Series Editor: Gr. eme J Hankey, MO, FRACP

11: Disorders of memory and intellect
John D G Watson

Development of effec tive treatments for dementia is a major challenge
L\IPAIRED IJIo'TELLECTUAL FUKCTlOS has profound effects
on pati ents, th eir famil ies and society, b ut medical and community services are far fro m optimal.
Dementia is mo re commo n with advancing age . In childhood, gene tic disorders acco u nt for a number of co gn itive
problems. In young and middle-aged adults, traumatic brain
injury is a m ajo r ca use; hypoxic/hypot en sive brain damage
(often associat ed with traumatic in jur y) also accounts for a
number of cases. In olde r adults, acquired d em entia is th e
major cause of disordered int ellect . At th e age of 60 year s,
the prevalence is about 1%; 3%--1 1% of all comm uni ty-living
adults old er than 65 years are affected ; by th e age of 85 ,
between 20 % and 50% are affected . M ore tha n 50 cau ses of
demen tia hav e been identified ; the m ost com mo n is
Alzheimer's di sease (AD), followed by vascular d ementias
(VAD) .

The current chall enges are early detection of patient s with
demen tia, ide nti fication of th e causes of d em en tia, and th e
development of major th era peu tic agents. As our po p ulatio n
ages, demen tia will incre asingly account for m edical, social
and finan cial p roblems.
Cognitive functions and neuroanatomy
Within the b rain, cognitive functions may be localised or dis-

tributed. C urrent th inking is th at th e brain has a gre at d eal
of distributed organisation, with many functions more specifically located to one part of th e br ain .
Cognitive ab ilities d epend on on e another, to some extent
in a hierarchical manner. Attention, conce ntra tio n, alertness
and moo d are card inal. People with attentional d isorders
may appear to have m ore specific defects (a false localising
sign), suc h as word-finding difficulties. However, pa tients
without overt art entional problems often pr esent beca use of
concerns about m em ory impairm en t. One can ofte n diagnose the cause as anxiety, d epression, or sleep d ep rivation .
The se are very ame nable to tre atment, especially in comparison with true d em entia .
f.ocallS6(f brain functIons: Most researchers beli eve th at th e
brain has co mplex, overlap ping sets of connections th at
follow both parallel and hiera rch ical organisation. Some
regions in th ese connections are so cr ucial th at a fun cti on is
attributed to th em (ie, th e fun ctions arc relatively localised).
Roy. 1 PrInce Alfred Hospit.ill, Sydn e y, NSW,
John C G W, tl on , OPhil. MBBS. FRACP. D,rector . Neuropsychology
UIlII. and Associate Professor 01 Med 'c,ne, University of Sydney
Repronts will oct be availab le from me author Corr espondence
Associate Professor J D G ....atson , Neurop sychology Unit. Royal Prince
Allred Hosp'tal . Missenden Road , Camperdo wn. NSW 2050,
iwatsonO mail.usyd,edu.au

MJA

Vol175

15 October 200 1

:

. .

• The clinical approach to the patien t with a suspect ed
disorder of memory and intellect is to establish whether
it is dementia. which parts of the brain are affected.
what is the cause, what is the prognosis, and wha t can
be done about it.
• The diagnosis of demen tia usually requ ires the
involvement of memory and at least one other cognitive
system. Delirium and depression are important
differentia l diagnoses .
• Patient s with dementia should usually have some
simple investigations afte r a careful history-taking and
examinat ion to identify "reversible" caus es.
• The commonest ca use of dementia is Alzhe imer's
disease, in which short-term memory disturbance is
usually prominent. Other causes of dementia include
cereb rovascular disease, t ewv-ooov disease and
Pick's disease .
There is now hope for patients with Alzhei mer's
disease (which can be treated with some success with
cholinesterase inhibitors) and patients with vascular
demen tia. in whom aggressive control of caus al risk
factors may retard progression .

The classical exam ple is of left inferi or frontal dama ge cau sing language di sorders. Other localised functions include
vision (occipita l lobe) and praxis (th e ability to form in th e
mind th e steps needed to carry out a task), visuospatial ability and na vigat ional skills (non- d ominant parietal lobe) .
Dis tributed function s: These arc an atom ically less well-defined
cognitive processes than langu age or vision. Attention and
memor y are the two m ost impo rt ant exam ples.
Di sturbed atten tion lead s to distra ctibility and inability to
perform the task. at hand. Pa tients may pe rsist inappropriately with actio ns or respo nses, not ch anging their mental set
to new requiremen ts (perseveration) . D isordered attentio n
leads to disorientation in time an d oft en in place . Delirium
or the CKll U confusWnal SUl U is the corres po ndin g clinical syndrome. Consciousness m ay also be im pai red or cloudy, but
th is is not th e prima ry prob lem.
The fun ctional anatomy of th e anentional system includes
the brains tem , th alamus and fronta l cortex, with some righ tsid ed emphasis for the latter. Focal or diffuse brain in jury
may cause att entional probl em s. The causes can be metabolic, toxic, pharmacological o r st ru ctu ral; th e first thr ee arc
more com mon . I Gathering an accurate history is the be st

433

MJA Pract ice Essentials

approach to such patients. Investigations may be useful,
includ ing brain scanning (CT or MRI), electroencepha lography (EEG) and lum bar puncture .
There are vario us conc epts of memory, but all agree that
there is mo re tha n one type. This is of practical benefit, as
differ ent diseases have different effect s on memor y.a-s For
example, patien ts with Alzheim er 's disease ma y recall how
to d rive a car, but no t where they left the keys or the route.
Box I shows a diagno stic approach based on pattern of
memory loss.

We do not know the cau se of AD . About 2% of cases are
dependent on a single gene mutation, involving chromosome
2 1, 14 or 1. Families with these mutations have early-onset
AD and are much studied in the hope of finding the underlying path ological mechanisms. There is recent evidence that
genes on chromosom e 10 may be risk fact ors for the more
common late-onset AD. ~ ·6
On chromos ome 19 lies the apolipoprotein E (APOE)
gene, which has a number of allelic forms. The APO EE~
allele confers an increased risk of develop ing late-onset AD
(heterozygotes have a two- to fourfold increas e in risk,
homozygotes four- to eightfold). This allele ma y affect the
general tendency to neuronal damage rather th an the
specifics of developing AD; it may also increase the risk of
poorer outcome after acute head injury."
Risk factors for AD are increasing age and family history.
Significant head injury may confer a slight increase in risk
(say, twofold), but there are conflicting studies. Possible protective facto rs include higher levels of educ ation, oestrogen
ingestion , non-steroidal anti-inflamm atory drug use, and viramin E. Pathogenesis proba bly depends on predisposition
and environmental factors. This will have therapeutic implications, as it seems unli kely tha t tr eatm ent ta ken after AD
onset will reverse the damage. If an agent is developed to be
taken for many years by tho se at risk of AD, precise screening will be ne eded and Costs are likely to be high .

Causes of dementia

Alzheimer's disease
This dis ease is defined by its neuropathology of extraneuronal senile plaq ue s, containing a core of am yloid protein,
and neurofibrillary tangles, conta ining paired filaments in a
helix formation. It is believed that cells containing tangles are
degenerating and will subs equently die . T heir loca tion and
numbers are likely relat ed to the type and severity of cogniti ve dysfunction. Major sites are the pyramidal cells of
association cortex and in the hippocam pu s. They are also
found in a number of subcort ical nuclei with diffu se projections to the cor tex, such as the nucleus basalis of Me yner t, the proje ction of which is cho linergic. This is the
rationale for using centr ally act ing cholinergic medi cations.

1 : Al gorithm of decision processes In diagnosing non-d elirious pat ients with memory problem s

I

Mild mem ory pr ob lem s

Psychological
factors ,

stress ere.

I

Mod era te m emory p ro blems

No

Sudden

psychological
factors

Temporal

Seve re me mory p roble m s

Poor test
co nsistency

or

I

PSyc!lological

I

anxiefy,
depression ,
etc.

Mal inge ring

sub trontal
palllOlogy
(eg, stroke
or lumo ur)

ex~anafion:

Memory
probl am in
prcooruon
1001ilE/r

cogn itive
problems

Sudden

Gradual

Temporal or
sub!ron la l
pathology
(eg. small
stroka or
lumour)

oerscaterer
or parieta l
lumour,
Parkinson's
disaasa ,
mull iple

scle rosis

Memory
problam not
in proportion
to omer
cogni tive
problems

Long

dueane n

Gradual
Tho ught
disorder or
hall oo natio ns
Psychosis

Fronl al
lumo ur
wilh CSF
obs truction

NO thought

ersoreer
Hunlingl on's
d isease ,
Pick's disease,
vascular

Lossot

demen~a

Personal ide ntify
intact

personal
idenlity

Shan
history

Long
hislory

BiITonlalor
temporal
lumour

Alzheimer s
dise ase or
simjlar

Transien t globa l
amnesia .
herpes s imp~x
erlCepha lilis ,
c aetererposienor
c ere bral arte ry
jntarc lion

Hysteria or
tugue state

Note mat the dec ision pat hs and examples of illnesses are not exhaustive. Adapted and much simpljlied Iro n Kap ur 198 8 , Fig ure t.t

434

MJA

Vol 175

15 Octobe r 2001

~' _ \ Practice Essentials

Vascular dementias
Al though th ere i s acceptance tha t cerebrovascular disease
causes deme ntia, there is less agreement on it s definition and
preval ence. Vascul ar d em entia (VA D ) o ften coexists wi th
other for ms of dem entia , especiall y A D. There is overlap in
the risk factors for bo th AD an d VA D, and trying to di stinguish ri gi dl y between th e two m ay be ccunte rp ro ducrive.s
VAD pr ob ably accounts for up to 20% o f d ementia in West e-n society. The clinical p icture differs between pati ents w-rth
multiple larg e-vessel infarc ts (m ulti -in f ar ct dementia) and
those wi th sm all-vessel disease affecting the wh ite m atte r .
The clin ical cou rse is often one of flu ctua ti on , and st ep wi se
deterior ation i s a hallmark , oft en with evidence o f stroke.
O cher suggestive features include early gait disorder, urinary
difficulties and parkinsonism.
Diseases oth er than stro ke can cau se VA D, including vasculitis. recurrent sm all emboli, recurrent haemorrhages, and
infections suc h as syphilis and L ym e disease. 'Whi t e ma tte r
changes on computed t omography (CT) and m agn etic resonance i m agin g (.\t Rl) are com mo n ly th ought to indicate

cereb ral isch aem ic di sease, b ut they al so occur i n n or ma l
agei ng. A recen t study showed th at th e prevalence and
d egr ee o f white m atter lesions in crease wi th age; at any age
wom en have m ore lesions, particul arly i n the fron tal
lobes." The aut hors sp ecu late that th is m ay explain the
h igher incidence of dem entia in women tha n men , especially
at l at er ages.
A t pr esent , m ost au thoriti es wou ld suggest that cer ebrova scula r r i sk factor m anagem ent is the correct app roach
to VAD, but the evidence i s not stro ng. There i s on ly one
me thodol ogicall y sound trial, i n a group o f 3 7 patient s who
received 325 mg o f aspirin daily. Compared ....ith 33 controls,
th e r at e of cogniti ve decl ine slowed. w Another observational
stud y found that patients older than 50 years taking sta tin s
arc less li kely to have d ementia , with th e effect seem i ngly
rela te d to the m edications per se, and not the ch olester ol

level.'!
Large multicent re, r and om i sed , d ouble-blind, pla cebocontrolled trials would be best in assessing the effect of
2 : Criteria for e1in'ul d iaenosis of probal* AD

C.se history: ProC.....lve eocnitt'te chances

A ez-veer-oro man presented bec ause of worsening
memory. One 01his adult children had suddenly
separated. so the pat ient had more responsibility for
. ing atter his grandchildren. He was thought to be
depressed. Antidepressants caused some improvement in
mood. but had little effect on memory. The patient still
vcrkeo part time in an unskilled jOb he had held for many
rears and drove a car without getting lost His father had
displayed some features of dementia at about the age of
80.
He was slow and vague in conversation and scored 14/30
on the Fclstein MMSE," well below the cut-off for likely
severe co gnitive impairment. There were no primitive
reflexes or focal neurological finding s. The gail was a little
slow and stooped, with poor arm swing.
A CT scan of the brain was normal. Neuropsychological
assessment demonstrated severe constructional apraxia,
impaired semantic memory, executive dysfunction , with
some impairment of recent memory, A positron emission
tomography scan of brain metabolism'Sdemonstrated
impairments in the pari etal and temporal lobes, consistent
with dementia of the Atzneener type.
Within a year he had to give up work, Repeat
nevoc svcnocarcat assessment showed worsening
function, now with impaired ep isodic memory. He was
commenced on ooneoezn. but could only tolerate a dose
of 5 mg per oav His rate of cognitive dec line slowed over
the ensuing three years.
Comment: This case is not unusual. There is some family
history, but not especially strong . Depression was
considered and treated . The constant pattern 01the
patient's daily life probably enabled him to continue at
work for longer Ihan might have been expected . His
presentation in normal conversation was slow and vague,
but only when he was ch allenged by a lormal cognitive
assessment did the severity of the prob lem become
evident

MJA

Vol 175

15 October 2001

Gener al
• Onset between 40 and 90 years
• Dementia established by clinic al examination and/or
cognitive tests
• Deficits in more than one area of cognition
• Prcqressive worsening of cogn itive state
• No evidence of d elirium (at least initially)
• Absence of general Of neurologic al disease that could
account for the progressive cognitive deficits
Supporting l eatur es
• Progressive deterioration of spec ific functions such as
language, praxis and percept ion
• Impaired activities of daily living
• Altered behaviour
• Similar family history
• Progressive cereb ral atrophy· on computed
tomography or magnetic resonance imaging: normal
lumbar puncture: electroencephalog ram normal or nonspecific chang es
Features c onsIstent with diagn o sis
• Plateaus in progression
• Associa ted symptoms: depre ssion, insomnia,
incontinence, delusions, illusions, hallucinations,
behavioural outbursts, sexual disorders, weight loss
• Certain neurological signs: increased muscle tone,
myoclonus, gail disorder, seizures
• Structura l brain scanning normal for age·
Featur es that make dIag nosis uncert etn or un likely
• Sudden onset
• Foca l neurological signs , especially if early in the
course
• Seizures or gait disturbance early in the course
AOap!l!'CIlrom McK haM et ai " ,ge,. ." " Espec lllllV initially !her e are otten no
(itagnosbC features on S!l'uCIU'a1 ornag.ng....m C T Of MRI . aIIhougn hn::bonaI
sc anr>ing tecIYliques . in parbcular QU8ntJl allve POSItron emoSSIQfl ItllTlOg-"raphy . may be very helptvl

435

MJA Practice Essentials

3 : Fe atures di s tin guishing d elirium, d ementia an d d epre ssion
Feature

Delirium

Dementia

Depression

Onset

Acute

Insid ious

Ac ute or subacu te

Time cou rse

Fluctua tes dur ing single day

Worsens over lengthy time ;
cognit ive loss precedes depression;
behaviour worse at night

Depression precede s co gnitive loss

Weeks to mont hs; otten prior history

Duration

Hours to weeks

Months to years

Attention

Impaired

Little affected; impaired in later stages Unimpaired

Orientation

Usually impaired in time, often p lace,
sometimes person

Impaired in later stages

Unimpaired

Alertness

Abnorma l: low or high

Normal, at least initially

Norma l or low

Cons istency of
performance

Variab le

Consistent

Variable and slow

Sleep-wake cycle

Always affected

Affected in later stages

Difficu lty sleepi ng .
early morning wakening

Short-term memory

Always impaired

Not affected initially

Poor effort may lead to variable
impairment

Lonq-term ep isod ic Impaired
memory

Impaired; recent loss e xceeds remote Equal recent and remote loss

Perception

Otten visual illusions and halluc inations Commo nly impai red in later stages

Normal , unless psychotic de pression

Cogniti ve loss

Unaware or den ied

Minimised by patient

Maximised by patient

Word-find ing d ifficulty,
though ts of deat h and dying

Suicidal thoughts

"Near miss' frequent

"Don't know" freq uent

Speech and thought Disorganised , deluded, slow or rap id
Ramb ling , impersistent

Answers given

Partly adap ted from Feinberg and Goodma n. 1984.'·
,.-

z

aspmn, sratins and other treatments. H owever, there are
many technical issues, such as which cognitive, be havioural
and global domains to assess. Because vascular risk modification benefits p atients in other ways, there wou ld be et hical difficulties in administering a placebo to pa tients with a
diagnosis ofVAD, or with vascular risk factors.
Other causes of dementia

T h ere are many other causes of d em entia, so me less ra re
th an previously thought. F or exa mple, Lewy bod ies are
found in the b rains of2%-9% of elde rly p eop le, an d de mentia with Lewy bodies (D LB) accounts for 12%- 27% of cases
previou sly diagnosed as AD. 12 D L B has the features of fluetuating cognition with m arked va riation in attention and
ale rtness, visual hallu cin ation s and p arkinsonism (u su ally
without response to L-dopa) . N eurolep tic agents, presc ribe d
for the hallu cin ations , often considerably worsen patients
with DLB. As with VAD, DLB overla ps co ns ide ra bly with
AD .
Another condition, Pick's disease, is strictly d efined by the
presence of characteristic neuropatho logy, but the rubric is
now often used clinically for dementia of th e frontal-lob e
type, an d frontotemporal dementia.'? Some cases are
fa milial. In its complete form , patients d isp lay diso rdere d
executive fu nction (goal-sett in g, plan ni n g), disi nhibited
behaviour and lan guage dysfunction. Cogni tive p rob lems
usualIy seen in AD , such as visuospatial diffic ulties and
ap raxia, are not common early in Pick's disease. Box 1 conta ins some other causes to be co nsidered.
436

Dia gnosi s
Do es the patient wi th cognitive impairmen t have
dementia?

In deme ntia, th ere is acq uired cognitive loss in an alert
p er son, interferin g with daily living, soc ial an d occ u pational
ab ilities. The d iagnosis is m ad e on history and/or cog nitive
testing . T he usual d efinitions require th at at least two areas
of cog nit ion (me mory, prax is, langu age, visual percep tion,
visuospatial abilities, and executive skills) be affected, one of
th em m em o ry.
Simple cogn itive sc reen ing tests that cover a n um ber of
cognitive domai ns (and take about five mi nutes) can help
d etect pa tients with dernentia.v-" Some kind of form al testing is important, as m any patients with mi ld to m oderate
dementia do not p resent too badly in no rmal co nversation
(Case histo ry). Box 2 outlines the features needed for a diagn osis of AD, th e m ost common d ementi a. Fam ily m em bers
are ab le to ma ke the d iagnosis: a meta-an alysis of studies
co m paring infor m an t questio nnaires with brief co gnitive
tests fou nd that the for mer were a litt le more sen sitive with
the same specificity, and d eserved to be u sed m ore
widely."
T h e m ajor di fferential di agnoses ar e de lir ium an d d epressio n (Box 3) . When a pa tient's confusion leads to florid
behavioural disturbance, the dia gnosis of d eliri um is usually
sim p le. M ore problematic is th e patien t with quie t delir ium:
the impai rment of co nsciousness is sub tle and the behavio ural ch ang e ofte n n egative rat h er than di srup tive, so the
d isord er may be mi ssed. A me ta- analysis of me thods to disMJA

Vol 175

150cl ober200 1

I'U A Pract ice Essentia ls

nnguish de pression and dem entia also concl u ded th at info rmant questionnaires performed as we ll as co gnitive scree ning rests .w

Sub cortical demen tia m ay be found in H u nt ington 's and
Parkinson 's d iseases, AIDS en ceph alopath y, mu lt iple sclerosis and cereb rova scu lar disease, alt hou gh the latt er ofte n
causes a mixed pictu re.

What part of th e brain is affected ?

It helps p ractically and theoretically to di vid e d em entia
according to what p art of th e brain is affected : cort ical o r
subcortical. AD and C reutzfel dt-jakob disease are examples
of the former. Lo ss of co rt ical grey matter lead s to co rresponding cognitive d efects; th us, dist urbed m em ory, lan guage, praxis an d visu ospa tial abi lities arc th e pr ed omi nant
featu res in AD, wh ile attention and frontal-e xecu tive fu nctions are well p reserved , at least in itially.
In subcortic al de ment ias, the cogn it ive d efects ari se from
the loss of sub cort ical regu lat ion o f cortex, espe cially p refrontal. Typically, the re is slowin g o f cognition and
responses, p er son ality change, and mood di sturba nce.
Patients are oft en uninterested in th e world and m ent ally
inert. Reduce d attent ion leads to p oor encod ing and th us
poor memory. Recogn ition mem ory is b ett er than free recall.

Whom to r efer

Box 4 outlines th e diagn ostic ap p roach in a p atient with
d em entia. O n e co u ld argu e that if syste ma tic enquiry an d
exa mina tion provid e eno ug h infor mation to d iagnose AD
with out any u n usu al fea tures, th en spec ialist referral is no t
n eeded. H owever, the d iagn osis is so serio us for all con cerned that patients, relatives an d G Ps often feel more comfort abl e if it is su ppo rted by a specialist, usually a general
physician , neurologist, geriatrician or psychiatr ist. O ther factors det er mining referral inclu de regi on al var iat ion in th e
o rgan isation of assessm ent services, access to C T and M R
scannin g, and requirements for access to subs idised d ementia med ication s and subs id ised nu rsing home care .

4: Dementia: c li nical approach and Inves tig at ions

History
Obtain the history from a reliable informant as well as from the patient. Key points are the areas of cognit ion affected ,
especially compa red with previous funct ion and skills, and the mode of onset and progression. Actual examples of cog nitive
impairment are very usetot
Becoming hopelessly lost in previously familiar territory is significant. Forgetting how to use familiar household objects or
appliances sugg ests apra xia; not recognising objects for what they are suggests agnosia. Not being able to plan and prepare
a meal sugges ts a d isorder of execut ive func tion (often related to frontal lobe problems) .
Neurolog ical and ph ys ical exa mi nation
Carry out genera l and neurological examinations. For example , left ventricu lar hype rtrophy, hypertension, and other vasc ular
risk factors, when co upled with foca l neurolog ical features, suggest vascu lar d ementia. Parkinsonian features sugg est
Parkinson's disease or dementia with Lewy bodies .
Brief co gni t ive test s
The Mini Mental State Examination (MMSE) and similar short instruments are only screening tests." The lower the score the
more likely is dement ia to be prese nt. Early demen tia, espec ially in a patient with bette r education and intellect, is unlikely to be
picked up by these tests,
Labo rato ry tests and " reversi b le" dementia
Lists of laboratory tests have been suggested for detecting "reversible" dementia. Recent evidence suggests that such gene ral
screening is unlikely to c hange outcomes. If the case is typica l of Alzheimer's disease , then there is lilli e ind ic ation tor a wide
range of tests. Patients with "reversible" dementias , even if identified and treated , rarely improveet.ee
Appropriate laboratory tests include standard haematology and biochem istry (includ ing serum calcium), thyroid function,
vitamin 8 '2 levels, and syphilis serology. If history and examination are sug gestive , other tests may be valuab le, such as
immunological tests, erythrocyte sediment ation rate, toxico logical and heavy metal screens, cerebrospinal fluid examination,
and HIV tests, Rarer cause s of dementia are possib le: later Investigations may include tests for Whipp le's and coel iac d isease,
and genetic tests. Occasionally even a brain biopsy is done, especially in rapid ly prog ressive dement ia in a younge r person .
Neuroimagi ng
In the absence of conclusive evidence as to the utility of computed tomog raphy (CT) sc anning, I believe this should be done
on all patients to rule out certain causes ot cogn itive c eclin e such as subdu ral haematoma. Magnetic resonance imagi ng (MRI)
may be more useful, at least in a subset of patients. In c ases of dementia, CT and MRI may be co mpletely normal, may show
focal or general atrophy, or may show other relevant abno rmalities. In the tertiary referral setting, funct ional imag ing , especially
glucose uptake as measured by pos itron emission tomogra phy. can be very helpf ul in early or atypical cases.
Neurop sy c ho lo gical test s
Dementia present for some l ime may not need specia lised neuropsycholog ical assessmen t. However, formal
neuropsychological testing is valuab le early in the cours e of dementia, or if there are unusual features, with dementia in a
younger perso n, or when there is the possibility of depression or similar.

MJA

Vol 175

15 October 2001

437

MJA Practice Es s e n tj a ~ s

Pro gno sis
The p rogn oses of AD an d VAD d o n ot differ mu ch . Progression rate before the tim e of d iagnosis in fluences subsequent p rogress. If two pa tient s ha ve similar cognit ive
diffic ult ies at d iagnosis, th e one with th e lon ger time to th at
point will do better than the othe r. T he estimated med ian
su rviva l after onset varies fro m five to n ine years. A recen t
Canadian study has readd ressed this issue, arguing that there
has been failure to cons ide r pa tients who progresse d and
d ied too rap idly to be included in studies (length bias).H F or
both AD and VAD, the un adjusted med ian survi val was in
accordance with other stu d ies at 6 .6 years. T he ad justed
median su rvival was much less at 3 .1 years for prob able AD ,
and 3 .3 years for VAD . The two-year su rvival rate for
pa tients in the co mmun ity is about 50% more th an for th ose
in nu rsing homes.w
M anagement
There are no speci fic tre atments for most causes o f
dementia. M any therap ies have been tried (Box 5) . T he
effect iveness of most h as not bee n well pr oved or disp roved.
It is important to expla in to th e p atient and family the featu res of the de me nt ia an d p rognosis, and to d iscuss th e
m ajor cognitive prob lem s. Review of m edications that m ay

worsen matters is nee ded, an d advice on minimising alcohol
in take. T reat m en t of coexisting con ditions, such as d epres.
sion or poor hearing, can h elp many.
M anagement of behavioural and psychiatr ic m anifestations
is very ch allenging, especially agit ation and wandering.
Simple beh aviou ral techniques, environmental manipula tion
and a stron gly structured routine may help. M ed ications are
ofte n used , but their safety and efficacy are n ot well demonstra ted . As mo re b asic activi ties of da ily living are lost, the
level of ca re n eed ed in creases, eventually requ iring nursing
home care. I d iscuss this with the family early to initiate steps
to ident ify op tions an d costs.
I recomm en d early contact with the Alzheim er 's Associatio ns, and discussion of financial, estate and gu ardia nship
issu es. Driving sho u ld be d iscussed. Earl y in the illne ss it
may be possible to drive, with geogr aphical or diurnal licence
rest rictions. I often refer for an on-road d riving assess ment.
Caregiver distr ess, fatigue, de press ion, physical illness, and
subs tance m isuse are ofte n serious problem s. For fami lies of
those wit h m ild to moderate AD, counse lling and supp ort
may delay the tim e to n ursing home placement by a year."
Respite care is likely to be useful, but is often h ard to obtain.
In Australia, there are strong Alzheimer's Associations. Th ey
organise semina rs and education programs for carers and
he alth profession als, an d provid e publications and video-

5 : Treatments that have been tried for dementia
Acetyl chOline manipulation s
Certain acetylcho linesterase inhibitors (donepe zil, rivastig mine. galantamine) cente r modest improvements in patients with
Alzheimer's d isease (AD) in the areas of cognition, behaviou r and activities of daily liVing (El ),h Z!
There is no evide nce that nicotine is a useful treatment for Alzheimer's disease (E3z),2lI Extra cons umption 01lecithin. a major
dietary source of cho line, is hypothesised to increase the production of acety lch oline and reduce symptoms of dementia, but
randomised trials do not support its use (El), 29
Neuroprot ectlon
Vitamin E may benefit peop le with AD, but cu rrently there is insuffic ient evidence of efficacy (E2).30 There is promising evidence
for a benefic ial effect of seleg iline in AD, but not enough to recommend its routine use (El ).31
Hormonal t he rapy
A meta-analysis of obse rvational studies on hormone replacement therapy and cogn ition sugges ted a decre ased risk of
dementia, However, better, larger studies are needed (E3~) . 32
There is no p resent evidence for any benefit 01oebvdroeptandrosterone (El),33
Anti -Inflamm ato ry drugs
There is no evidence that aspirin is effect ive in treating patients with vascula r dementia (E3 ~) . Jo<
Obse rvational studies in patients with arthritis suggest that NSAIDs may have a protect ive effect on the incidence and onset 01
AD. Prospective, double-blind , placebo-controlled studies are needed to determine their role (E3z),35
Antipsy chotIc medication
Classica l neuroleptic s have little efficacy over p lacebo in behav ioural diso rders associated with dement ia, with efficacy rate
eq uivalent to side effect rate (Et ).36 However, thioridaz ine, one of the most commonly presc ribed drugs because it is thought to
produce less frequent motor side effec ts, has marked anticho linergic propert ies that could have a detr imental effect on
cognition. A meta-analysis showed that, com pared with placebo, its only positive effect is in the reduction of anxiety, and it has
equal or higher rates of adv erse effects than p lacebo , other neurolep tics and othe r sedatives (El),37 Risperidone may be more
r eiofur and safer (E2).36
Oth er th erap ies
Reality orientation (the presentation of orientation information (eg . time. plac e and person-related] to provide patients with a
greater understand ing of their surround ings) has some benefits on both cogn ition and behaviour. A continued program may be
needed to sustain potential benefits (E1).39
No firm conclusions are possible regarding the effect iveness of reminiscence therapy (the vocal or silent reca ll of events in a
person's life, either alone , or with another person or group ot people). The trials are of poor quality (El) .40

43.

MJA

Vol 175

15 October 2001

MJAPrad ice Essentials

tapes. There is a strong co mmitme nt to ca rer support : in
New South Wales alone there are more th an 160 de mentia
carer support gro ups . Understanding of dementia and its
progression, help in man aging difficult behaviour, and advice
on other co mm unity support opti ons can all be obtai ne d.
The Alzheimer's Association of Austra lia can be contacted at
PO Box 108 , Higgins, ACT 26 15. Phone: (02) 6254 4233.
Fax: (02) 6278 7225 . Web: <http://www.alzheimers.o rg.au>.

Evidence-based guidelines

• Certain acetylcholinesterase inhibitors can benefit
people with Alzheimer's dtseese.e-" (E1)
• Hormone replacement therapy may reduce risk 01
oemenue.» (E32)
• Non-steroidal anti-inflammatory drugs may have a
protective effect on incidence and onset of Alzheimer's
oteeeee.» (E32)

New medications for dementia

If, Febr uary 2001, the Commonwealth subsidised, unde r
strict conditions, donepezil and rivastigmine for use in mild
to moderately severe AD. These drugs increase brain acetylcholine concentrations, thus modestly helping m em ory difficulries and oth er cognitive or behavioura l problems (see
E0X 5). They are not cures, and at present th eir continued
subsidy will require evidence of improveme nt (not just stabilisation) after six mon ths' use . Other medications are
appearing on the market, incl ud ing galan tam ine.
It is hoped that effective agen ts will emerge for th e treatll• ent of th e various form s of dementia. The difficulty of this
task cannot be overstated. Basic knowledge of the mechanisms of de m ent ia, th e ident ification of th ose at risk, good
study design , eth ical issues, cost and side effect pro files are
ali factors th at make thi s a challenging endeavour.
References
Rummans TA. Evans JM. Krahn LE. et al. ()(!Iirium in elde rly patients; evaluation
and manage ment. Mayo Cion Proc 1995; 10, 989·998
2 HQdges JA. COgmtive assessment lor clinicians Oxford: O.ford University Press.

""

3 BunefS N. Dells DC. Lucas JA Clinical essessmeot 01memory disO/dllls in amne·
s,a and dementia, Annu Rev Psycholl 995; 46 493·523
4 Kapur N, Memory di sord ers in clinical p ractice . Lond on; Butlerworths , 1008
5 En ekin·Taner N. GraH-Radlord N, Younkin LH. et al. Linkage 01plasma AI}42 to
a quantitativ e loc us On chromosome 10 in rate-co set Alzheimer 's disease pedi.
grees. Science 2000 ; 290; 2303·2304,
6, Myers A. Holmans p. Marshall H. et al. Susceptibi lity locus fO/ AIZheimer's dis ·
ease on chromosome 10. Science 2000 : 290: 2304-2305.
Teasdale GM. Nico ll JA. Murray G. el at. Assoc iation of apol ipoIXote in E polymorphism with outc ome after head injury. Lancet 1997; 350: 1069· 1071.
8 Schmidt R. Schmidt H, Fazekas F. Vascular risk lactors in demen tia. J Neurol
2000; 247: 8 1-8 7.
9. de Leeuw FE, de Groot JC. Achten E. et al. Preva lence of ce rebral white maner
lesions in elderly people a population based magnetic resonance imaging study,
The Ronerda m Scan Sludy, J Neurol Neurosurg Psych,atry 200 1. 10, 9·1 4.
10. Meyer JS. Rogers RL. McClintic K. et at Randcm ized clinical trial of dai ly aspirin
therapy in multi·infarct demel1tia, A pilot study, JAm Geriatr Soc 1989; 31: 549·

' 55

1' . Jick H. ZornOOrg GL, Jick SS. et er. Slatins and the risk 01 dementia Lancet 2000:
356: 1627·16 31.
12 Perry RH. Irving 0 , Tomlinson BE, Lewy body prevalence in the aging brain: rerationship to neu ropsyc hiatric disorders. Alzheimer-type pathology and catecholaminergic nucje. J Neurol Sci 1990: 100 223-233
13 Kertesz A. Murooz 0, Pick's di sease, Irontolem poral dementia. and Pick comple x:
eml!(ging co nce pts. Arch Neurol 1996: 55: 302-304
14 FoIstein MF, Fcrstetn SE. McHugh PR. -Mini-mental stale- A pract ical mettlOd lor
grading the cog nitive state of palients for the clin ician. J Psychiarr Res 1975; 12:
189-100
15, Petersen RC. Smitll GE. Ivnik R.I. et aI. Apo lipoprotein E status as a p redictor of
the develop ment of Alzheimer's disease in memory -impaired individuals, JAMA
1995: 213: 1274· 1278,
16 Mazziotta JC. Frackowiak RS.!, Phe lps ME. The use of posit ron emission tomography in !he clini cal essess n-ent 01deme ntia S8min Nuc! Mad 1992: 22: 233-

'"

17, McKhann G. Orachman D. Foisleln M. el at Clinical diagnos is 01Alzheimer's di sease: report of the NINCOS-AORDA WO/k Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's disease,
Neurology 1984; 34; 939·944 .

MJA

Vol 175

15 October 200 1

18. Jorm AF. Method s 01 screening lor dementia; a meta-anary sis of studies conpanng an informant q ueslionnaire wilh a brief cog nitive test. Alzheimer DiSAssoc
Disord1997: 11: 158-162.
19 Feinoerg T. Goodman B. Afteclive illness, dementia, and cseoecoeeeone. J Clin
Psychiatry 1984 ; 45; 99· 103,
20, Lachner G. Engel RR. Difterenlia tion of dementra and depression by memo ry
tests. A meta-analys is. J New Men! Dis 1994: 182; 34·39
21. Freter S. Bergm an H. Gold S. et at. Prevalence of po tentially reversib le dem entias and aClual reve rsibility in a memory cl inic co hort. Can Med Asscc J 1998;
159; 657·662.
22, Burke O. Sengoz A, Schwa rtz R. Polenlial ly reversib le cogni live impa irment in
patients presenting to a memo ry disorders clin ic. J Clin Neurosci 2000: 7: 120·

tea
23. Wolfson C. Woilson DB, Asghar;an M. er er. A reevalualion of the duration of survival eeer the onset of dementia N Engl J Med 2001 , 344: 1111·11 16
24. van Dijk PT. Dippe l OW. Hab bema JD Survival of patients wilh deme ntia, JAm
Geriatr Soc 199 1: 39 603-6 10.
25. Birks JS. Melzer D. Beppu H. Oonepez il for mild and mccerate Alzheimer'S crsease (Cochrane Review), In: The Cochrane Library, 2.2001 . Oxlord: Update Soft.
ware.
26 Birks J. GrimHly Evans J. lakovidou V,et at. Rivastigmine lor Alzheimer's disease
(Coch rane Review), In : The Cochrane Library, 2, 2001. Oxford u ccaie Software.
27. Olin J. Schneider L. Galanlam ine 10/ AIZheimer's ceeeee (Coc hrane Review). In:
The COC hrane Library. 2, 200 1. O. ford Update Software
28 Lopez-Arrieta JM. Rodriguez JL . Sanz F, Efficacy and safety of nic otine on
Alzheime (s disease patients (Coc ll rane Review) , In: The Coc hrane Li brary. 2.
2001. Odord: Update Software.
29 Higgins JPT. Flicker L. LeCithin for demen tia and cog nitive impairment
(Cochrane Review) In: The Cochrane Library. 2, 2000, O.ford: Updale Software,
30, Tabet N. Birks J. Grim ley Evans J. et al. Vitamin E for Alzheimer'S disease
(Cochr ane Review). In: The Cocllrane Library, 2. 2001. Oxford: Updare Software
31 Birks J. FliCker L. Selegiline 10/ Alzheimer's disease (Cochrane Review) In: The
Cochrane Lib rary, 2, 200 1, O.ford: upoate Softwa re,
32. LeBlanc ES. Janowsl<y J. Chan BK. eI er. HO/mor>e reptacerrem therapy and cog.
nition: systemat IC review and meta-analysis, JAMA 200 1: 285: 1489·1499.
33. Huppert FA. Van Niekerk JK. Dehydroepiandrosferone (DHEA) supplementation
for cogn itive funclion (Coch rao;e Review) . In: The Coch rane Libra ry. 2. 200 1.
O.ford: Update Softwa re,
34 Winiams PS, Rands G. Orrel M. et at. Aspi rin lor vascu lar dementia (Coc hrane
Review). In: The COChrane Library . 2. 2001. O dord: Up date scnwe -e
35 Flynn BL, Th~sen KA. Pharmacologic manag ement of Alzheimer disease par t
III; nonsteroidal antiinflammatory drugs-MlElr ging protective evidence? Ann Phar·
mac other 1999 ; 33: 840·8 49,
36 Lanclol KL. Best TS. Mlttmann N, et at Efticacyand salety of neurolepttcs in
behavioral disorders assoc iated with dementia J elin PSychiatry 1998; 59 550-

SO ,
37 . Kirchner V, Kelly CA , Harvey RJ, Thioridazine for dementia (Coc hrane Review).
In: The Coc hrane Library. 2, 200 1. oxtcre: Update Software ,
38. l aud ig M. A risk-benefit assessment of risperioone fO/ the treatment of eenev ioural and psychological sympt oms in demenl ia. DrugSa12OOO: 183-195 ,
39 Spector A. Orrell M, Davies S, et al. Reality orient ation lor dementia (Cochrane
Review), In: The Coch rane Libr ary. 2. 2001 O Xford: Update Softwa re
40. Spector A. Orrell M. Davies S, el ai, Reminiscence l herapy for dementia
(Coc hrane Revrew). Inc The Cochrane Library, 2, 200 1 ceceo: Update Software,
41. Miffelman MS. Ferris SH. Shulman E, el al. A lamily inlervention to del ay nursing home placemen t of patients with Alzheimer disease, A randomized conlroiled
l risl. JAMA 1996; 216; 1725-1731.
0

439