubsequentlearning,a phenomenonreferred toaslatentinhibi-tion. Although more than one psychologicalexplanationof latent inhibitionhas beenoffered, a numberof explanationsconvergeon the conceptthatdecrementsin attention to, or processingof, the preexposedcue serveas abasis for latent inhibition. In this context it is notablethat either selectivedamageto the hippocampus (Han et al 1995) or selectiveremoval of thecholinergic projection to the hippocampus impairs latent inhibition in rats(Baxteret al 1995b).Theconceptthata latentinhibition deficit might exist inagedratsbecauseof diminishedfunction of the cholinergicprojectionto thehippocampus hasnotyetbeentesteddirectly.It is interesting to note,however,thata recentstudyshowedthatinformation encodingof neuronsin hippocam-pus in young ratswill becomeunresponsive tocues that are not reliablefeaturesof a spatialenvironment,aneffectnot seento thesameextentin agedrats(Tanila et al 1996).Thus,apparentlythe selectionof informationthat issubjectto processingandencodingby hippocampal neuronsis alteredduringaging in a manner that might be predictedfrom the effects of removingcholinergicneuronsin youngrats.

CONCLUSION

As informationaccumulatesaboutthealterationsthatoccurduringagingin thebrain,it becomesincreasingly clearthatanumberof differenttypesof changescanbeidentifiedin differentneuralsystems.Moreover,theseverityof age-re-lated changesin particularbrain systemsoften coincideswith the extentofdeclinein cognitive functionsassociatedwith thosesystems.Certainevidence


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

has also indicated that heterogeneityin the effects of aging may exist indifferent cognitive domainsand neurobiological systems. All theselines ofevidencesuggestthat aging is not a global process,a conclusionthat canbeapplied to the information derived from animal modelsas well as humanstudies.

In thosecaseswherecomparisonscan be madeacrossstudiesof humanbeingsandlaboratoryanimals, an importantinsight into the neurobiologyofaging is emerging.Researchusingadvancedstereologicalmethods indicatesthat neuronlossis not characteristicof cortical systems, including thehippo-campus,but thatneurodegenerationdoesaffectdistinctpopulationsof subcor-tical neuronsthat providecortical innervation. Apartfrom such structuralfeaturesof the brain, other aspectsof functional integrity are also affectedduringaging.Althoughmany ofthedetailedanalysesavailablefrom studiesofanimal modelshaveyet to be extendedto studiesof humanbrains,neuro-imagingresearchprovidessupportfor the conceptthat processingwithin theexistingcircuitsof the braincan be compromisedduringaging.

The comparisonof functional analysesacrossspecies,including neuro-imaging researchwith humanbeings,highlights the needto advancenewmodelsfor understandingthe neurobiologicalbasisof cognitive alterationsthat occur late in life. Aging differs from many conditionsinvolving braindamage,which neuropsychological studieswere originally intendedto ad-dress.Animal models for thoseconditions frequentlyentail thevirtual destruc-tion of a brain systemto test hypothesesaboutthe underlyingsubstrateforcognitive functionsof interest,e.g. the medial temporallobe anddeclarativememory.During normal aging, in contrast,substantial structuralintegrity ispreservedover the entire life span,and neuronsthat do exhibit appreciableneurodegeneration,suchas thosein the basalforebrain cholinergicsystem,areby no meanseliminated.For this reason,lesion modelsmay have limi tedutility for capturingthe performanceof neuralsystemsin the agedbrain, inwhich considerableremodelingoccursanda variety of functionalalterationswithin theexistingsystems can be detected(Gallagher et al1994).

Finally, the themeof individual differencesin aging is well supportedbystudiesof laboratoryanimals,including behavioralmodelsdevelopedfor theirsensitivity to memoryfunctionssubservedby themedialtemporallobe.More-over,individual differencesin behavioralcapacitieswithin thesemodelsoftencorrelatewith the severityof neurobiological alterationsin the relevantbrainsystems.Theselinesof researchgive credenceto theconceptthatage-associ-atedmemorydeclinein humanbeingscanreflect a normalagingprocess,asdistinct from a preclinicalconditionthat heraldsdementia.An understandingof the basisfor individual differencesin the effectsof aging is likely to beadvancedby further studies on theneurobiologyof agingusinganimalmodels


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

asanimportantadjunctto thestudyof humanbeings.In this endeavor,it willbe particularly importantto determinethe factorsthat distinguishthoseagedindividuals that maintain preservedfunction from thosethat experiencede-cline, an undertakingthat will benefit from the bridgesthat can be formedbetweenhumanbeingsandthestudyof agingin well-developedanimalmod-els.

ACKNOWLEDGMENTS

The authorswish to acknowledgeLisa Brooksfor preparationof the figures,and grantsAG09973and KO5 MH 01149to MG, and AG10606to PR forsupportof thework.

LiteratureCited

Albert MS,Moss MB. 1996. Neuropsychologyof aging: findings in humans and monkeys.In Handbook of the Biologyof Aging, ed.EScheider, J Rowe, 4:217–30. San Diego:Academic

Anderson JR, Anthouard M, de Monte M,KempfJ.1993. Differencesin performanceof young andold monkeys on a visuospa-tial memory task. Q. J. Exp.Psychol. 46B:391–98

ArnstenAF. 1993. Catecholamine mechanismsin age-relatedcognitive decline.Neurobiol.Aging14:639–41

Arnsten AF, Cai JX, Steere JC, Goldman-Rakic PS. 1995.Dopamine D2 receptormechanismscontribute to age-relatedcog-nitive decline: the effectsof quinpirole onmemory and motor performancein mon-keys.J. Neurosci.15:3429–39

Bachevalier J, Landis LS, Walker LC, Brick-sonM, Mishkin M, et al. 1991. Agedmon-keys exhibit deficits indicative of wide-spread cerebral dysfunction. Neurobiol.Aging12:99–111

BarnesCA. 1979. Memory deficits associatedwith senescence: neurophysiological andbehavioral study in the rat. J. Comp.Physiol. Psychol. 93:74–104

Barnes CA, McNaughton BL. 1985. An age-comparisonof the ratesof acquisition andforgetting of spatial information in relationto long-term enhancementof hippocampalsynapses.Behav.Neurosci.99:1040–48

Barnes CA, McNaughton BL, O’Keefe J.1983. Loss of place specifici ty in hippo-campal complex spike cells of senescentrat.Neurobiol. Aging8:521–45

Barnes CA, Treves A, Rao G, ShenJ. 1994.Electrophysiological markers of cognitiveaging: region specifici ty andcomputationalconsequences.Semin. Neurosci.6:359–67

BartusRT, DeanRL, BeerB, LippaAS. 1982.

The cholinergic hypothesis of geriatricmemory dysfunction. Science217:408–18

Bartus RT, DeanRL, Fleming DL. 1979. Ag-ing in the rhesus monkey: effectson visualdiscrimination learning andreversallearn-ing. J. Gerontol. 34:209–19

BartusRT, FlemingD, JohnsonHR. 1978. Ag-ing in the rhesus monkey: debil itating ef-fects on short-term memory. J. Gerontol.33:858–71

Baxter MG, Bucci DJ, Gorman LK, Wiley R,Gall agher M. 1995a.Selective immuno-toxic lesions of basalforebrain cholinergiccells: effects on learning and memory inrats.Behav. Neurosci.109:714–22

Baxter MG, Bucci DJ, Sobel TJ,Will iamsMJ,Gorman LK, Gallagher M. 1996. Intactspatial learning following lesions of basalforebrain cholinergic neurons. NeuroRe-port 7:1417–20

Baxter MG, GallagherM. 1996. Intact spatiallearning in both young and agedrats fol-lowing selective removal of hippocampalcholinergic input. Behav. Neurosci. 110:460–67

Baxter MG, Gallagher M, Holland PC.1995b.Disruption of decremental attentional proc-essingby selective removal of hippocam-pal cholinergic input. Soc. Neurosci.Abstr.21:935

Berger-SweeneyJ,HeckersS,MesulamM-M,Wiley RG, Lappi DA, Sharma M. 1994.Dif ferential effectson spatial navigation ofimmunotoxin-induced cholinergic lesionsof the medial septal area and nucleusbasalis magnocellularis. J. Neurosci. 14:4507–19

Bondi MW, Salmon DP,MonschAU, GalaskoD, Butters N, et al.1995. Episodic memorychanges are associated with the APOE-4allele innondemented olderadults.Neurol-ogy 45:2203–6


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

BrizzeeKR, OrdyJM, BartusRT.1980. Local-ization of cellular changeswithin multimo-dal sensoryregions in agedmonkey brain:implications for age-relatedcognitive loss.Neurobiol. Aging1:45–52

Buckner RL, PetersenSE. 1996. What doesneuroimaging tell us about the role of pre-frontal cortex in memory retrieval? Semin.Neurosci.8:47–55

BunseyM, EichenbaumH. 1996. Conservationof hippocampal memory function in ratsandhumans.Nature379:255–57

Chiba AA, Bucci DJ, Holland PC, GallagherM. 1995b. Basal forebrain cholinergic le-sions disrupt increments but not decre-ments in conditioned stimulus processing.J. Neurosci.15:7315–22

Chiba AA, Bushnell PJ, Oshiro WM, Gal-lagher M. 1995a. Altered selective atten-tion in ratswith cholinergic lesions of thesubstantia innominata. Soc. Neurosci.Abstr. 21:936

Chouinard ML, Gallagher M, Yasuda RP,Wolfe BB, McKinney M. 1995. Hippo-campal muscarinic receptor function inspatial learning-impairedagedrats.Neuro-biol. Aging16:955–63

Clark AS, Magnusson KR, Cotman CW. 1992.In vitro autoradiography of hippocampalexcitatory amino acid binding in agedFis-cher 344 rats: relationship to performanceon the Morris watermaze.Behav. Neuro-sci.106:324–35

Coffey CE, Wilkinson WE, Parashos IA,Soady SAR,SullivanRJ,etal. 1992. Quan-titative cerebralanatomy of the aging hu-man brain: a cross-sectional study usingmagnetic resonance imaging. Neurology42:527–36

Cohen NJ, Squire LR. 1980. Preservedlearn-ing andretention of pattern-analyzing skillin amnesia: dissociation of “knowing how”and “ knowing that.” Science 210:207–9

Corder EH, Saunders AM, Strittmatter WJ,Schmechel DE, Gaskell PC, et al. 1993.Gene doseof apolipoprotein E type 4alleleandthe risk of Alzheimer’s diseasein lateonsetfamilies.Science261:921–23

Corkin S.1984. Lasting consequencesof bilat-eral medial temporal lobectomy: clinicalcourseandexperimentalfindings in H. M.J. Neurosci.2:1214–29

Coyle JT, Price DL, DeLong MR. 1983.Alzheimer’s disease:a disorder of corticalcholinergic innervation. Science219:1184–90

Craik FIM, Jennings JM. 1992. Humanmem-ory. In Handbook of Aging and Cognition,ed. FIM Craik, TA Salthouse, pp. 51–83.Hil lsdale,NJ:Erlbaum

Craik FIM, Morris LW, Morris RG, LoewenER.1990. Relationsbetweensourceamne-

sia and frontal lobe functioning in olderadults.Psychol. Aging5:148–51

Crook T, BartusRT, Ferris SH,WhitehouseP,CohenGD, GershonS. 1986. Age-associ-ated memory impairment: proposeddiag-nostic criteria and measures of cli nicalchange: report of a NIMH work group.Dev.Neuropsychol. 2:261–76

CrystalH, DicksonD, Fuld P, Masur D, ScottR, etal. 1988. Clinico-pathologic studiesindementia: nondemented subjects withpathologically confirmed Alzheimer’s dis-ease.Neurology 38:1682–87

Cupp CJ, Uemura E. 1980. Age-relatedchangesin prefrontal cortexof Macacamu-latta: quantitative analysis of dendriticbranching patterns. Exp. Neurol. 69:143–69

Daigneault S, Braun CM. 1993. Workingmemory and the self-orderedpointing task:further evidenceof earlyprefrontal declinein normal aging. J. Clin. Exp. Neuropsy-chol. 15:881–95

DamAM. 1979. Thedensity of neurons in thehuman hippocampus. Neuropathol. Appl.Neurobiol. 5:249–64

Davies P, Maloney AJF. 1976. Selective lossof central cholinergic neurons in Alzhe-imer’s disease.Lancet2:1403

Davis KL, Thal LJ, Gamzu ER, Davis CS,Woolson RF, et al. 1992. A double-blind,placebo-controlled multicenter study oftacrine for Alzheimer’s disease.NewEngl.J. Med.327:1253–59

DeanRL, BartusRT. 1988. Behavioralmodelsof aging in nonhuman primates.In Hand-book of Psychopharmacology, ed. LLIversen, SD Iversen, SH Snyder, 20:325–92. NewYork: Plenum

DeKeyserJ, Ebinger G, Vanquelin G. 1990.Age-relatedchanges in the human nigros-triatal dopaminergic system.Ann. Neurol.27:157–61

deLeonMJ, Golomb J, George AE, Convit A,Tarshish CY, et al. 1993. The radiologicprediction of Alzheimer Disease:the atro-phic hippocampal formation. Am. J.Neuroradiol. 14:897–906

Dywan J, Segalowitz SJ,Will iamsonL. 1994.Sourcemonitoring during name recogni-tion in older adults: psychometric and elec-trophysiological correlates.Psychol. Aging9:568–77

Eberling JL, RobertsJA, De Manincor DJ,Brennan KM, Hanrahan SM, et al. 1995.PET studies of cerebral glucose metabo-lismin conscious rhesus macaques.Neuro-biol. Aging 16:825–32

EichenbaumH, Otto T, Cohen NJ. 1994. Twofunctional components of the hippocampalmemory system. Behav. Brain Sci. 17:449–518

Everitt BJ, Robbins TW. 1997. Central cholin-


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

ergic systems and cognition. Annu. Rev.Psychol. 48:649–84

FischerW, GageF,BjörklundA. 1989. Degen-erativechangesin forebrain cholinergic nu-clei correlatewith cognitive impairmentsinagedrats.Eur. J. Neurosci.1:34–45

Foster TC, Barnes CA, Rao G, McNaughtonBL. 1991. Increasein perforant pathquan-tal sizein agedF-344 rats.Neurobiol. Ag-ing 12:441–48

Gage FH, Dunnett SB, Björklund A. 1989.Age-relatedimpairments in spatial memoryare independent of those in sensorimotorskills. Neurobiol. Aging 10:347–52

Gage FH, Kelly P, Björklund A. 1984. Re-gional changes in brain glucose metabo-lism reflect cognitive impairments in agedrats.J. Neurosci.4:2856–65

Gallagher M, Burwell R, Burchinal M. 1993.Severity of spatial learning impairment inaging: development of a learning indexforperformancein the Morris water maze.Be-hav. Neurosci.107:618–26

Gallagher M, Colombo PJ. 1995. Aging: thecholinergic hypothesisof cognitivedecline.Curr. Opin. Neurobiol. 5:161–68

Gallagher M, Gill TM, Baxter MG, Bucci DJ.1994. The development of neurobiologicalmodels for cognitive decline in aging.Semin. Neurosci.6:351–58

Gallagher M, Nagahara AH, Burwell RD.1995. Cognition andhippocampal systemsin aging: animal models. In Brain andMemory: Modulation and Mediation ofNeuroplasticity, ed.JL McGaugh, N Wein-berger, G Lynch, pp. 103–26. New York:Oxford Univ. Press

GazzaleyAH, Siegel SJ,Kordower JH, Muf-sonEJ,MorrisonJH.1996. Circuit-specificalterations of N-methyl-D-aspartate recep-tor subunit 1 in the dentate gyrus of agedmonkeys. Proc. Natl. Acad. Sci. USA 93:3121–25

GeddesJW,MonaghanDR, CotmanCW, LottIT, Kim RC, Chui HC. 1992. Plasticity ofhippocampalcircuitry in Alzheimer’s dis-ease.Science230:1179–81

GeinismanY, deToledo-Morrell L, Morrell F.1986. Loss of perforated synapses in thedentate gyrus: morphological substrate ofmemory deficit in aged rats. Proc. Natl.Acad. Sci. USA83:3027–31

GeinismanY, deToledo-Morrell L, Morrell F,Persina IS, RossiM. 1992. Age-relatedlossof axcospinous synapses formedby two af-ferent systemsin the rat dentate gyrus asrevealedby the unbiasedstereological dis-sectortechnique.Hippocampus2:437–44

Glisky EL, PolsterMR, RouthieauxBC. 1995.Double dissociation between item andsource memory. Neuropsychology 9:229–35

Goldman-Rakic PS, Brown RM. 1981. Re-

gional changesof monoaminesin cerebralcortex and subcortical structures of agingrhesus monkeys.Neuroscience6:177–87

Golomb J,de LeonMJ, Kluger A, George AE,TarshishC, Ferris SH. 1993. Hippocampalatrophy in normal aging: an associationwith recent memory impairment. Arch.Neurol. 50:967–73

GradyCL, McIntosh AR, Horwitz B, MaisogJ, Ungerleider L, et al. 1995. Age-relatedreductions in human recognition memorydue to impaired encoding. Science 269:218–21

Greenwood P, Parasuraman R, Haxby JV.1993. Visuospatial attention across theadult li fe span. Neuropsychologia 31:471–85

HanJ-S, Gallagher M, Holland PC.1995. Hip-pocampal lesions disrupt decrements butnot increments in condit ioned stimulusprocessing. J. Neurosci.15:7323–29

Hel kala EL , K oi vi sto K , Hänninen T,Vanhanen M, Kervinen K, et al. 1996.Memory functions in humansubjectswithdifferent apolipoprotein E phenotypesdur-ing a 3-year population-based follow-upstudy. Neurosci.Lett. 204:177–80

Irwin I, DeLanney LE, McNeill T, Chan P,Forno LS, et al. 1994. Aging and the ni-grostriatal dopamine system:a nonhumanprimate study. Neurodegeneration 3:251–65

IssaAM, Rowe W, Gauthier S, MeaneyMJ.1990. Hypothalamic-pituitary-adrenal ac-tivity in aged, cognitively impaired andcognitively unimpaired rats. J. Neurosci.10:3247–54

Janowsky JS,Shimamura AP, Kritchevsky M,Squire LR. 1989a. Cognitive impairmentfollowing frontal lobedamageandits rele-vanceto humanamnesia.Behav. Neurosci.103:548–60

Janowsky JS, Shimamura AP, Squire LR.1989b. Source memory impairment in pa-tients with frontal lobe lesions. Neuropsy-chologia 27:1043–56

JonesDNC, Barnes JC, Kirkby DL, HigginsGA. 1995. Age-associatedimpairments ina test of attention: evidence for involve-ment of cholinergic systems.J. Neurosci.15:7282–92

Kil lianyRJ,MossMB, RoseneDL, HerndonJ,Lai ZC. 1995. Age-relatedchanges in therhesus monkey: memory executive func-tion and “ IQ” in a nonhuman primatemodel of normalhumanaging. Soc. Neuro-sci.Abstr. 21:1564

Koliatsos VE, Martin LJ, Price DL. 1990. Ef-ferentorganizationof themammalianbasalforebrain. In Brain Cholinergic Systems,ed. SM Biesold, pp. 120–52. New York:Oxford Univ. Press

Lane F, Snowden J. 1989. Memory and de-


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

mentia: a longitudinal survey of suburbanelderly. In Clinical and Abnormal Psychol-ogy, ed.PLovibond,PWilson, pp. 365–76.New York: Elsevier

Luine V, Bowling D, HearnsM. 1990. Spatialmemorydeficits in agedrats:contributionsof monoaminergic systems. Brain Res.537:271–78

Markus EJ, Barnes CA, McNaughton BL,Gladden VL, Skaggs WE.1994. Spatial in-formation content andreliabili ty of hippo-campalCA1 neurons: effectsof visual in-put. Hippocampus4:410–21

McIntyre JS,Craik FIM. 1987. Age differencein memory for item and source informa-tion. Can. J. Psychol. 41:175–92

Meaney MJ, Aitken DH, van Berkel C, Bhat-nagar S, Sapolsky RM. 1988. Effect ofneonatal handlingon age-related impair-ments associated with the hippocampus.Science239:766–68

Meunier M,BachevalierJ,Mishkin M, MurrayEA. 1993. Effectson visualrecognition ofcombinedandseparate ablations of the en-torhinal and perirhinal cortex in rhesusmonkeys.J. Neurosci.13:5418–32

Mizumori SJY,Lavoie AM, Kalyani A. 1996.Redistribution of spatial representation inthe hippocampusof agedratsperforming aspatial memory task.Behav. Neurosci.110:1006–16

Moore CI, Browning MD, Rose GM. 1993.Hippocampalplasticity inducedby primedburst, but not long-term potentiation,stimulation in areaCA1 of agedFisher 344rats.Hippocampus3:57–66

Morris JC, McKeel DW, Storandt M, RubinEH, Price JL,etal. 1991. Very mild Alzhe-imer’s disease: informant-based clinical,psychometric and pathological distinctionfrom normal aging. Neurology 41:469–78

Morse CK. 1993. Does variabil ity increasewith age?An archival study of cognitivemeasures.Psychol. Aging8:156–64

MoscovitchM, Ulmita C.1991. Consciousandnonconscious aspects of memory: aneuropsychological framework of modulesand central systems. In Perspectives onCognitiveNeuroscience,ed.RG Lister,HJWeingartner, pp. 229–66. New York: Ox-ford Univ. Press

MossMB, Rosene DL, PetersA. 1988. Effectsof aging on visual recognition memory inthe rhesus monkey.Neurobiol. Aging 9:495–502

Mouloua M, ParasuramanR. 1995. Aging andcognitive vigilance: effects of spatial un-certainty and event rate. Exp. Aging Res.21:17–32

Muir JL, Everitt BJ, Robbins TW. 1994.AMPA-inducedexcitotoxic lesions of thebasal forebrain: a significant role for the

cortical cholinergic systemin attentionalfunction. J. Neurosci.14:2313–26

Murphy BL, Arnsten AF, Goldman-Rakic PS,Roth RH. 1996. Increaseddopamine turn-over in the prefrontal cortex impairs spatialworking memory performance in rats andmonkeys. Proc. Natl. Acad. Sci. USA 93:1325–29

Murray EA. 1996. What have ablation studiestold usabout theneural substratesof stimu-lus memory?Semin. Neurosci.8:13–22

NagaharaAH, Otto T, Gallagher M. 1995. En-torhinal/Perirhinal cortex lesions impairperformance on two versions of placelearning in the Morris watermaze.Behav.Neurosci.109:3–9

Naveh-Benjamin M, Craik FIM. 1995. Mem-ory for context andits use initem memory:comparisonsof youngerandolder persons.Psychol. Aging10:284–93

Nicoll e MM, Bizon J, Gallagher M. 1996.Ionotropic glutamatereceptors in the hip-pocampus and striatum of agedrats: rela-tionship to cognitive decli ne. Neurosci-ence.In press

Otto T, EichenbaumH. 1992. Complementaryrolesof theorbital prefrontal cortexandtheperirhinal-entorhinal cortices in an odor-guided delayed-nonmatching-to-sampletask.Behav.Neurosci.106:762–75

Parasuraman R, Haxby JV. 1993. Attentionandbrain function in Alzheimer’s Disease:a review.Neuropsychology 7:242–72

Parkin AJ,Walter BM,Hunkin NM. 1995. Re-lationships between normal aging, frontallobe function,and memory for temporaland spatial information. Neuropsychology9:304–12

Peters A, Leahu D, MossMB, McNally J.1994. Theeffectsof agingonarea46of thefrontal cortexof therhesusmonkey.Cereb.Cortex6:621–35

PetridesM, Ali visatos B, EvansAC, MeyerE.1993a. Dissociation of human mid-dorso-lateral from posterior dorsolateral frontalcortex in memory processing. Proc. Natl.Acad. Sci. USA90:873–77

PetridesM, Ali visatos B, Meyer E, EvansAC.1993b. Functional activation of the humanfrontal cortex during the performance ofverbal working memory tasks.Proc. Natl.Acad. Sci. USA90:878–82

PrestySK, Bachevalier J, Walker LC, StrubleRG, PriceDL, et al. 1987. Age differencesin recognition memoryof the rhesusmon-key (Macacamulatta). Neurobiol. Aging 8:435–40

Quirion R, WilsonA, RoweW, Aubert I, Rich-ard J, et al. 1995. Facilitation of acetyl-choline releaseandcognitive performanceby an M2-Muscarinic receptor antagonistin agedmemory-impairedrats.J. Neurosci.15:1455–62


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

Rapp PR. 1993. Neuropsychological analysisof learning andmemoryin the agednonhu-manprimate.Neurobiol. Aging 14:627–29

Rapp PR. 1995. Cognitive neuroscienceper-spectiveson aging in nonhumanprimates.In Emotion, Memoryand Behavior, ed. TNakajima, T Ono, pp.197–211. Tokyo:Jpn.Sci.Soc.

Rapp PR, Amaral DG. 1989. Evidence fortask-dependent memory dysfunction in theagedmonkey. J. Neurosci.9:3568–76

Rapp PR, Amaral DG. 1991. Recognitionmemorydeficits in a subpopulation of agedmonkeys resemble the effects of medialtemporal lobe damage. Neurobiol. Aging12:481–86

Rapp PR, Gallagher M. 1996. Preservedneu-ron number in the hippocampus of agedrats with spatial learning deficits. Proc.Natl. Acad. Sci.USA. 93:9926–30

Rapp PR,HeindelWC.1994. Memory systemsin normal and pathological aging.Curr.Opin. Neurol. 7:294–98

Rapp PR, KanskyMT, Eichenbaum H. 1996.Learningandmemory for hierarchical rela-tionships in the monkey: effectsof aging.Behav. Neurosci.110:887–97

Rasmussen T, Schliemann T, Sørensen JC,Zimmer J, West M. 1996. Memory im-pairedagedrats:no loss ofprincipalhippo-campaland subicular neurons. Neurobiol.Aging17:143–47

RobbinsTW, Everitt BJ,MarstonHM, Wilkin-son J, Jones GH, Pagae KJ. 1989. Com-parative effectsof ibotenic acid and quis-qualic acid-inducedlesions of the substan-tia innominataonattentional function in therat: further implications for the role of thecholinergic neurons of the nucleus basalisin cognitive processes. Behav. Brain Res.35:221–40

Roberts JA, Eberling JL, Rapp PR,TuszynskiMH, JagustWJ. 1996. Reductions in cere-bral glucose metaboli sm are associatedwith memory deficits in agedrhesusma-caques.Soc. Neurosci.Abstr. In press

Schacter DL, Savage CR, Alpert NM, RauchSL, Albert MS. 1996. The role of the hip-pocampusandfrontal cortex in age-relatedmemorychanges:a PET study. NeuroRe-port. In press

Scovil le WB, Mil ner B. 1957. Loss of recentmemory after bilateral hippocampal le-sions.J. Neurol. Neurosurg. Psychiatry 20:11–21

ShenJ,BarnesCA. 1996. Age-relateddecreasein cholinergic synaptic transmission inthree hippocampal subfields. Neurobiol.Aging17:439–51

Shimamura AP. 1994. Neuropsychologicalperspectiveson memory andcognitive de-cl ine in normal human aging. Semin.Neurosci.6:387–94

Shimamura AP, Janowsky JS, Squire LR.1990. Memory for the temporal order ofevents in patients with frontal lobe lesionsand amnesic patients. Neuropsychologia28:803–13

Smith ML, Booze RM. 1995. Cholinergic andgabaergic neurons in thenucleusbasalis re-gion of young andagedrats.Neuroscience67:679–88

Smith TD, Gallagher M, Lesli e FM. 1995.Cholinergic binding sites in rat brain:analysis by age and cogni tive status.Neurobiol. Aging 16:161–73

Spencer WD, RazN. 1994. Memory for facts,source, and context: can frontal lobe dys-function explain age-related differences?Psychol. Aging 9:149–59

Squire L, Zola-Morgan S. 1991. The medialtemporal lobe memory system. Science253:1380–86

StackEC,Gallagher M, Rapp PR.1995. Reor-ganization of hippocampal circuitry in theagedrat. Soc. Neurosci.Abstr. 21:472

Stenvers KL, Lund PK, GallagherM. 1996.Increasedhippocampalexpressionof type1 insulin-like growth factor (IGF) receptormessengerRNA is associated with cogni-tive decline in agedrats.Neuroscience 72:505–18

Stroessner-Johnson HM, Rapp PR, AmaralDG. 1992. Cholinergic cell lossandhyper-trophy in the medial septalnucleus of thebehavioral ly characterized aged rhesusmonkey. J. Neurosci.12:1936–44

Suzuki WA, Zola-MorganS,Squire LR, Ama-ral DG. 1993. Lesionsof the perirhinal andparahippocampal cortices in the monkeyproduce long-lasting memory impairmentin the visual and tactual modalit ies. J.Neurosci.13:2430–51

Tanila H, Shapiro M, Eichenbaum H. 1996.Hippocampalplacefields in agedratswithspatial memory defici t. Soc. Neurosci.Abstr. 21:943

Thal LJ, Masur DM, Blau AD, Fuld PA,KlauberMR. 1989. Chronic oral physostig-mine without lecithin improvesmemory inAlzheimer’s disease.J. Am. Geriat. Soc.37:42–48

TorresEM, PerryTA, Blokland A, WilkinsonLS, Wiley RG, et al. 1994. Behavioural,histochemical and biochemical conse-quencesof selective immunolesions in dis-creteregions of the basalforebrain cholin-ergic system.Neuroscience63:95–122

Undie AS, WangH-Y, FriedmanE. 1995. De-creasedphospholipaseC-β immunoreactiv-ity, phosphoinositide metabolism, andpro-tein kinaseC activation in senescentF-344ratbrain. Neurobiol. Aging 16:19–28

Voytko ML, Olton DS, Richardson RT, Gor-manLK, Tobin JR,Price DL. 1994. Basalforebrain lesions in monkeys disrupt atten-


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

tion but not learning and memory. J.Neurosci.14:167–86

Wang YH, Luo JH, YasudaRP,Gallagher M,Kellar KJ, Wolfe BB. 1996. Age-relatedchangesin NMDA receptor subunits,NR1,NR2A, andNR2B in rat striatumandhip-pocampus.Soc. Neurosci.Abstr. 22

Wenk GL, PierceDJ, Struble RG, Price DL,Cork LC. 1989. Age-related changes inmultiple neurotransmitter systems in themonkeybrain. Neurobiol. Aging 10:11–19

West MJ. 1993.Regionally specific loss ofneurons in the aging humanhippocampus.Neurobiol. Aging14:287–93

West MJ, Amaral DG, Rapp PR. 1993. Pre-servedhippocampal cell number in agedmonkeyswith recognitionmemorydeficits.Soc. Neurosci.Abstr. 19:599

West MJ, ColemanPD, Flood DG, TroncosoJC.1994. Dif ferencesin the patternof hip-pocampal neuronal loss in normal agingand Alzheimer’ s disease. Lancet 344:769–72

Wil l iams GV, Goldman-Rakic PS. 1995.Modulation of memory fieldsby dopamineD1 receptors in prefrontal cortex. Nature376:572–75

Wilner J,Otto T, GallagherM, EichenbaumH.1993. Hippocampal lesions that impairplacelearning facilitatedelayednonmatch-ing performance in rats. Soc. Neurosci.Abstr. 19:358

Wilson FAW, Scalaidhe SP, Goldman-RakicPS.1993. Dissociationof objectandspatialprocessing domains in primate prefrontalcortex.Science260:1955–58

Winocur G. 1991. Conditional learning in agedrats:evidenceof hippocampal andprefron-tal cortex impairment. Neurobiol. Aging13:131–35

Winocur G.1992. A comparison of normal oldratsandyoung adult rats with lesions tothehippocampus or prefrontal cortex on a testof matching-to-sample. Neuropsychologia30:769–81

Youngjohn JR, Crook TH. 1993. Stabili ty ofeveryday memory in age-associatedmem-ory impairment: a longitudinal study.Neuropsychology 7:406–16

Zyzak DR, Otto T, EichenbaumH, GallagherM. 1995. Cognitivedeclineassociated withnormal aging in rats: a neuropsychologicalapproach. Learn. Mem.2:1–16


Ann

u. R

ev. P

sych

ol. 1

997.

48:3

39-3

70. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tex

as -

Dal

las

on 0

8/25

/11.

For

per

sona

l use

onl

y.

the use of animal models to study the effects of aging on cognition

Documents