The Korsakoff syndrome.

M D Kopelman
BJP 1995, 166:154-173.
Access the most recent version at DOI: 10.1192/bjp.166.2.154

References This article cites 0 articles, 0 of which you can access for free at:
http://bjp.rcpsych.org/content/166/2/154#BIBL
Reprints/ To obtain reprints or permission to reproduce material from this paper, please write
permissions to permissions@rcpsych.ac.uk

You can respond http://bjp.rcpsych.org/letters/submit/bjprcpsych;166/2/154

to this article at
Downloaded http://bjp.rcpsych.org/ on July 10, 2014
from Published by The Royal College of Psychiatrists

To subscribe to The British Journal of Psychiatry go to:

http://bjp.rcpsych.org/site/subscriptions/
British Journal of Psychiatry (1995), 166,154—173 Review Article

The Korsakoff Syndrome

MICHAEL D. KOPELMAN

Background. Investigations of the Korsakoff syndrome by researchersfrom different disciplines

have proliferatedin recent years, making it appositeto review the variousfindings.
Method. This review is based on the author's knowledge of reports in the major clinical
and neuropsychologicaljournals, supplementedby Medline searchesto update particular
subtopics.
Results.The Korsakoffsyndromeis defined as a disproportionateimpairmentin memory,
relativeto otheraspectsof cognitivefunction,resultingfroma nutritional(thiamine)depletion.
The initialmanifestationsof the disorderare variable, and a persistentmemory impairment
can resultfrom a non-alcoholicaetiology,althoughthisseemsto happenmuchlesscommonly
than in the past —¿
presumablybecauseof generallyhigherstandardsof nutrition. Although
there is agreementon the underlyingneuropathology,the critical lesionsites for memory
disorderhavebeendebated.Recentevidencesuggeststhat the circuitinvolvingthe mammillary
bodies,the mammillo-thalamictract andthe anteriorthalamus,ratherthan the medialdorsal
nucleus of the thalamus, is particularlycritical in the formation of new memories. The
relationshipof these deficitsto thiaminedepletionremainsa topic of current investigation,
as doesthe purportedroleof neurotransmitterdepletionsin the cholinergic,glutamate/GABA
andcatecholamineandserotonergicsystems.Neuro-imagingstudieshaveconfirmedautopsy
findingsof morewidespreadstructuraland metabolicabnormalities,particularlyinvolvingthe
frontal lobes.
Conclusions.The relationshipof these neuropathological,neurochemical,and metabolic
abnormalitiesto cognitivefunctioning,with particularreferenceto specificaspectsof memory
processing,has been consideredin some detail. Whereas structuraland/or neurochemical
abnormalitieswithin the limbic/diencephaliccircuitsaccountfor anterogradeamnesia,some
other factor, suchas frontal lobedysfunction,must underliethe severeretrogradememory
loss which is characteristicallyfound in this syndrome.

The fascination oftheKorsakoff syndromeisthat Clinical description of the syndrome

relatively small,and sometimesabrupt,neuro
pathological lesions give rise to a severe amnesic Korsakoff's description
syndrome, disproportionate toanyother impairments
incognitive functioning.Theneurochemical processesAlthough earlier writers noted the association
involved and the precise specification of the nature between alcohol abuse and the development of
of the memory deficit remaintopicsof intense profound amnesia (e.g. Hooke, 1680; D'Assigny,
investigation. The recent emergence of neuro 1697), the first formal descriptions in the modern
imaging techniques for specification of structural scientific literature were by Lawson (1878) and
and metabolic abnormalitiess, as well as the existence Korsakoff (1887, 1889a,b). Lawson is one of the least
of increasingly flexible neurochemical, neuro acknowledged, and Korsakoff one of the most
pharmacological, and neuropsychological techniques, misquoted, authors in medical literature, and con
now providetheopportunity forrelating neuro sequently it is important to clarify what they actually
biological abnormalities to specific cognitive deficits said.
much more closely. Lawson (1878) described a syndrome, frequently
This paper reviews clinical and neuropsychological but not always of alcoholic causation, in which there
fmdings in this disorder, relating them to observations was a severe loss of recent memory. He distinguished
from neuropathological, neurochemical, and neuro cases in which there was a specific loss of memory
imagingstudies. As Korsakoff's own clinicalfrom those in whom alcohol induced a progressive
description of the syndrome has commonly been mis dementia, difficult to differentiate from general
interpreted, and turns out to be much closer to recent paresis; he also discussed a case of psychogenic
neuropsychological findings than many textbook amnesia. Inthecourse ofhisdiscussion,heremarked
interpretations, hisaccountwouldseema suitablethat “¿in cases where organic change has been
placetobegin. produced in the brain, the nature of the symptoms

154
 THE KORSAKOFF SYNDROME 155

will be caused not so much by the character of the processes . . . (and) facts are remembered―
exciting cause (e.g. alcohol), as by the physiological although their retrieval requires “¿specially
favourable
functions of the regions diseased―. conditions―. In some instances, events may be
In a series of three articles, Korsakoff (1887, remembered “¿but not the time when they occurred―.
1889a,b) described a syndrome of characteristic In more severe cases, “¿the amnesia is much more
memory disturbance, which he had witnessed in “¿notprofound. . . the memory of facts is completely
less thanthirty― casesofchronic alcohol abuse,as lost―.
well as 16 patients in whom alcohol had not played As is well known, Korsakoff mentioned that “¿such
a role. The latter cases included instances in which patients invent some fiction and constantly repeat
the syndrome had developed following persistent it. . . (for example) of conversations which have
vomiting (eight patients), postpartum (sepsis, never occurred. . . so that a peculiardelirium
macerated foetus), in acute or chronic infection develops, rooted in false recollections (pseudo
(typhus, tuberculosis), after toxic poisoning (carbon reminiscences)―. However, he tended to place
monoxide,lead,arsenic), or otherchronicdisease greater emphasis upon the confusion of “¿old
(neoplasm, lymphadenoma, diabetes). Although recollections with present impressions―,and he gave
Korsakoff didnotmake any specific reference to several examples of this:
Wernicke's syndrome, which had been described in
“¿In
teffing of something about the past, the patient
1881, he did mention that a “¿prodromalagitation and would suddenly confuse events and would introduce the
confusion― commonly preceded the appearance of events related to one period into the story about another
the memory disorder; that “¿sometimesthere are period. . . Telling of a trip she had made to Finland
ophthahnoplegia externa, nystagmus andlike mani before her illnessand describing her voyage in fair detail,
festations―; and thatgaitdisturbance (ataxia), the patient mixed into the story her recollections of
dysarthria and dysphagia may also be present. Crimea, and so it turned out in Finland peoplealways
Furthermore, the ‘¿psychic disturbance' was an aspect eat lamb and the inhabitants arc Tartars―.
of a ‘¿multipleneuritis' in which peripheral neuro
pathy was commonly, but not necessarily, present Subsequent clinical studies
(Korsakoff, 1889a).
The characteristic memory disturbance occurred Subsequent studies have confirmed the broad outline
in a setting of clear consciousness. of Korsakoff's clinical description, and that a
characteristic confusional state and memory disorder
“¿At
first, during conversation with such a patient... and/or neuropathological changes can result from
(heor she)givesthe impressionof a personin complete
possession of his (or her) faculties; he (she) reasons a number of factors besides chronic alcohol abuse.
about everythingperfectlywell,drawscorrectdeductions Moll(1915) described 30alcohol-related cases,
of
from givenpremises,makeswittyremarks, playschess whom 25 (83.3°lo) weremen. LikeKorsakoff, he
or a gameof cards,in a wordcomportshimself(herseli) emphasised the patients' intact attention, their
as a mentally sound person.― disorientation in time, and the variability of
However, their retrograde loss. The disorientation in time
resulted in the patients' underestimation of the
“¿the
patient constantly asks the same questions and duration of their hospitalisation; the retrograde com
repeats the same stories. . . may read the same page ponent of the amnesia could sometimes be extensive.
over and again sometimesfor hours. . . is unable to Similarly, Zangwill (1953) described two Korsakoff
rememberthosepersonswhomhe (she)met onlyduring patients who gave a striking underestimation of their
the illness, for example, the attending physician or
age, and whose retrograde amnesia extended back at
nurse―.
least ten years before the onset of the disorder.
Characteristically, “¿thememory of recent events... Talland (1965), Victor eta! (1971) and Luria (1976)
is chiefly disturbed. . . everything that happened all followed Korsakoff in arguing that confabulation
during the illness and a short time before―.However, often results from the inappropriate recall of genuine
insomecases“¿not onlymemory ofrecent eventsis memories,jumbledintemporalsequence.Victoret
lost, but also that of the long past―,in which case a! (1971) alsoemphasisedthe extensivenature of the
the impairment may involve memories of up to retrograde component of theamnesia:
30 yearsearlier. Korsakoffalsoemphasisedthe “¿(Whilst)
it is true that remote memorieswere better
variability in the severity of the disorder. In mild preserved than recent ones.. . it was our impression
cases, recent memories are “¿remembered vaguely that memoriesof the distantpast wereimpairedto some
(without) complete abrogation. . . the forget extent in practically all patients with Korsakoff's
fulness chiefly affects the patient's own thought psychosisand seriouslyimpaired in most of them―.
156 KOPELMAN

Spurning an emphasis upon either a specific deficit et a!, 1972; Wallis et a!, 1978), whereas others may
of recent memory or the presence of confabulation, never have been diagnosed in life, surviving in the
Victor et a! (1971) defined the disorder as “¿ancommunity with (presumably) a milder degree of
abnormal mental state in which memory and learning memory impairment (Harper, 1979, 1983; Torvik et
are affected out of all proportion to other cognitive a!, 1982). In short, it appears that the initial clinical
functions in an otherwise alert and responsive manifestation of the disorder may vary from acute
patient―.The diagnosis of the syndrome nowadays coma, through the classical Wernicke syndrome, to
requires both clinical and neuropsychological evidence an insidious onset of memory impairment, and that,
of this disproportionate memory impairment, as well in some cases, the disorder may not be identified until
as evidence of an underlying alcoholic or nutritional the subject comes to autopsy.
aetiology. Furthermore, various modern writers have con
Two topics in which clinical findings have been firmed that the Wericke-Korsakoff pathology can
more conflicting are the mode of onset of the result from a number of debilitating disorders, all
disorder and the prevalence of Wernicke features. of which produce malnutrition or malabsorption. De
Moll (1915) followed Kraepelin(1910) in arguingthat Wardener & Lennox (1947) described 52 prisoners
the disorder can have either an acute or an insidious of-war in south-east Asia who developed a Wernicke
onset, and he also noted a better prognosis in encephalopathy as a result of malnutrition, of whom
women. Cravioto et a! (1961) emphasised that the 32 (6l.501o) developed loss of recent memory. The
disorder most commonly first appears as an ‘¿organic onset was 6—14 weeks after the start of captivity, and
mental syndrome'(confusion, disorientation,memory was usually preceded by episodes of diarrhoea and
impairment, and dyscalculia) with or without vomiting, self-starvation, or concurrent infection
peripheral neuropathy, and that eye signs and ataxia (Cruikshank, 1950). Interestingly, in the early
were relatively rare. The full Wericke ‘¿triad' stages, the subjects complained of wavering vision
(confusion, eye signs, ataxia) was present in only four on looking sideways, double vision, memory loss,
(14.3°1o)of their 28 (pathologically-confirmed) cases, and impaired time appreciation, indicating that
and in none of 42 cases described in an earlier series ‘¿insight'
was initially preserved. Other writers have
by Riggs & Boles (1944). On the other hand, 96% described Wernicke symptoms and/or pathology
of the Korsakoff patients described by Victor & occurring after: self-starvation (Devathason & Koh,
Adams (1953) and Victor et a! (1971) had manifested 1982; Pentland & Mawdsley, 1982); intravenous
signs of a preceding Wericke syndrome, and only feeding, especially in the presence of a glucose load
lO°loof these patients had exhibited an insidious (Wallis et a!, 1978; Harper, 1980; Luda, 1980); the
onset (Victor & Adams, 1985). However, Cravioto persistent vomiting of hyperemesis gravidarum
et a! (1961) pointed out that the prevalence figures (Campbell & Russell, 1941; Nightingale et a!, 1982;
obtained in the Victor & Adams (1953) series would Wood eta!, 1983);and carcinoma of the oesophagus,
have reflected the nature of the referrals and the stomach, or intestine (Campbell & Russell, 1941;
diagnostic criteria employed within an acute neuro Malamud & Skillicorn, 1956; Ebels, 1978). Ebels
logicalservice;in short, this seriesmay have represented (1978) reported 29 cases of Wernicke-Korsakoff
a selected subsample of a wider population of pathology detected at autopsy, in whom chronic
Korsakoff patients. alcohol abuse had been excluded. He subdivided the
Consistent with this view, Cutting (1978a) found underlying clinical disorders into four main groups:
that, in a retrospective series of 50 Korsakoff patients (a) gynaecological (hyperemesis gravidarum, malig
diagnosed within a psychiatric service, the disorder nancy); (b) gastrointestinal (carcinoma, other causes
sometimes had an acute onset and sometimes an of malabsorption); (c) other malignancy; and (d)
insidious onset, defined in terms of the interval other causes of debility and malnutrition (e.g.
between the onset of symptoms and admission to haemodialysis, severe self-neglect).
hospital. An ‘¿acute' onset (less than eight weeks Although the above papers described Wernicke
between onset and admission) was more commonly symptoms and/or Wernicke-Korsakoff pathology
associated with the classical Wericke eye signs, resulting from nutritional disorders, most did
whereas an ‘¿insidious' onset (more than eight weeks not provide sufficient or unequivocal evidence
between onset and admission, the mean being one of persistent memory impairment, except for
year) occurred more commonly in women and had Korsakoff himself (1889), Dc Wardener & Lennox
a better long-term prognosis. More recent studies (1947), and Cruikshank (1950). Three recent papers
have indicated that some patients, in whom a have attempted to do this, using modern neuro
pathological diagnosis of Wernicke's encephalopathy psychological techniques (Beatty eta!, 1989; Becker
is eventually made, may first appear in coma (Torvik et a!, 1990; Parkin et a!, 1991). Beatty et a! (1989)
 THE KORSAKOFFSYNDROME 157
described a patient with a “¿Korsakoff-likeamnesic find his way about within the immediate vicinity, but
syndrome―following severe anorexia and vomiting, wouldgetlostifhe travelled further.
but this patient had very abnormal liver function B was formerly an electrician, and he had a full
tests, including a very high gamma glutanyl trans scale IQ of99withananterograde memory quotient
ferase (‘yOT), which raises the suspicion of past (AMQ) of approximately 51. He also showed fairly
alcohol misuse. Becker eta!(1990) described a woman severe impairment across various tests of retrograde
with malabsorption from an inflammatory cause in memory with a pronounced ‘¿temporal gradient'
the small intestine, who had a past history of heavy (relative sparing of early memories). On tests of
drinking, although that seems to have resolved by frontal lobe function, he was also quite severely
the time of her amnesic disorder. Parkin eta! (1991) impaired, scoring only 19on FAS verbal fluency, and
described a patient who had had prolonged intra obtaining only two out of six categories on a card
venous feeding, but the amnesia appeared following sorting test with 81Oloperseverative errors.
a surgical operation, in which she had to have a
blood transfusion, and there were multiple small
Case 2
lesions in the deep white matter on a magnetic
resonance imaging (MRI) scan, probably vascular in Born in 1936,C had been a receptionist and
origin, making unclear the possible contribution of secretary. She had consumed one bottle of sherry or
hypotension and/or hypoxia during the operation. as much as two-and-a-half bottles of wine per day
Inshort, itissurprisinglyhardtofindunequivocalfor approximately 15—20 years, and then consumed
cases of the amnesic syndrome from non-alcoholic three-quarters of a bottle of gin plus half a bottle
nutritional depletion nowadays, presumably because of wine per day for seven years u?ktilher admission.
of generally higher standards of nourishment than In June 1983, her general practitioner made a
when Korsakoff himself was writing. However, there domicilary visit and found that she had opthalmo
is little doubt that non-alcoholic causes may plegia, marked ataxia, confusion, and profuse
complicate and compound alcoholic causation, and confabulation. The opthalmoplegia and confabulation
the present author has seen one case in whom a resolved over the course of a month, following
previous gastrectomy, and another case in whom thiamine administration, but some degree of
carcinoma of the stomach, probably contributed to nystagmus and a pronounced ataxia remained, the
the onset of Wericke and Korsakoff symptoms. latter of which was secondary to both a sensory and
Finally, various forms of localised brain pathology a motor neuropathy. She was first seen by the present
can,of course,producean amnesicsyndrome, author in a general hospital, and, when visited in a
although the term ‘¿Korsakoffsyndrome' is nowadays long-stay psychiatric hospital two years later, she had
best reserved for those cases of persistent memory no memory of ever having been in the general
impairment with the characteristic pathological hospital.
changes (discussed below), resulting from nutritional C had a full scale IQ of 115 and an AMQ of 70.
(thiamine) depletion. On retrograde memory tests, she showed moderately
severe impairment with a pronounced sparing of
early memories. For example, she was able to
Caseexamples describe how she had won a tennis cup at the age
of 15, beating “¿GorgeousGussie―on the way, and
Case 1
how, after the prize-giving, her father had made a
Patient B,who wasbornin1925, hasbeenincluded derogatory comment about her maths. On the other
instudies of theKorsakoff syndromeby various hand,shewasabletogiveonlya veryfragmentary
authors. He has consistently claimed that he account ofhercurrent dailylife,andwasunableto
consumed ten pints of beer per night for 20—25 years, recognise the author from previous visits. On frontal
but only at weekends. If true, the effects of this lobe tests, she showed relative preservation, scoring
alcohol consumption may have been exacerbated by 45 on FAS verbal fluency and obtaining six out of
a partial gastrectomy which was performed in 1952 six categories on card-sorting with no perseverative
for a perforated duodenal ulcer. However, it is clear errors.
that B went on a four-week alcoholic binge following
thedeathof hiswifein1970,and thisculminated
Case 3
in an admission, in which nystagmus, ataxia, and
confusion were noted, as well as a peripheral D wasa school teacher who hadbeenbornin1923.
neuropathy. When last seen by the author, B was His drinking history was somewhat unclear. It was
living in a long-stay hospital, where he was able to probable that he had started consuming alcohol fairly
158 KOPELMAN

heavily either whileintheFleet AirArm duringthe Skifficorn, 1956; Cravioto et a!, 1961; Victor
Second World War, or during three years in the et a!, 1971; Harper, 1983). There were petechial
MerchantNavy following that,buthe admitted to haemorrhages, endotheial proliferation, focal areas
drinking heavily for only about seven years before of parenchymal necrosis, demyelination, gliosis, and
the onset of his disorder. During that time, he said variable degrees of neuronal loss (Jolluffe eta!, 1941;
that he had been drinking eight or nine pints of beer Malamud & Skillicorn, 1956; Cravioto et a!, 1961;
a night, plus a third of a bottle of spirits daily. The Victor et a!, 1971). As mentioned above, similar
onset of his disorder had occurred when he was changes have also been observed in subjects in whom
staying ina hotel inLagosin1976, whenhesuddenly a diagnosis was not made during life, but in
became ataxic, abusive and excessively talkative. whom there had been a history of alcohol abuse
Medical details of his admission were lacking, but (Harper, 1979, 1983; Torvik et a!, 1982). In some
in the 1980s he still showed evidence of nystagmus, ofthese lattersubjects, acute or‘¿active'
pathological
marked ataxia, and some degree of disorientation. In changes (haemorrhages, endotheial proliferation)
the past, he had been described as confabulating. D were prominent, but, more commonly, chronic or
was a linguist, who had previously been accomplished ‘¿inactive'changes (necrosis, ghosis) were predominant
in current European, classical, and several African (Torvik et a!, 1982; Harper, 1983).
languages. These linguistic skills appeared to be In addition, Malamud & Skillicorn (1956) reported
largely intact, providing evidence of the relative that cortical atrophy was present in 8.5% of their
preservation of the more semantic (conceptual/ Wernicke-Korsakoff cases, Victor eta! (1971) in 26°lo
linguistic) aspects of memory. of their series, and Harper (1983) in 21% of his
Current full scal@IQ was 129, and his AMQ was subjects. More recently, Harper et a! (1987, 1988)
69. On retrograde memory tests, he showed fairly have drawn attention to the gross atrophy, reduced
severe impairment across all tests, with a marked neuron count and increased hydration of the frontal
temporal gradient. On tests of frontal lobe function, lobesinalcoholic subjects atautopsy, particularly
he showedfairly severe impairment, scoring 24 on those inwhom a pathological diagnosis ofWernicke's
FAS verbal fluency and obtaining only two out of encephalopathy has been made.
six categories on card-sorting, with 45% perseverative
errors. Critical lesion sites for memory disorder
Although there is now general agreement about the
Neuropathology of the Wemicke-Korsakoff
overlapping distribution of the lesions in Wernicke's
syndrome encephalopathy and Korsakoffs syndrome, debate
has ensued regarding the critical lesion(s) for the
Type and distribution of lesions development of an amnesic syndrome, the thalamus
Bender & Schilder (1933) established that patients and the mammillary bodies being the sites most
with a patho!ogica! diagnosis of Wernicke's commonly implicated. Victor et a! (1971) pointed out
encephalopathy had often displayed a “¿mental that of 24 cases in whom the medial-dorsal nucleus of
picture . . . like that of a Korsakoff psychosis―. the thalamus was affected, all had a clinical history
Consistent with this, Joffiffe eta! (1941) reported that of persistent memory impairment (Korsakoff's
of 27 patients who were given a dinica! diagnosis of syndrome), whereas five cases in whom it was un
Wernicke's syndrome, 13 survived, of whom 12 affected had a history of Wernicke features without
were left with a residual Korsakoff syndrome. any recorded clinical history of subsequent memory
Subsequently, the pathological series of Malamud disorder. By contrast, the mammillary bodies were
& Skillicorn (1956) and Victor et a! (1971) clearly implicated in all the Wernicke cases examined,
established that the distribution and nature of the whether or not there was subsequent memory
pathological lesions in the two disorders were impairment. Amnesia in association with unilateral
virtually identical, resulting in the coining of the term or bilateral involvement of the thalamic nuclei, with
‘¿the
Wernicke-Korsakoff syndrome'. apparent sparing of the mammillary bodies, has also
These studies, and also that of Cravioto et a! been described in cases of tumour or infarction
(1961), confirmed that the pathological abnormalities (McEntee et a!, 1976; Speedie & Heilman, 1982;
lay in the paraventricular and peri-aqueductal grey Guberman & Stuss, 1983; Winocur eta!, 1984; Graff
matter, the walls of the third ventricle, the floor of Radford et a!, 1985, 1990; Katz et a!, 1987; Parkin
the fourth ventricle, and the cerebellum. The et a!, 1994).
mammillary bodies and the thalamus were among However, Mair et a! (1979) provided a careful
the sites most commonly affected (Malamud & pathological and neuropsychological description of
 THE KORSAKOFFSYNDROME 159

two Korsakoff patients, whose autopsies showed totheeffect ofdiscretestructurallesions (Speedie

lesions in the mammillary bodies and the midline and & Heilman, 1982; Markowitsch, 1984; von Cramon
anterior portion of the thalamus but not in the medial eta!, 1985; Valenstein eta!, 1987; Graff-Radford et
dorsal nuclei. They suggested that these lesions might a!, 1990). On the other hand, Zola-Morgan et a!
‘¿disconnect'
a critical circuit running between the (1989a,b) have produced evidence from primate
temporal lobes and the frontal cortex, and Warrington studies arguing against theroleoftheamygdaloid
& Weiskrantz (1982) put forward a hypothesis circuit, and indicating that parallel projections from
concerning the functional significance of this circuit. the perirhinal cortex maybe more important (compare
Mayes eta! (1988) presented findings in two further Meumer et a!, 1993; Murray et a!, 1993). A careful
patients, who had had careful neuropsychological MRI study of an amnesic patient with a left thalamic
and autopsy investigations, which closely replicated infarction hasrecently beeninterpreted inthese terms
the findings in the Mair et a! (1979) study. Von (Parkin et a!, 1994), although the involvement of
Cramon eta! (1985) reviewed findings in a personal both the mammillo-thalamic tract and the dorso
series of six cases of thalamic infarction (of whom medial nucleus in their patient leaves equivocal the
four were amnesic and two were not), together with relative importance of these two structures.
a further five cases from the literature. These authors In summary, the balance of the more recent
found that the region of the thalamus always evidence appears to suggest that it is a circuit
implicated in the amnesic cases (and not implicated comprising the hippocampus, entorhinal and perirhinal
in the non-amnesic patients) lay in the anterior cortex, the mammillary bodies, mammillo-thalamic
thalamus, involving the mammillo-thalamic tract, tract, and the anterior (rather than the medial dorsal)
ventral to the medial dorsal nucleus (cf. Graff nucleus of the thalamus, which is particularly critical
Radford et a!, 1990). More recently, Squire et a! in memory formation.
(1989a) and Dusoir et a! (1990) have described two
patients whose anterograde amnesia appears to have
Neurochemistry of the Korsakoff syndrome
resulted from traumatic lesions to the mammillary
bodies, consistent with the view that a circuit involving
Thiamine
the mammillary bodies, the mammillo-thalamic tract
and the anterior thalamus is crucial for memory In an elegant series of experiments, Alexander (1940)
formation. Findings generally consistent with these and Alexander eta! (1938) demonstrated that: (a) the
observations have also been reported by Castaigne lesions of Wernicke's encephalopathy could be
et a! (1981), Mori et a! (1986) and Gentilim et a! produced in pigeons following thiamine deprivation;
(1987). (b)these lesions were not produced by deprivation of
Markowitsch (1984) has argued that lesions within all other vitamins, when doses of thiamine were
the thalamic nuclei (or hippocampi) do not always adequate; and (c) the histology and topography of
give rise to amnesia and, when they do, there is these lesions were identical to those obtained in a
usually concomitant involvement of other nuclei. It series of 16 Wernicke patients. Subsequently, Joffiffe
follows from von Cramon et a?s (1985) fmdings that et a! (1941) reported clinical evidence of thiamine
there are critical regions within the thalami that must
deficiency in 27 Wernicke patients, of whom 12
be implicated for amnesia to occur; and Markowitsch survivors manifested the Korsakoff syndrome. The
(1984) acknowledged that there may be ‘¿nodal points'
latter group showed an incomplete and inconsistent
in the neural substrate of memory (c.f. Zola-Morgan response to thiamine therapy, although Bowman et
eta!, 1986). Markowitsch (1984) emphasised that the a! (1939) had earlier shown that disorientation and
thalamic nuclei and hippocampi do not themselves confabulation often show a favourable response to
represent discrete “¿functional
entities―,but are thiamine therapy in very early Korsakoff cases.
embedded in complex neural circuits and “¿neuronal Similarly, Campbell & Russell (1941) felt that
assemblies―. Consistent withthis, Mishkin(1978, thiamine deficiency was ‘¿strongly incriminated' in
1982) argued from animal experiments that there are a series of 21 cases of Wericke's encephalopathy,
two limbic circuits in which combined lesions are and these authors also noted the common association
required to produce severe amnesia: the hippocampal with Korsakoff's syndrome in survivors. Subsequently,
(medial limbic) and amygdaloid (basolateral limbic) Dc Wardener & Lennox (1947) reported their obser
circuits,(butseeMeuniereta!,1993;Murrayeta!, vations in malnourishedprisoners-of-war.The onset
1993). The anterior and dorsomedial thalamus and of symptoms,whichincluded mentalchangesin
the medial temporal lobes are the anatomical 78% of cases and loss of recent memory in 61%,
locations (‘nodalpoints') where these two circuits generallyoccurred6-14 weeksafter captivity;this is
converge,and are, therefore, particularlyvulnerable thesametimeassymptomsofberiberi develop (also
160 KOPELMAN
causedby thiaminedepletion), but beforethe subsequently, Leigh et a! (1981) reported wide
symptoms of other vitamin deficiencies. Moreover, variability in transketolase activity among Korsakoff
thesymptomsshoweda veryfavourable response to patients and other alcoholics. Nixon et a! (1984)
thiaminetreatment by injection, when thiswas reported that one particular pattern in the isocnzyines
available, butresulted ina veryhighmortality when of human erythrocytic transketolase was associated
thiamine was unavailable. with the presence of Wernicke-Korsakoff syndrome.
In retrospect, thecommon factorin both the However, the affinity (Km) for the interactionof this
alcoholic andthenon-alcoholic Korsakoff patients,particular transketolase variant with TPP did not
who havedeveloped thesyndromeintheabsence of differ betweenpatients and controls, leavingit
any other brain pathology, is almost certainly a unclear why thesepatients shouldbe particularly
thiamine deficiency. However, it is sometimes vulnerable to the effects of thiamine depletion (Nixon
claimed that a combination of thiamine deficiency et a!, 1984). More recently, Jung et a! (1993)
and the direct neurotoxic action of alcohol is required demonstrated thattransketolase activity in the
toproduce a persistent memory loss (unresponsive to fibroblasts ofthree Korsakoff patientswasreduced
treatment) intheKorsakoff syndrome(Freund, 1973; by 45%, but theirtransketolase showed normal
Butters & Cermak,1980). Althoughitisprobable immunochemical characteristics. Consistent with
thatthenon-alcoholic casestendtoshow a better this, McCool et a! (1993) have identified the trans
response tothiamine therapy, itshould benotedthat ketolase gene,findingthatno particular alleic
(a) Korsakoff's own series included 15 non-alcoholic variants could account for the biochemical properties
cases; (b)one of Dc Wardener & Lennox's (1947) oftheenzyme,when cultures fromtwo Wericke
and threeout of eightof Cruikshank's (1950) Korsakoff patients with extremely low affinity (Km)
prisoners-of-war showed persistent
mentalsymptoms for the TPP co-enzymewerecomparedwith cultures
despite treatment; and (c) in many other series from two non-alcoholic controls with extremely
adequate follow-up details are not provided, and high affinity. Hence, McCool et a! (1993) argued
consequently cognitive outcome is unknown. It seems that(extragenic) differences in post-translational
likely thattheresponse totreatment isdetermined processing and modification of thetransketolase
bytheageofthepatient (Tallaksen eta!,1993), the polypeptide, rather than allelic variation, might
abruptness of theonsetof thedisorder, and the underlie variable susceptibility tothedevelopment
rapidity with which treatment is instituted, and it is of thesyndromeamong alcoholics.
plausible thattreatment isinstitutedmorerapidly in Witt (1985) has provided an excellent review of the
thenon-alcohol cases. Moreover, McEntee(1993) has possible ways by which thiamine depletion might
pointed out that alcohol reduces both the absorption predispose to the characteristic pathological lesions
of thiamine and theactivity of theenzymewhich of the Wernicke-Korsakoff syndrome, although the
converts ittoitsactive form. precise mechanism involved is not known. TPP, the
Ithasbeensuggested thatnicotinic aciddeficiencyactive form of thiamine, appears to be involved in
may contribute tothedevelopment ofthedisorder DNA synthesis aswellasthree enzymatic reactions
in some cases (Jolliffe et a!, 1941; Lishman, 1981; which areessential forglucosemetabolismand
Ishii & Nishihara, 1981), although thissuggestion is neurotransmitter production; and the metabolic hetero
difficult toevaluate inviewofthefactthatallthe geneity of different brain regions might explain why
B group vitamins have generally been depleted (and some areas are more vulnerable to thiamine depletion
replaced) together. However, recent scares concerning than others. Six neurotransmitter systems are
anaphylaxis following Parentrovite administration affected by thiamine depletion, either by reduction
may mean thatinadequate replacement oftheother of TPP-dependent enzyme activity, or by direct
B vitamins willbecomemorecommon (cf. Serdaru structural damage, and four of these neuro
et a!, 1988). transmitters, acetylcholine, glutamate, aspartate, and
A genetic factor hasbeenpostulated toexplain gamma-amino butyric acid (GABA) are directly
why onlya minority ofheavydrinkers develop the related to glucose metabolism (Fig. 1).
syndrome, whichisfarless common thanthehepatic
orgastrointestinal complications ofalcohol abuse.
Acetyicholine
Transketolase isan enzymewhichrequires thiamine
pyrophosphate (TPP) as a co-factor, and Blass & Particular interest lies in the possible contribution
Gibson(1977) postulatedthat a hereditaryabnormalityof cholinergic depletion to the memory disorder of
of transketolase metabolismpredisposed some theKorsakoff syndrome, inviewoftheevidence that
alcoholics to theKorsakoffsyndrome.However, cholinergic blockade inhealthy subjectsmay induce
theirstudywas basedon onlyfourcases,and memoryimpairment(Kopelman&Corn, 1988),and
 @1@
THE KORSAKOFF SYNDROME 161

Thiamine thatcholinergic
depletion
produces
an anterograde
amnesia, similar to that seen in the Korsakoff
syndrome,butnotany retrograde
loss(Kopelman
Thiamine
pyrophosphate
(TPP) & Corn, 1988; Kopelman, 1992). More particularly,
Mayes et a! (1988), in a detailed autopsy study of
two Korsakoff patients, failed to findmore than
ITransketolase Pyruvate ketoglutarate minimalchangeswithin thebasalforebrain.
$ activityI +decarboxylase dehydrogenase
activity4a actMty
/ Glutamate/GABA

1@
I Acetyicholine,
glutamate,
aspartate,
GABA
Glutamateappearsto themain excitatory
transmitter tothepyramidal
neuro
neuronsofthehippo
campus, and GABA is the inhibitory neurotransmitter
4 turnover/synthesis
within those neurons (Shepherd, 1988). In addition,
and/or Bowen eta! (1992) have suggested that glutaminergic
Structural damage transmission may bethechief factor thatnormally
sustains theactivity of cortico-cortical neurons.
Glutamate activates the N-methyl-D-aspartate
Noradrenaline,serotonin
(NMDA) receptor complexes in the hippocampi,
whicharewidelythoughttoplaya critical rolein
t Acetyicholine,
glutamate,aspartate, GABA memory processes (Lynch & Baudry, 1988).Although
turnover/synthesis
thereisevidenceof hippocampalhypometabolism
intheKorsakoff
Fig. 1 Thiamine and neurotransmitter production (based on syndrome(Fazio eta!,1992), and
Witt, 1985). that glutaminergic depletion may contribute to the
memory deficit inAlzheimer's disease(Ellisoneta!,
thatacetylcholine isdepleted inAlzheimer's disease 1986;Hyman et a!,1987;Bowen, 1990),this
(Rossor et a!, 1984; Kopelman, 1986a). TPP neurotransmitter systemhas not,as yet,been
depletion wouldbeexpected toproducediminished implicated in the Korsakoff syndrome.
levels of acetyl-CoAor high-energy phosphates
(ATP, ADP, AMP), thereby resulting in depleted
Noradrenaline and serotonin
acetylcholine synthesis (Schenker et a!, 1980; Witt,
1985). Although the data concerning acetylcholine In a series of papers, McEntee and Mair have
levels following thiamine deficiency remain contro implicated thenoradrenergic systeminthememory
versial, a consistent finding is that acetylcholine disorder of the Korsakoff syndrome (McEntee &
turnoveris indeedreducedfollowingdietarythiamine Mair, 1978, 1979; McEntee et a!, 1984). Their
deficiency (Barclayeta!,1981),and thereissome argument has been based upon the findings of: (a)
evidence that inhibition of (TPP-dependent) pyruvate reduced 3 methoxy 4-hydroxy phenyiglycol (MHPG)
decarboxylase causes a measurable impairment of levels in the cerebrospinal fluid (CSF) of Korsakoff
acetylcholine synthesis (Gibson et a!, 1975). patients; (b)a strong correlation between MHPG
An alternative possible link between Korsakoff's levels and an index of memory impairment; and (c)
syndrome and the cholinergic system has been a small but statistically significant improvement in
mooted by Arendt et a!(1983), who reported a 47010 memory test performance on administering clonidine
reduction of the neuron count in the nucleus basalis to these patients. Similarly, O'Donnell et a! (1986)
of three Korsakoff brains at autopsy, thereby reported that one week's oral administration of
implicating involvement of the ascending cholinergic methylphenidate (an adrenergic agonist) produced
projections. This led Butters (1985) to propose that a statistically significant improvement in six
Korsakoff's syndrome primarily involves a cholinergic Korsakoff patients on a verbal recall task.
(basalforebrain), ratherthan a diencephalic, On the other hand, (a) Martin et a! (1984) failed
pathology. Studies of basal forebrain lesions in to replicate the finding of reduced MHPG; (b)the
traumatic and vascular casesseemtoconfirmthat MHPG - memory test correlation was obtained in
this region contributes to memory processes (Damasio a sampleof only nine patients and was not reported
eta!, 1985a,b; Salazar eta!, 1986; Irle&Markowitsch, in the enlarged series; and (c) O'Carroll et a! (1993)
1987; Phillips et a!, 1987; von Cramon & Schuri, have recently failed to replicate the findings in a
1992). However, psychopharmacological studies of clonidine trial, not obtaining any significant improve
cholinergic blockade in healthy subjects indicate ments in 18 Korsakoff patients across a number of
162 KOPELMAN

cognitive tests. Animal studies implicate the adrenergic non-Korsakoff alcoholics, and seven healthy
system in mechanisms mediating attention and controls, Shimamura eta! (1988) reported significant
arousal, rather than memory (Robbins, 1984; Sahgal, enlargement of the third ventricle and widening of
1984). Mair & McEntee (1983, 1986)themselves have the Sylvian fissures and the left frontal sulci in
acknowledged that the noradrenergic system may Korsakoff patients, relativeto healthycontrols.
operate by a general effect upon attentional and In general, non-Korsakoff alcoholic subjects had
arousal mechanisms. Although it is very plausible measures approximately midway between the control
that an attentional deficit may explain some of the and the Korsakoff groups. Within the Korsakoff
characteristic personality and behavioural disorders group, a computerised measure of frontal sulcal
seen in the Korsakoff syndrome (e.g. apathy), it enlargement was reported to show a statistically
seems unlikely that it would account for the ‘¿core' significant median rank correlation (R = —¿0.43) with
memory disorder, especially in view of the relative 12 memory measures.
preservation of primary or working memory (see In a much more thorough study of 38 alcoholic
below). Korsakoffpatients, 100 non-Korsakoffalcoholics
A new development was reported by Martin et a! and 50 control subjects, Jacobson & Lishman (1987,
(1989) who found an improvement in free recall 1990) found that Korsakoff patients as a group had
performance in five patients with an alcoholic wider third ventricles, larger lateral ventricles, and
Korsakoff syndrome, following administration of wider interhemispheric fissures than non-Korsakoff
fluvoxamine, a serotonin reuptake inhibitor. They alcoholics or healthy controls. However, sulcal and
argued that these improvements in memory correlated Sylvian fissure widths did not differ between
significantly with reductions in the levels of 5- Korsakoff patients and other alcoholics. A com
hydroxy-indole-acetic acid (5HIAA). This finding is puterised measure of enlargement of the anterior
broadly consistent with an earlier study in alcohol interhemispheric fissure correlated with general
treated volunteers by Weingartner et a! (l983a). intellectual decline (indicated by the differences
However, at least three studies have failed to fmd between estimated premorbid IQ, based on a reading
significant gain on cognitive tasks following the test, and current IQ). On the other hand, the degree
administration of 5-HT reuptake inhibitors to of third ventricular enlargement (which presumably
Alzheimer patients (Cutler et a!, 1985; Dehlin et a!, reflects thalamic and hypothalamic pathology) was
1985; Lawlor et a!, 1988), and the finding requires related to the severity of memory impairment
replication in Korsakoff patients. (indicated by the difference between current IQ and
a memory quotient). A comparison of subgroups of
patients with (a) a small anterior interhemispheric
Neuro-imaging studies
fissure and a large third ventricle, or (b)a larger
anterior interhemispheric fissure and a small third
Structural imaging
ventricle, produced differences in the degree of
Clinical and autopsy studies of the Wericke intellectual decline and memory impairment, signifi
Korsakoff syndrome have been complemented by in cant at a 0.06 level (Jacobson & Lishman, 1987).
vivo computerisedtomography (CT) or MRI investi Two case reports have indicated areas of CT scan
gations, the most important of which have been part hypodensity in the thalamus of Wernicke-Korsakoff
of larger studies of non-Korsakoff alcoholic patients. patients (McDowell & Le Blanc, 1984; Mensing et
Three CT scan studies have employed planimetric a!, 1984). Similarly, Shimamura eta! (1988) reported
or computerised measures of ventricular enlargement, a significant reduction in mean thalamic density in
and either rating scales or computerised measure Korsakoff patients, relative to healthy controls, and
ments of sulcal, interhemispheric and Sylvian fissure a significant median rank correlation between density
widening. Carlen et a! (1981) found that alcoholic and twelve memory measures within the Korsakoff
subjects (n = 93) showed significantly greater ventri group (R = 0.34). However, density measurements
cular enlargement and sulcal widening than a non are very prone to a number of important sources of
alcoholic control group, but that differences between artefact, which require careful control (Jacobson et
a Wericke-Korsakoff group (n = 25) and other a!, 1985). After controlling for these in his much
alcoholics were minimal and non-significant. Within larger study, Jacobson (1987; Jacobson & Lishman,
the Korsakoff group, there was a significant 1990) found that left and right thalamic density
correlation between sulcal widening and an overall measurements were indeed significantly reduced in
index of cognitive impairment, but other radiological! male Korsakoff patients, relative to non-Korsakoff
cognitive correlations were not impressive. In a small alcoholics (P< 0.05), but that there were only weak,
study of seven alcoholic Korsakoff patients, seven non-significant correlations with measures of memory
 THE KORSAKOFFSYNDROME 163
impairment. Furthermore, aftercontrolling for underlying aetiology. Finally, Martin et a! (1992)
artefacts, therewere no significant differenceshave recently examineda mixed group of ten
between Korsakoff patients, non-Korsakoff alcoholics Korsakoff and alcoholic dementia patients, finding
and healthy controls in terms of frontal density. metabolic disruption within anterior brain regions
Christie eta! (1988) conducted an MRI study and andright posterior whitematter, withhyperactivity
confirmed the presence of cortical atrophy in in cerebellar-cortical connections on FDG PET.
Korsakoff patients, but these authors obtained only
one significant correlation between measures of
Neuropsychology of the alcoholic Korsakoff
cortical atrophy and memory test scores. Squire et
syndrome
a! (1990) conducted an MRI study in four alcoholic
Korsakoff patients, finding abnormally small
General cognitive and frontal lobe deficits
mammillary nuclei, barely detectable by magnetic
resonance, whereas the temporal lobes, hippocampi, As mentionedabove,Jacobson& Lishman(1987)
and parahippocampal gyri were of normal size. showed a variable degree of impairment on a
These findings in the Korsakoff patients contrasted standard IQ test. Their findings suggested a con
with the MRI results in another group of amnesic tinuum between those patients who showed a dis
patients, who were considered to have probable proportionate impairment of anterograde memory
temporal lobe pathology. Finally, Jernigan et a! and those patients with a fairly severe, general
(1991) have reported widespread reductions in grey cognitive decline, suggestive of ‘¿alcoholic dementia',
matter volumes in Korsakoff patients, particularly even though those alcohol patients with clear clinical
in the anterior portion of the diencephalon, but evidence of generalised dementia had been excluded
also in medial temporal and orbito-frontal cortex. from their study. This calls into question the
definition of the alcoholic Korsakoff syndrome as
a disproportionate memory impairment, relative to
Functional imaging
other cognitive functions, given above. However,
Hata et a! (1987) used xenon-contrast computed Jacobson & Lishman (1987) employed the original
tomography to make three-dimensional measure version of the Wechsler Memory Scale (WMS;
ments of local cerebral blood flow in seven Wernicke Wechsler, 1945) as their measure of anterograde
Korsakoff patients (at a very early stage in their memory. As others have pointed out (Warrington,
disorder), ten non-Korsakoff alcoholics, and 17 1982; Weiskrantz, 1985; Brandt & Butters, 1986), the
healthy controls. They claimed that the most WMS is confounded by the inclusion of digit span
prominent reductions of local cerebral blood flow and mental control as determinants of the memory
in the Korsakoff patients were found in the hypo quotient, and most Korsakoff patients perform
thalamus and basal forebrain, although close within normal limits on those subtests. The revised
inspection of their data indicates that the findings Wechsler Memory Scale (WMS—R)avoids this
were equivocal, as only two healthy controls had problem, and tends to give substantially lower
measurements in these regions. Subsequently, Hunter quotients to Korsakoff and other amnesic patients
eta! (1989) used single photon emission computerised than the previous scale (Butters eta!, 1988). It seems
tomography (SPECT) in Korsakoff and Alzheimer quite likely that, had Jacobson & Lishman been able
patients, finding that the Korsakoff group showed a to use the WMS—R(which was not available at the
general trend towards reduced tracer uptake throughout time of their study) or some other, relatively ‘¿pure'
the cortex, except for the posterior temporal cortex, measure of anterograde memory, their patients
where regional cerebral blood flow was preserved. The would indeed have shown a disproportionate impair
impaired flow showed significant rank correlations ment of anterograde memory, although this remains
with clinical (CAMCOG) measures of orientation to be demonstrated. In that case, it would remain true
and recent memory (Roth et a!, 1986). Fazio et a! that the degree of generalised cognitive impairment
(1992) studied 11 patients with ‘¿pure amnesia', varies widely between individual Korsakoff patients,
including two Korsakoff patients, using fluoro as Jacobson & Lishman (1987) indicated, but the
deoxy-glucose (FDG) positron emission tomography neuropsychological fmdings would be consistent with
(PET), finding significant bilateral reductions in the clinical definition of the disorder (i.e. dis
metabolism in interconnected limbic-hippocampal proportionate memory impairment).
regions, including thethalamic nuclei,aswellasthe More specifically, there is evidence that Korsakoff
frontal basal cortex. This metabolic impairment did patients show deficits on tasks requiring visuo
not correspond closely to alterations in structural perceptive and visuospatial capacities, including the
anatomy, and appears to have arisen irrespective of digit-symbol substitution test, hiddenorembedded
164 KOPELMAN
figures tests, and various concept formation tests fluency task for categories and on a computerised
requiring sorting and discrimination of complex category sorting test. Kopelman (1989, 199la) found
visual stimuli (Oscar-Berman, 1973; Kapur & significant impairments on the FAS verbal fluency,
Butters, 1977; Butters & Cermak, 1980; Brandt the Nelson version of card-sorting, and on a cognitive
& Butters, 1986). These visuoperceptual deficits are estimates test. As discussed below, various aspects
also found in chronic alcoholics who are not clinically of the memory deficits found in Korsakoff patients
amnesic, and probably arise from atrophy in cortical have been attributed to frontal dysfunction (e.g
association areas (Brandt & Butters, 1986). Jacobson Moscovitch, 1982; Squire, 1982; Schacter eta!, 1984;
ci a! (1990) found that Korsakoff patients are Mayes cia!, 1985; Janowsky eta!, 1989; Shimamura
impaired, relative to other alcoholics and healthy ci a!, 1990; Kopelman, 199la).
controls, on an automated digit—symboltest, and
that female Korsakoff patients were also impaired
at a spatial orientation task. Memory deficits
There is also considerable evidence of frontal lobe
There are many studies of the pattern of impaired
or ‘¿executive'dysfunction in Korsakoff patients
and preserved memory function in the alcoholic
(Shimamura ci a!, 1988; Leng & Parkin, 1988;
Korsakoff syndrome, using concepts of memory such
Janowsky ci a!, 1989; Kopelman, 1989, 1991a;
as those illustrated in Fig. 2 (for review see Squire,
Jacobson ci a!, 1990; Shoqeirat ci a!, 1990; Joyce
1987; Kopelman, 1987a; Mayes, 1988; Baddeley,
& Robbins, 1991), consistent with the neuro-imaging
1990; Squire & Butters, 1992).
and autopsy findings of frontal lobe atrophy. For
example, Janowsky ci a! (1989) found impairment
Working memory
on the Wisconsin Card-Sorting test and on the
initiation and perseveration subscales of the Dementia Beyond the immediate sensory memory stores
Rating Scale (Mattis, 1976). Leng & Parkin (1988) (holding information for a matter of milliseconds),
also found impairment on the Wisconsin Card primary or working memory refers to the capacity
Sorting test, but not a cognitive estimates test. to hold small quantities of information for brief
Jacobson ci a! (1990) found impairment on a verbal periods of time (up to about 30 seconds) without

Remote
(anterograde) (retrogiade)
Sensory memory Primary(‘workIng@
memory Secondarymemory
Fig. 2 Current concepts in memory research.
 THE KORSAKOFF SYNDROME 165
active rehearsal (James, 1890; Baddeley, 1990). It or encoding of some types of semantic information,
is usually measured by performance on span tasks, it seems unlikely that this accounts for the severity
the ‘¿recency'
component of free recall, or measures of the memory disorder in the Korsakoff syndrome
of ‘¿short-term'forgetting (Baddeley, 1986, 1990). (McDowell, 1981). Similarly, although Korsakoff
There are many investigations showing that per patients show a disproportionate deficit in the
formance on span tests (verbal and non-verbal) is encoding of such aspects of context as temporal
preserved in the Korsakoff syndrome (e.g. Zangwill, order, spatial location, modality of presentation, and
1946; Baddeley & Warrington, 1970; Kopelman, source of information (e.g. Huppert & Piercy, 1976,
1985, 199lb). Performance on short-term forgetting 1978a; Schacter et a!, 1984; Meudell et a!, 1985;
tasks is much more variable, with some studies Kopelman, 1989; Shoqeirat & Mayes, 1991; Mayes
showing preserved ability (Baddeley & Warrington, et a!, 1991), such deficits may be incidental to the
1970) and some showing severe impairment (Butters memory impairment for ‘¿target'information, being
& Cermak, 1980). However, the consensus now evident in some patients but not others (Schacter et
appears to be that there is a variable pattern of a!, 1984; Shimamura & Squire, 1987; MacAndrew
performance by Korsakoff patients on both verbal & Jones, 1993). Moreover, although Korsakoff
and non-verbal tasks (Kopelman, 1985, l99lb; Mayes patients' performances on contextual memory tests
et a!, 1988; Leng & Parkin, 1989). Leng & Parkin often show statistically significant correlations with
(1989) suggested that this variability in performance their scores on ‘¿target'
memory tasks, the degree of
correlated with the degree of frontal lobe dys shared variance is usually relatively low (Kopelman,
function, whereas Kopelman (1991b, 1992) argued 1989; Pickering et a!, 1989; Shoqeirat & Mayes,
that the impairments on verbal and non-verbal short 1991). It may well be the case that contextual
term forgetting tasks were correlated with left information is particularly difficult to learn, but that
hemisphere and right-hemisphere cortical atrophy it does not represent the core component of the
respectively, as measured on CT scans. amnesic disorder (Baddeley, 1990; Kopelman, 1992).
The ‘¿long-term' forgetting of explicit memories
refers to the rate of decline in retrieving ‘¿target'
Exp!icit memory
information which had previously been learned to
Secondary memory (Fig. 2) refers to the capacity to an adequate level (Huppert & Piercy, 1976, l978b;
hold much larger quantities of information for Kopelman, 1992). In empirical studies, this involves
periods beyond a few seconds (James, 1890; ‘¿matching'the initial learning of amnesic patients
Baddeley, 1990). It is usually assumed that some to that of healthy controls, usually by prolonged
specific ‘¿encoding'
process or underlying physiological exposure of the stimulus material to the amnesic
mechanism of ‘¿consolidation'is required to store group. Using such a technique, several studies have
information in secondary memory (Milner, 1966; shown that rates of long-term forgetting are normal,
Butters & Cermak, 1980), and that appropriate once target information has been learned to an
retrieval cues may greatly facilitate its recall (Tulving adequate level for as long as ten minutes (Huppert
& Thomson, 1973; Roedigger & Blaxton, 1987). The & Piercy, 1978b; Squire, 1981; Kopelman, 1985;
‘¿explicit'
component of secondary memory refers to Martone eta!, 1986). Two possible qualifications to
those memories of which we are, or can be made, this finding are: (a) that all these studies have
consciously aware (Shimarnura, 1986; Schacter, involved recognition memory, as ‘¿matching'is
1987; Tulving & Schacter, 1990), and, by definition, extremely difficult to achieve on recall tasks; and (b)
‘¿explicit
memory' is severely impaired in Korsakoff that it is plausible that there are differences in
patients. It is plausible that this deficit results from forgetting rates over intervals shorter than ten
an underlying dysfunction in the process of physio minutes (Kopelman, 1992).
logical ‘¿consolidation',and this may have neuro
chemical and electrophysiological correlates (Meudell
Imp!icit memory
& Mayes,1982;Kopelman,l986a,1987a).
However,various psychological deficits havealso Implicit memory refers to learning of which the
been postulated, including an impairment in the subject is not consciously aware (Shimamura, 1986;
encoding ofsemantic (or‘¿meaningful') informationSchacter, 1987; Tulving & Schacter, 1990). It
(Cutting, 1978b;Butters & Cermak,1980)orinthe encompasses simple conditioning tasks, the acquisition
encoding of contextual information (Huppert & and retention of perceptuo-motor skills (‘procedural
Piercy, 1976, 1978a; Hirst, 1982; Mayes eta!, 1985; memory'), and ‘¿priming', which refers to the facili
Mayes, 1992; Parkin, 1992). While there does indeed tation of a particular response to a cue by an earlier
appear to be some degree of deficit in the processing stimulus. The cue may be either the same as the
166 KOPELMAN

earlier stimulus (‘repetitionpriming'), a fragment eta!, 1993). However, Korsakoff patients are impaired
of the earlier stimulus, (e.g. ‘¿word-completion at speeded tasks such as verbal fluency (Shoqeirat
priming'), or semantically related to the stimulus eta!, 1990; Kopelman, l991a); and amnesic patients
(‘semanticpriming'). There are many studies which often fail to learn the definitions of words which have
demonstrate that Korsakoff patients show preserved come into the language since the onset of their
capacity on simple conditioning and procedural amnesia (Gabrielli et a!, 1988) or the names of the
memory tasks (Brooks & Baddeley, 1976;Weizkrantz current Prime Minister or other public figures
& Warrington, 1979; Martone eta!, 1984) as well as (Kopelman, l986b).
at various priming tasks (Warrington & Weiskrantz,
1970; Graf et a!, 1984; Shimamura & Squire, 1984;
Retrograde amnesia
Schacter & Graf,1986;Tulving & Schacter, 1990).
Such findings are generally interpreted as demons Each of these components of memory potentially has
trating the existence of an ‘¿implicit'
memory system, a retrograde as well as an anterograde component
preserved in amnesia and mediated by cortical (Fig. 2), but this has been studied in detail only with
structures (priming) and/or subcortical structures respect to explicit memory. There is clearly an
(procedural memory), distinct from the impaired extensive retrograde memory loss in the Korsakoff
‘¿explicit'memory system, mediated by limbic syndrome, extending back several decades, as
diencephalic structures. Korsakoff patients also show Korsakoff (1889) himself noted. Modern neuro
preserved affective or evaluative memory responses psychological studies have confirmed that this
(Johnson et a!, 1985; Frith et a!, 1992), possibly retrograde memory loss extends back at least 25—30
mediated by Mishkin's (1982) amygdaloid circuit, as years (Albert et a!, 1979; Cohen & Squire, 1981;
opposed to the hippocampal/anterior thalamic Kopelman, 1989; Squire et a!, 1989b; Parkin et a!,
circuit implicated in explicit memory (see above). The 1990).This extensive retrograde loss includes memory
findings of preserved implicit memory processes in for remote public or ‘¿semantic' information, facts
the Korsakoff syndrome may well have implications about a patient's own life (‘personalsemantic
for the design of rehabilitation and retraining memory'), and autobiographical memory for incidents
programmes (Glisky eta!, 1986; Baddeley & Wilson, or events from the patient's past (Kopelman, 1989).
1994). All these aspects of retrograde memory show a tem
poral gradient, with relative sparing of the most
distant memories, and the gradient is significantly
Semantic memory
steeper than that seen in dementing disorders such
Semantic memory is a conglomerate term referring as Alzheimer's disease (Kopelman, 1989). The
to knowledge of language, concepts, and well relative sparing of early memories may result from
rehearsed facts (e.g. ‘¿Paris
is the capital of France'), their greater salience and rehearsal, such that they
independent of the recall of any particular episode have become assimilated within semantic memory
or incident specific in time and place (Tulving, 1972). (Cermak, 1984; Weiskrantz, 1985). Consequently,
Many researchers believe that, with increasing there are two possible reasons why Korsakoff
rehearsal, explicit memories of particular episodes patients have a steeper temporal gradient than
gradually become assimilated within the knowledge dementing patients: (a) semantic memory is relatively
based semantic memory system (Cermak, 1984; preserved in Korsakoff's patients, allowing better
Weiskrantz, 1985), implying a continuum between retrieval of early memories than in dementing
‘¿explicit'
and ‘¿semantic'
memory. However, there is patients; and (b)a progressive anterograde impair
also neuropsychological evidence of their dissociation, ment during the Korsakoff patients' periods of heavy
some patients showing a disproportionate loss of drinking may have made the loss of recent memories
personal or autobiographical memories (case 3 above; particularly severe.
Dalla Barba et a!, 1990; O'Connor et a!, 1992; Although confabulation may occur in retrieving
Ogden, 1993; Evans et a!, 1993), while others remote or autobiographical memories, it is relatively
manifest a relatively specific loss of semantic memory, rare except for fleeting intrusion errors (Victor et a!,
as in patients with ‘¿semanticdementia' secondary to 1971; Berlyne, 1972; Kopelman, 1987b), which may
atrophy or pathology in the ‘¿dominant' temporal lobe well represent a normal response to a weak or failing
(De Renzi et a!, 1987; Snowden et a!, 1989; Hodges memory (Kopelman, 1987b). In those relatively rare
et a!, 1992). In general, Korsakoff patients show cases where spontaneous and florid confabulation
relative preservation of performance on semantic persists beyond the initial confusional (Wernicke)
memory tests compared with Alzheimer or other stage, it may result from either a disorganised
dementia patients (Weingartner eta!, l983b; Perani and disinhibited retrieval of memories and their
 THE KORSAKOFFSYNDROME
associations (Baddeley & Wilson, 1986; Kopelman,
1987b)or from a deficit in ‘¿reality
monitoring' (Dalla
Barba, 1993). In either event, there is substantial
evidence that frontal lobe dysfunction (possibly
resulting from damage to orbital medial pathways)
may underlie ‘¿spontaneous confabulation' (Luria,
1976; Stuss et a!, 1978; Kapur & Coughlan, 1980;
Baddeley & Wilson, 1986; Kopelman, l987b), and
that frontal lobe atrophy and dysfunction are
characteristic in the Korsakoff syndrome.
However, it is more typical for Korsakoff patients
to show an impoverished, aspontaneous recollection
of remote memories than florid confabulation. This
aspontaneity is analogous to the ‘¿non-fluent'
pattern
of retrieval described by Baddeley & Wilson (1986)
and Della Sala et a! (1993) in other patients with
frontal lobe lesions (possibly involving the more
dorso-lateral regions). Several studies have shown
that the severity of the retrograde and anterograde
memory impairments are poorly correlated in this
group of patients (Shimamura & Squire, 1986;
Kopelman, 1989, 199lb; Parkin, 1991). While
performance on retrograde and anterograde memory
tests showed only 21% shared variance in one study,
there was a significant association between perfor
mance on frontal lobe tests and retrograde memory,
such that a regression equation on the basis of three
‘¿frontal'
tests predicted 68.5% of the variance on
retrograde memory tests (Kopelman, 1991b, 1992).
This suggested that frontal lobe dysfunction, resulting
from the known abnormalities in this region revealed
in neuro-imaging and autopsy studies, may contribute
to a failure in organising retrieval processes for
remote and autobiographical memories in Korsakoff
patients (Kopelman, 1992, 1993). Studies of patients
Table 1
Neuropsychology of Korsakoff's syndrome
Component
Primary/working memoryRelatively intact (span tests, ‘¿short
forgetting)‘Psychological' term'
encodingsemanticDeficits
probably insufficient to
amnesiacontextualDeficits account for
incidental to amnesia rather
feature?‘Physiological than core
manRetention/storage/forgettingIntact
consolidation'Difficult to assess in
accomplishedRetrievalDeficits if leaming
secondary to encoding!
consolidation (Meudell & Mayes,
1982)Priming/procedural
learning(skills)PreservedRetrograde
amnesiaExtensive impairment (25 years or
more) with ‘¿temporal
gradient'
167
with vascular lesions in the diencephalon, in whom
retrograde memory was either relatively spared
(Parkin & Hunkin, 1993; Parkin et a!, 1994) or
disrupted (Hodges & McCarthy, 1993) in the
presence of intact or impaired ‘¿frontal'
test function,
respectively, can be interpreted as broadly consistent
with this hypothesis.
Table 1 summarises the pattern of preserved
and impaired memory processes in the Korsakoff
syndrome.
Conclusions
At the end of the 19th century, Lawson and
Korsakoff both described a syndrome, commonly
but not always associated with alcohol abuse, in
which memory was profoundly impaired while other
cognitive functions remained intact. The modern
definition of the syndrome, provided by Victor et
a! (1971), is in terms of an impairment of “¿memory
and learning . . . out of all proportion to other
cognitive functions―, resulting from nutritional
(thiamine) depletion. At an early stage, the association
with the clinical features of Wericke's encephalo
pathy was recognised, but it now appears that a
history of Wericke's syndrome is not invariably
present, and that coma or an insidious onset are
alternative initial manifestations of the disorder.
Moreover, the characteristic neuropathology can
sometimes be found at autopsy in alcoholic subjects
who have never been diagnosed during life as having
the Korsakoff syndrome.
The characteristic pathological features of the
disorder are found in the paraventricular and peri
aqueductal grey matter. There is consensus that
lesions in either the thalamus or the mammillary
bodies, or both, are critical for the production of
memory impairment. Controversy has concerned the
relative importance of the thalamus and mammillary
bodies, whether the anterior or dorso-medial thalamus
is critical for memory disruption, and the way in
which the diencephalon interacts with the hippocampi
and basal forebrain in memory formation, but the
most recentstudies emphasise theimportance of
the circuit comprising the mammillary bodies, the
mammillo-thalamic tract, and the anterior thalamus.
Neuro-imaging studies have corroborated the finding
in autopsy studies that cortical atrophy, particularly
frontal atrophy, is also present in many cases.
A variety of causes of malnutrition or mal
absorption have been reported to produce either the
Wernicke or the Wernicke-Korsakoff syndrome.
The belief that a history of alcohol abuse is required
to produce a permanent memory impairment is not
well substantiated, although ‘¿pure' non-alcoholic
 168 KOPELMAN
cases have proved hard to find in recent times. The —¿ & WILSON, B. (1986) Amnesia, autobiographical memory,

common factor underlying the development of the and confabulation. In Autobiographical Memory (ed. D. C.
Rubin). Cambridge: Cambridge University Press.
Wericke-Korsakoff syndrome is almost certainly —¿ & —¿ (1994) When implicit learning fails: amnesia and
thiamine depletion; but why some subjects should the problem of error elimination. Neuropsychologia, 32, 53—68.
be especially vulnerable to thiamine depletion, and BARCLAY, L. L., GnuoN, G. E. & BLASS, J. P. (1981) Impairment
the precise manner by which it produces the of behaviour and acetylcholine in thiamine deficiency. Journal
of Pharmacology and Experimental Therapeutics, 217, 537-543.
characteristic Wernicke-Korsakoff lesions, are not Bv.ArrY,W. W., BAILY,R. C. & FISHER,L. (1989) Korsakoff-like
understood. One possibility is that thiamine deficiency amnesic syndrome in a patient with anorexia and vomiting.
produces a reduction in acetyicholine turnover and International Journal of Clinical Neuropsychology, 11, 55—65.
synthesis, although depletions in the noradrenergic BECKER, J. T., FURMAN, J. M. R., PANISSET, M., et a! (1990)
Characteristics of the memory loss of a patient with Wernicke
and serotonergic systems have also been mooted. Korsakoff's syndrome without alcoholism. Neuropsychologia,
However, psychopharmacological studies concur 28, 171—179.
with neuropsychological investigations in suggesting BENDER, L. & SCHILDER, P. (1933) Encephalopathia alcoholica:
that lesions in partially independent brain structures polioencephaJitis haemorrhagica superior of Wernicke. Archives
of Neurology and Psychiatry, 29, 990-1053.
and neurochemical systems may underlie anterograde BERLYNE,N. (1972) Confabulation. British Journal of Psychiatry,
and retrograde amnesia. There is now general 120, 31—39.
agreement concerning the pattern of the memory BLASS,J. P. & GIIISoN,0. E. (1977) Abnormality of a thiamine
deficit in the Korsakoff syndrome, involving a severe requiring enzyme in patients with Wenrncke-Korsakoff syndrome.
New England Journal of Medicine, 297, 1367-1370.
impairment of new learning and an extensive retro Bowse@,D. M. (1990) Treatment of Alzheimer's disease: molecular
grade loss, with intact or well-preserved working pathology versus neurotransmitter-based therapy. British Journal
memory, priming, procedural memory and the rate of Psychiatry, 157, 327—330.
of ‘¿long-term'forgetting. The present paper has —¿, FRANCIS, P. T., PANGALOS, M., et al (1992) Treatment

strategies for Alzheimer's disease. Lancet, 339, 132—133.

suggested that, while structural lesions or neuro Bowswi, K. M., 000DHART, R. & Jow@, N. (1939) Observations
chemical deficiencies involving the limbic-diencephalic on the role of vitamin Bl in the aetiology and treatment of
circuits produce anterograde amnesia, some other Korsakoff psychosis. Journal of Nervous and Mental Disease,
factor, such as frontal lobe dysfunction, must account 90, 569—575.
Bwesr, J. & Burreas, N. (1986) The alcoholic Wernicke-Korsakoff
for the severe retrograde memory loss, which syndrome and its relationship to long-term alcohol abuse. In
commonly extends back 25 years or more in this Neuropsychological Assessment of Neuropsychiatric Disorders
disorder. (ads I. Grant & K. M. Adams), pp. 441—477.Oxford: Oxford
University Press.
BRoOKS, D. N. & BADDELEY,A. D. (1976) What can amnesic
Acknowledgements patients learn? Neuropsychologia, 14, 111—122.
The author is grateful to the Wellcome Trust for grants at the Burress, N. (1985) Alcoholic Korsakoff's syndrome: some un
Section of Neuropsychiatry at the Institute of Psychiatry. The resolved issues concerning etiology, neuropathology, and
author is now supported by the West Lambeth Community Care cognitive deficits. Journal of Clinical and Experimental Neuro
Trust, and he is also very grateful to Miss Claire Hook for patiently psychology, 7, 181—210.
—¿ & CERMAK, L. S. (1980) Akoholic Korsakoff's Syndrome:
typing the manuscript.
An Information-Processing Approach to Amnesia. New York:
AcademicPress.
References —¿, SALMON, D. P., MUNRO CULLUM, C., et al (1988)

Differentiation of amnesic and demented patients with the

ALBERT. M. S., Btrrrras, N. & LEvIN, J. (1979) Temporal gradients Wechsler Memory Scale —¿Revised. Clinical Neurologist, 2,
in the retrograde amnesia of patients with alcoholic Korsakoff's 133—148.
disease. Archives of Neurology, 36, 211—216. CAMPBELL, A. C. P. & RUSSELL, W. R. (1941) Wernicke's en
ALEXANDER, L. (1940) Wernicke's disease: identity of lesions cephalopathy: the clinical features and their probable relationship
produced experimentally by Bl avitaminosis in pigeons with to vitamin B deficiency. Quarterly Journal of Medicine, 34,
haemorrhagic polioencephalitis occurring in chronic alcoholism 41—64.
inman.AmericanJournal ofPathology, 16,61—70. CARLEN, P. L., WILKINSON, D. A., WORTZMAN, 0., et al (1981)
PIJOAN, M. & MYERSON, A. (1938) Ben-ben and scurvy. Cerebral atrophy and functional deficits in alcoholics without
Transactions of the American Neurological Association, 64, apparent liver disease. Neurology, 31, 377—385.
135—139. CASTAIGNE,P., LHERMITrE, F., BUGE, A., et al (1981) Paramedian
ARENDT, T., BIOL, V., ARENDT, A., et a! (1983) Loss of neurons thalamic and midbrain infarcts: clinical and neuropathological
in the nucleus basalis of Meynert in Alzheimer's disease, paralysis study. Annals of Neurology, 10, 127-148.
agitans and Korsakoff's disease. Acta Neuropathologica, 61, CERMAK, L. S. (1984) The episodic-semantic distinction in amnesia.
101—108. In The Neuropsychology of Memory (ads L. R. Squire & N.
BADDELEY, A. D. (1986) Working Memory. Oxford: Clarendon Butters). New York: Guilford Press.
Press. CHRISTIE, i. E., KEAN, D. M., DOUGLAS, R. H. B., et al (1988)
—¿ (1990) Human Memory: Theory and Practice. London: Magnetic resonance imaging in presenile dementia of the
Lawrence Erlbaum. Alzheimer-type, multi-infarct dementia, and Korsakofrs
—¿ & WARRINGTON, E. K. (1970) Amnesia and the distinction syndrome. Psychological Medicine, 16, 3 19-329.
between long- and short-term memory. Journal of Verbal COHEN, N. J. & SQUIRE, L. R. (1981) Retrograde amnesia and
Learningand VerbalBehaviour,9, 176—189. remote memory impairment. Neuropsychologia, 19, 337—356.
 THE KORSAKOFF SYNDROME 169
CRAvIOTO,H., KOREIN,J. & SILBERMAN,J. (1961) Wernicke's GENTILINI, M., Du Reuzj, E. & Cam, 0. (1987) Bilateral paramedian
encephalopathy. Archives of Neurology, 4, 54—63. thalamic artery infarcts: report of eight cases. Journal of
CRUIKSHANK, E. K. (1950) Wernicke's encephalopathy. Quarterly Neurology, Neurosurgery and Psychiatry, 50, 900-909.
Journal of Medicine, 19, 327—338. GIBSON,0. E., JosE, R. & B@ss, J. P. (1975) Decreased synthesis
CUTLER, N. R., HAXBY, J., KAY, A. D., et a! (1985) Evaluation of of acetylcholine accompanying impaired oxidation of pyruvic
zimelidine in Alzheimer's disease. Archives of Neurology, 42, acid in rat brain minces. Biochemical Journal, 148, 17—23.
744—748. GLISKY, E. L., SCHACTER, D. & TULVING, E. (1986) Computer
Curnit@o,J. (l978a) The relationship between Korsakoff's syndrome learning by memory impaired patients: acquisition and retention
and “¿alcoholic
dementia―.British Journal of Psychiatry, 132, of complex knowledge. Neuropsychologia, 24, 313—328.
240—251. GRAF, P., SQUIRE,L. R. & MANDLER,0. (1984) The information
(l978b) A cognitive approach to Korsakofrs syndrome. that amnesic patients do not forget. Journal of Experimental
Cortex, 14, 485—495. Psychology: Learning, Memory and Cognition, 11, 501-5 18.
DALLABARBA,0. (1993) Confabulation: knowledgeand recollective GRAFF-RADFORD, N. R., DAMASIO,H., YAMADA,T., et a! (1985)
experience. Cognitive Neuropsychology, 10, 1-20. Non-haemorrhagic thalamic infarction: clinical, neurological,
Cwoiorri, L. & DENES, 0. (1990) Autobiographical and electrophysiological findings in four anatomical groups
memory loss and confabulation in Korsakoff's syndrome:a case defined by computerised tomography. Brain, 408, 485-516.
report. Cortex, 26, 525-534. TRANEL, D., VAN HOESEN, 0. W., eta! (1990) Diencephalic
DAMASIO, A. R., GRAFF-RADFORD, N. R., ESLINGER, P. J., et a! amnesia. Brain, 113, 1—26.
(l985a) Amnesia following basal forebrain lesions. Archives of GUBERMAN, A. & S-russ, D. (1983) The syndrome of bilateral
Neurology, 42, 263—271. paramedian thalamic infarction. Neurology, 33, 540-546.
ESLINGER, P. J., DAP.IASI0, H., et a! (l985b) Multimodal HARpER,C. 0. (1979) Wernicke's encephalopathy: a more common
amnesic syndrome following bilateral temporal and basal disease than realised: a neuropathological study of 51 cases.
forebrain damage. Archives of Neurology, 42, 252—259. Journalof Neurology,NeurosurgeryandPsychiatry,42,226-231.
D'ASSIGNY,M. (1697) The art of memory: a treatise useful for such (1980)Sudden, unexpected death and Wernicke's encephalo
as areto speakin publick. In ThreeHundred Yearsof Psychiatry, pathy: a complication of prolonged I/V feeding. Australian and
(eds R. Hunter & I. Macalpine, 1963), pp. 262—264.Oxford: New Zealand Journal of Medicine, 10, 230-235.
Oxford University Press. (1983) The incidence of Wernicke's encephalopathy in
DEHLIN,0., HEDENRUD,B., JANSSON,P., et al (1985) A double Australia: a neuropathological study of 131 cases. Journal of
blind comparison of alaproclate and placebo in the treatment Neurology, Neurosurgery and Psychiatry, 46, 593—598.
of patients with senile dementia. Acta Psychiatrica Scandinavica, KRIL, J. & DALY, J. (1987) Are we drinking our neurones
71, 190—196. away? British Medical Journal, 294, 534—536.
DELLA SALA, S., LAIACONA, M., SPINNLER, H., et a! (1993) & (1988) Brain shrinkage in alcoholics is
Impaired autobiographical recollection in some frontal patients. not caused by changes in hydration: a pathological study. Journal
Neuropsychologia,31, 823-840. of Neurology, Neurosurgeryand Psychiatry, 51, 124—127.
DE RENzI, E., Llorn, M. & NICHELU,P. (1987) Semantic amnesia HATA, T., MEYER, J. S., TANAHASHI,N., et al (1987) Three
with preservation of autobiographical memory. Cortex, 23, dimensional mapping of local cerebral perfusion in alcoholic
575—597. encephalopathy with and without Wernicke-Korsakoff
DEVATHASON,
0. & KOH, C. (1982) Wernicke's encephalopathy in syndrome. Journal of CerebralBlood Flow and Metabolism, 7,
prolonged fasting. Lancet, 2, 1108—1109. 35—44.
DE WARDENER,H. E. & LENNOX, B. (1947) Cerebral beriberi HIRST, W. (1982) The amnesic syndrome: descriptions and ex
(Wernicke's encephalopathy): review of 52 cases in a Singapore planations. Psychological Bulletin, 91, 435—460.
PoW hospital. Lancet, 1, 11—17. HOOO1tS,J. R., PATrERSON,K., OXBURY,S., et a! (1992) Semantic
DUSOIR, H., KAPUR, N., BYRNES,D. P., et a! (1990) The role dementia. Progressivefluentaphasiawith temporallobe atrophy.
of diencephalic pathology in human memory disorder: Brain, 115, 1783—1806.
evidence from a penetrating paranasal injury. Brain, 113, & MCCARThY,R. A. (1993) Autobiographical amnesia
1695—1706. resulting from bilateral paramedian thalamic infarction. A case
EBELS, E. J. (1978) How common is Wernicke-Korsakoff study in cognitive neurobiology. Brain, 116, 921—940.
syndrome? Lancet, 2, 781—782. HOOKE, R. (1680) Lectures of light. In Three Hundred Years of
ELLISON, D. W., BEAL, M. F., MAZUREK, M. F., et a! (1986) A Psychiatry: 1535-1860(eds R. Hunter & I. Macalpine, 1963).
post-mortem study of aminoacid neurotransmitters in Oxford: Oxford University Press.
Alzheimer's disease. Annals of Neurology, 20, 616-621. HUNTER,R., MCLUSKIE,R., WYPER,D., et a! (1989) The pattern
EVANS, J., WILSON, B., WRIGHT, P. W., et a! (1993) Neuro of function-related regional cerebral blood flow investigated by
psychological and SPECT scan findings during and after single photon emission tomography with @“¿Tc-HMPAO in
transient global amnesia: evidence for the differential impairment patients with presenile Alzheimer's disease and Korsakoff's
of remote episodic memory. Journal of Neurology, Neurosurgery psychosis. Psychological Medicine, 19, 847—856.
and Psychiatry, 56, 1227—1230. HUPPERT,F. A. & PIERCY, M. (1976) Recognition memory in
FAZI0, F., PERANI, D., GILARDI, M. C., et a! (1992) Metabolic amnesic patients: effect of temporal context and familiarity of
impairment in human amnesia: a PET study of memory material. Cortex, 12, 3-20.
networks. Journal of Cerebral Blood Flow and Metabolism, 12, & (l978a) The role of trace strength in recency and
353—358. frequency judgements by amnesic and control subjects. Quarterly
Fseutm, 0. (1973) Chronic central nervous system toxicity in Journalof ExperimentalPsychology,30, 347-354.
alcohol. Annual Review of Pharmacology, 13, 217-227. & (l978b) Dissociation between learning and
FIUm, C. D., CAHILL,C., RIDLEY,R. M., et a! (1992) Memory for remembering in organic amnesia. Nature, 275, 317—318.
what it is and memory for what it means: a single case of HYMAN, B. T., VAN HOESEN,0. W. & DAMASI0, A. R. (1987)
Korsakoff's amnesia. Cortex, 28, 53—67. Alzheimer's disease: glutamate depletion in the hippocampal
GABRIELLI, J. D. E., CoHEN, N. J. & CORKIN, 5. (1988) The perforant pathway zone. Annals of Neurology, 22, 37—40.
impaired learning of semantic knowledge following bilateral IRLu, E. & MARKOWITSCH, H. J. (1987) Basal forebrain-lesioned
medial temporal-lobe resection. Brain and Cognition, 7, monkeys are severely impaired in tasks of association and
157—177. recognition memory. Annals of Neurology, 22, 735-743.
170 KOPELMAN
Isim, N. & NISHIHARA, Y. (1981) Pellagra among chronic alcoholics - (1992)The ‘¿new'
and the ‘¿old':
components of the anterograde
clinical and pathological study of 20 necropsy cases. Journal of and retrograde memory loss in Korsakoff and Alzheimer
Neurology, Neurosurgery and Psychiatry, 44, 209-215. patients. In The Neuropsychology of Memory (2nd edn) (eds L.
JAcoBSON, R. R. (1987) Clinical, CAT Scan and Psychometric R. Squire & N. Butters), pp. 130-146. New York: Guilford.
Studies in Korsakoff's Psychosis. MD Thesis, University of (1993) The neuropsychology of remote memory. In
Cambridge. Handbook of Neuropsychology, Vol. 8 (eds H. Spinner & F.
TURNER,S. W., BALDY,R. E., et a! (1985) Densitometric Boller), pp. 215—238.Amsterdam: Elsevier.
analysis of scans: important sources of artefact. Psychological & CORN, T. H. (1988) Cholinergic ‘¿blockade'as a model
Medicine, 15, 879—889. for cholinergic depletion: a comparison of the memory deficits
& LISHMAN,W. (1987) Selective memory loss and global with those of Alzheimer-type dementia and the alcoholic
intellectual deficits in alcoholic Korsakoff's syndrome. Korsakoff syndrome. Brain, 111, 1079—1110.
Psychological Medicine, 17,649—655. KORSAKOFF, 5. 5. (1887) Disturbance of psychic function in
& (1990) Cortical and diencephalic lesions in alcoholic paralysis and its relation to the disturbance of the
Korsakoff's syndrome: a clinicaland Cf scan study. Psychological psychic sphere in multiple neuritis of non-alcoholic origin.
Medicine, 20, 63—75. Quoted by Victor, M., Adams, R. D. & Collins, 0. H. (1971)
ACKER,C. & LISHMAN,W. A. (1990) Patterns of neuro The Wernicke-Korsakoff Syndrome. Oxford: Blackwell.
psychological deficit in alcoholic Korsakoffs syndrome. (1889a) Psychic disorder in conjunction with peripheral
PsychologicalMedicine, 20, 321—334. neuritis (trans. M. Victor & P. I. Yakovlev, 1955). Neurology,
JAMFS, W. (1890) Principles of Psychology, Vol. 1. New York: Holt. 5, 394—406.
JANOWSKY, J. S., SHIMAMURA, A. P., KRITCHEVSKY, M., et a! (1989) (1889b) Etude médico-psychologiquesur une forme des
Cognitive impairment following frontal lobe damage and its maladies de Ia mémoire.Revue Phiosophie, 20, 501—530.
relevance to human amnesia. Behavioural Neuroscience, 103, KRAEPELIN, E. (1910) Psychiatric (8th edn). Leipzig: Barth.
548—560. LAWLOR, B. A., MELWw, A. M., SUNDERLAND, 1., et a! (1988) A
JERNIGAN, T. L., SCHAFER, K., BUSTERS, N., et a! (1991) Magnetic pilot study of serotonergic system responsivity in Alzheimer's
resonance imaging of alcoholic Korsakoff patients. Neuro disease. Psychopharmacology Bulletin, 24, 127—129.
psychopharmacology, 4, 175-186. LAWSON,R. (1878) On the symptomatology of alcoholic brain
JOHNSON, M. K. & KIM, J. K. (1985) Do alcoholic Korsakoff's disorders. Brain, 1, 182—194.
syndrome patients acquire affective reactions? Journal of LEIGH, D., MCBURNEY,A. & MCILWAIN,H. (1981) Erythrocyte
Experimental Psychology: Learning, Memory and Cognition, 11, transketolase activity in the Wernicke-Korsakoff syndrome.
22—36. British Journal of Psychiatry, 139, 153—156.
JOLIIFFE,N., WORTIS,H. & FEIN, H. D. (1941) The Wernicke LENG, N. R. C. & PARKIN,A. J. (1988) Double dissociation of
syndrome. Archives of Neurology and Psychiatry, 46, 569—597. frontal dysfunction in organic amnesia. British Journal of
JoYcE, E. M. & ROBBINS, T. W. (1991) Frontal lobe function in Clinical Psychology, 27, 359—362.
Korsakoff and non-Korsakoff alcoholics: planning and spatial & (1989) Aetiological variation in the amnesic
working memory. Neuropsychologia, 29, 709-723. syndrome: comparisons using the Brown-Peterson task. Cortex,
JUNG, E-H., REX SHEU, K-F. & BLASS, J. P. (1993) An enzymatic 25, 251—259.
and immunochemical analysis of transketolase in fibroblasts Lusw@AN, W. A. (1981) Cerebral disorder in alcoholism: syndromes
from Wernicke-Korsakoff syndrome. Journal of Neurological of impairment. Brain, 104, 1-20.
Sciences,
114,123—127. LUDA, E. (1980) Wernicke's encephalopathy. Archives of Neurology,
KAPUR,N. & BUTTERS,N. (1977) Visuoperceptive deficits in long 37, 255.
term alcoholics with Korsakoff's psychosis. Journal of Studies LURIA, A. R. (1976) The Neuropsychology of Memory. New York:
on Akohol, 38, 2025-2035. John Wiley.
& COUGHLAN,A. K. (1980) Confabulation and frontal lobe LYNCH,0. & BAUDRY,M. (1988) Structure-function relationships
dysfunction. Journal of Neurology, Neurosurgery and Psychiatry, in the organization of memory. In Perspectives in Memory
43, 461—463. Research (ed. M. S. Gazzaniga), pp. 23-91. Cambridge, MA:
KATZ, D. I., ALEXANDER,M. P. & MANDELL, A. M. (1987) Massachusetts Institute of Technology Press.
Dementia following strokes in the mesencephalon and di MACANDREW, S. B. 0. & Jopms, 0. V. (1993) Spatial memory in
encephalon. Archives of Neurology, 44, 1127-1133. amnesia evidence from Korsakoff patients. Cortex, 29,235-250.
KOPELMAN, M. D. (1985) Rates of forgetting in Alzheimer-type MAIR, R. 0. & MCENTEE,W. J. (1983) Korsakoff's psychosis:
dementia and Korsakoff's syndrome. Neuropsycho!ogia, 23, noradrenergic systems and cognitive impairment. Behavioural
623—638. Brain Research,9, 1—32.
(1986a) The cholinergic neurotransmitter system in human & (1986) Cognitive enhancement in Korsakoff's
memory and dementia: a review. Quarterly Journal of Experi psychosis by clonidine: a comparison with L-dopa and ephedrine.
mental Psychology, 38A, 535-573. Psychopharmacology, 88, 374-380.
(1986b) Clinical tests of memory. British Journal of MAIR, W. 0. P., WARRINGTON,E. K. & WEISKRANTZ,L. (1979)
Psychiatry,148, 517—525. Memory disorder in Korsakoff's psychosis: a neuropathological
(l987a) Amnesia: organic and psychogenic. British Journal and neuropsychological investigation of two cases. Brain, 102,
of Psychiatry,150, 428-442. 749—783.
(1987b)Two typesof confabulation. Journal of Neurology, MALAMUD,N. & SICIWC0RN,S. A. (1956) Relationship between the
Neurosurgery and Psychiatry, 50, 1482—1487. Wernicke and Korsakoff syndrome. Archives of Neurology and
(1989) Remote and autobiographical memory, temporal Psychiatry,76, 585—596.
context memory, and frontal atrophy in Korsakoff and MARKOWITSCH, H. J. (1984) Can amnesia be caused by damage of
Alzheimer patients. Neuropsychologia, 27, 437—460. a single brain structure? Cortex, 20, 27-45.
(l991a) Frontal lobe dysfunction and memory deficits in MARTIN, P. R., WEINGARTNER, H., GORDON, E. K., et a! (1984)
the alcoholic Korsakoff syndrome and Alzheimer-type dementia. Central nervous system catecholamine metabolism in Korsakoffs
Brain, 114, 117—137. psychosis. Annals of Neurology, 15, 184—187.
(l99lb) Non-verbal, short-term forgetting in the alcoholic —¿, ADINOFF, B., EcVAP.Dr, M. J., et a! (1989) Effective

Korsakoff syndrome and Alzheimer-type dementia. pharmacotherapy of alcoholic amnesic disorder with fluvoxamine.
Neuropsychologia, 29, 737-747. Archives of General Psychiatry, 46, 617—621.
 THE KORSAKOFFSYNDROME 171

RIO, D., ADINOFF, B., et a! (1992) Regional cerebral MILNER, B. (1966) Amnesia following operation on the temporal
glucose utilization in chronic organic mental disorders associated lobes. In Amnesia (eds C. W. M. Whitty & 0. Zangwill),
with alcoholism. Journal of Neuropsychiatry and Clinical pp. 109-133. London: Butterworths.
Neurosciences, 4, 159—167. MISHKIN, M. (1978) Memory in monkeys is severely impaired by
MARTONE,M., BurrEIts, N., PAYNE,M., eta! (1984) Dissociations combined but not by separate removal of amygdala and hippo
between skill learning and verbal recognition in amnesia and campus. Nature, 273, 297-298.
dementia. Archives of Neurology, 41, 965—970. —¿ (1982) A memory system in the monkey. Philosophical
& TRAWlER,D. (1986) Some analyses of forgetting Transactions of the Royal Society, London, B298, 85-95.
of pictorial material in amnesic and demented patients. Journal MOLL, J. M. (1915) The amnesic or Korsakoffs syndrome with
of Clinical and Experimental Neuropsychology, 8, 161-178. alcoholic aetiology: an analysis of 30 cases. Journal of Mental
MATrIS,5. (1976) Dementia Rating Scale. In Geriatric Psychiatry / Science,61, 423—437.
(eds R. Bellack & B. Karasu). New York: Grune & Stratton. MORE, E., YAMADORIS,A. & MITANI, Y. (1986) Left thalamic
MAYES, A. R. (1988) Human Organic Memory Disorders. infarction and disturbance of verbal memory: a dlinicoanatomical
Cambridge: Cambridge University Press. study with a new method of computed tomographic stereotaxis
(1992) What are the functional deficits that underlie lesion localisation. Annals of Neurology, 20, 671—676.
amnesia? In The Neuropsychology of Memory (2nd edn) MOSCOVITCH,M. (1982) Multiple dissociation of function in
(eds L. R. Squire & N. Butters), pp. 23—35.New York: amnesia. In Human Memory andAmnesia (ed. L. S. Cermak),
Guilford. pp. 337-370. Hillsdale, NJ: Lawrence Erlbaum.
MEUDELL, P. R. & PICKERING, A. (1985) Is organic MURRAY, E. A., GAFFAN, D. & MISHKIN, M. (1993) Neural
amnesia caused by a selective deficit in remembering contextual substrates of visual stimulus-stimulus association in rhesus
information? Cortex, 21, 167—202. monkeys. Journal of Neuroscience, 13, 4549-4561.
MANN, D., et a! (1988) Location of lesions in NIGHTINGALE,
S., BATES,D., Hasm, P. D., eta! (1982) Wernicke's
Korsakoffs syndrome: neuropsychological and neuropathological encephalopathy in hyperemesisgravidarum. Postgraduate Medical
data on two patients. Cortex, 24, 367-388. Journal,58, 558—559.
& MACDONALD,C. (1991) Disproportionate inten NIXON, P. F., KACZMAREK,M. J., TATE, J., et a! (1984) An
tional spatial memory impairments in amnesia. Neuropsychologia, erythrocyte transketolase isoenzyme pattern associated with the
29, 771—784. Wernicke-Korsakoff syndrome. European Journal of Clinical
MCCOOL, B. A., PwNK, S. 0., MAR1N, P. R., et a! (1993) Cloning Investigation,14, 278—281.
of human transketolase cDNAs and comparison of the nucleotide O'CARROLL, R. E., MoFFooT, A., EBMEIER, K. P., et a! (1993)
sequence of the coding region in Wernicke-Korsakoff and non Korsakoff's syndrome, cognition and clonidine. Psychological
Wernicke-Korsakoff individuals. Journal of BiologicalChemistry, Medicine,23, 341—348.
268, 1397—1404. O'CoNNoR, M., BUTTERS, N., MILI0rIS, P., et a! (1992) The
MCDOWELL, J. (1981) Effects of encoding instructions on recall and dissociation of anterograde and retrograde amnesia in a patient
recognition in Korsakoff patients. Neuropsycho!ogia, 19, with herpes encephalitis. Journal of Clinical and Experimental
43—48. Neuropsychology, 14,159—178.
& Lit BLANC,H. J. (1984) Computed tomographic findings O'DoNNELL, V. M., Prrrs, W. M. & F@, W. E. (1986) Nor
in the Wericke-Korsakoff syndrome. Archives of Neurology, adrenergic and cholinergic agents in Korsakoffs syndrome.
41, 453—454. ClinicalNeuropharmacology,
9, 65—70.
MCENrEE, W. J. (1993) The effect of ethanol on thiamine OGDEN, J. A. (1993) Visual object agnosia, prosopagnosia,
absorption and metabolism: a possible explanation for so-called achromatopsia, loss of visual imagery, and autobiographical
alcoholic dementia. British Association of Psychopharmacology amnesia following recovery from cortical blindness: case MH.
Annual Meeting Abstracts, p. A6. Neuropsychologia, 31, 571—589.
—¿, BIBER, M. P., PERL, D. P., et a! (1976) Diencephalic OSCAR-BERMAN, M. (1973) Hypothesis testing and focusing behavior
amnesia:reappraisal.Journalof Neurology,Neurosurgery
and during concept formation for amnesic Korsakoff patients.
Psychiatry,39, 436—441. Neuropsychologia, II, 191-198.
—¿ & MAIR, R. 0. (1978) Memory impairment in Korsakoff's PARKIN,A. J. (1991) Recent advances in the neuropsychology of
psychosis: a correlation with brain noradrenergic activity. memory. In Memory: Neurochemical andAbnormal Perspectives
Science, 202, 905-907. (eds J. Weinman & J. Hunter), pp. 141—162.London:
& (1979) Memory enhancement in Korsakoff's Harwood.
psychosis by clonidine: further evidence for a noradrenergic —¿ (1992) Functional significance of aetiological factors in
deficit. Annals of Neurology, 7, 466-470. human amnesia. In The Neuropsychology of Memory (2nd edn)
—¿, —¿ & LANGLAIS, P. J. (1984) Neurochemical pathology (eds L. R. Squire & N. Butters), pp. 122-129. New York:
in Korsakoffs psychosis: implications for other cognitive Guilford.
disorders. Neurology, 34, 648-652. —¿, MONTALDI, D., LENG, N. R. C., et a! (1990) Contextual

Meu5ING, J. W. A., Ho0GLAND,P. H. & SUDoF,J. L. (1984) cueing effects in the remote memory of alcoholic Korsakoff
Computed tomography in the diagnosis of Wernicke's patientsand normal subjects. QuarterlyJournalof Experimental
encephalography - a radiological-neuropathological correlation. Psychology,42A, 585—5%.
Annals of Neurology, 16, 363-365. —¿, BLUNDEN, J., Rents, J. E., ci a! (1991) Wernicke -Korsakoff

MEUDELL, P. & MAYES, A. R. (1982) Normal and abnormal syndrome of nonalcoholic origin. Brain and Cognition, 15,
forgetting: some comments on the human amnesic syndrome. 69—82.
In Normality and Pathology in Cognitive Functions (ed. A. W. —¿ & HUNKIN, N. M. (1993) Impaired temporal context
Ellis), pp. 203-208. London: Academic Press. memory on anterograde but not retrograde tests in the absence
—¿, OSTERGAARD, A., et a! (1985) Recency and of frontal pathology. Cortex, 29, 267-280.
frequency judgements in alcoholics amnesics and normal people —¿, REES, J. E., HUNKIN, N. M., et a! (1994) Impairment of

with poor memory. Cortex, 21, 487-511. memory followingdiscrete thalamic infarction. Neuropsycho!ogia,
MEUNIER, M., BACHEVALIER, J., MISHKIN, M., et a! (1993) Effects 32, 39—52.
on visual recognition of combined and separate ablations of the PENTLAND,B. & MAWDSLEY,
C. (1982) Wernicke's encephalopathy
entorhinal and perirhinal cortex in rhesus monkeys. Journal of following ‘¿hunger
strike'. Postgraduate Medical Journal, 58,
Neuroscience, 13, 5418—5432. 427—428.
 172 KOPELMAN
PERANI, D., BRESSI,S., CAPPA, S. F., et a! (1993) Evidence of Performance on tests
sensitive
tofrontal lobelesions bypatients
multiple memory systems in the human brain. A [‘8FJ FDG with organicamnesia:Langand Parkin revisited.British Journal
PET metabolic study. Brain, 116, 903—920. of Clinical Psychology, 29, 401-408.
PHILLIPS, S., SANGALANG, V. & STERNS, 0. (1987) Basal forebrain —¿ & —¿ (1991) Disproportionate incidental spatial memory
infarction —¿
a clinicopathological correlation. Archives of and recall deficits in amnesia. Neuropsycho!ogia, 29, 749—770.
Neurology, 44, 1134—1138. SNOWDEN, J. S., GOULDING, P. J. & NEARY, D. (1989) Semantic
PICKERING, A., MAYES, A. R. & FAIRBAIRN, A. F. (1989) Amnesia dementia: a form of circumscribed cerebral atrophy. Behavioural
and memory for modality information. Neuropsycho!ogia, 27, Neurology,2, 167—182.
1249—
1259. SPEEDIE,L. J. & HEILMAN,K. M. (1982) Amnesic disturbance
RIGGS, H. E. & BOLES, R. S. (1944) Wernicke's disease: a clinical following infarction of the left dorsomedial nucleus of the
and pathological study of 42 cases. Quarterly Journal of Studies thalamus. Neuropsychologia, 20, 597-604.
on Akohol, 5, 361-370. SQUIRE, L. R. (1981) Two forms of human amnesia: an analysis
ROBBINS,1. W. (1984) Cortical noradrenaline, attention, and of forgetting. Journal of Neuroscience, 1, 635—640.
arousal. Psychological Medicine, 14, 13—21. —¿ (1982) Comparisons between forms of amnesia: some
ROEDIGGER, H. L. & BLAXTON, T. A. (1987) Retrieval modes deficits are unique to Korsakoff's syndrome. Journal of
produce dissociations in memory for surface information. In ExperimentalPsychology:Learning, Memory. Cognition,8,
Memoryand CognitiveProcesses:
TheEbbinghausCentennial 560—572.
Conference (eds D. Gorfein & R. R. Hoffman), pp. 349—379. —¿ (1987) Memory and Brain. Oxford: Oxford University
Hillsdale, NJ: Lawrence Erlbaum. Press.
ROSSOR, M. N., IVERSEN, L. L.. REYNOLDS, 0. P., et a! (1984) —¿, Aa@AL, D. 0., ZOLA-MORGAN, S., et al(1989a) Description

Neurochemical characteristics of early and late onset types of of brain injury in the amnesic patient N.A. based on magnetic
Alzheimer's disease. British Medical Journal, 288, 961—964. resonance imaging. Experimental Neurology, 105, 23—35.
ROTH, M., TYM, E., M0UNTJ0Y, C. Q., et a! (1986) CAMDEX: a —¿, HAIST, F. & SHIMAMURA, A. P. (1989b) The neurology of

standardised instrument for the diagnosis of mental disorder in memory:quantitative assessment ofretrograde amnesiaintwo
the elderly with special reference to the early detection of types of amnesic patients. Journal of Neuroscience, 9,828-839.
dementia. British Journal of Psychiatry, 146, 698—709. —¿, AMARAL, D. 0. & PRESS, 0. A. (1990) Magnetic resonance

SAHGAL, A. (1984) A critique of the vasopressin-memory hypothesis. imaging of the hippocampal formation and mammillary nuclei
Psychopharmacology, 83, 215-228. distinguish
medicaltemporallobeand diencephalic
amnesia.
SALAZAR, A. M., GRAFMAN, J., SCHLESSELMAN, S., ci a! (1986) Journal
ofNeuroscience,
10,3106—3117.
Penetrating war injuries of the basal forebrain: neurology and —¿ & BUl-rERS, N. (1992) Neuropsychology of Memory (2nd

cognition. Neurology, 36, 459-465. edn).New York:Guilford.

SCHACIER, D. L. (1987) Implicit memory: history and current Swss, D. T., ALEXANDER, M. P., LIBERMAN, A., et a! (1978) An
status. Journal of Experimental Psychology: Learning, Memory extraordinary form of confabulation. Neurology, 28, 1166-1172.
and Cognition,13, 501—5 18. T@w@xsEN, C. M. E., BELL,H. & BOlStER,T. (1993) Thiamin and
—¿, HARBLUK, J. L. & MCLACHLAN, D. R. (1984) Retrieval thiamin phosphate ester deficiency assessed by high performance
without recollection: anexperimental analysis ofsourceamnesia. liquid chromatography in four clinical cases of Wernicke
Journalof VerbalLearningand VerbalBehavior,23, 593—611. encephalopathy. Alcoholism: Clinical and Experimental Research,
—¿ & GRAF, P. (1986) Preserved learning in amnesic patients: 17,712—716.
perspectives from research on direct priming. Journal of Clinical TALLAND, G. A. (1965) Deranged Memory: A Psychogenic Study
and ExperimentalNeuropsychology, 8, 727—743. of theAmnesicSyndrome.New York:AcademicPress.
SCHENKER, S., HENDERSON, G. I., HOYUMPA, A. M., ci a! (1980) T0RvIK, A., L1NDBOE,C. F. & ROGDE,S. (1982) Brain lesions in
Hepatic and Wernicke's encephalopathies: current concepts of alcoholics. Journal of Neurological Science, 56, 233-248.
pathogenesis. American Journal of Clinical Nutrition, 33, TULVING, E. (1972) Episodic and semantic memory. In Organisation
2719—2726. of Memory (eds E. Tulving & W. Donaldson). London:
SERDARU, M., HAUSSER-HAUW. C., LAPLANE, D., ci a! (1988) The Academic Press.
clinical spectrum of alcoholic pellagra encephalopathy. Brain, —¿ & THOMSON, D. M. (1973) Encoding specificity and
111, 829—842. retrieval processes in episodic memory. Psychological Review,
SHEPHERD, 0. M. (1988) A basic circuit of cortical organization. 80, 353—373.
In Perspectives in Memory Research (ed. M. S. Gazzaniga). —¿ & SCHACrER, D. L. (1990) Priming and human memory
Cambridge, MA: Massachusetts Institute of Technology Press. systems. Science, 247, 301—306.
SHIMAMURA,A. P. (1986) Priming effects in amnesia: evidence VALENSTEIN, E., BOWERS, D., VERFAELLIE, M., et a! (1987)
for a dissociable memory function. Quarterly Journal of Retrosplenial amnesia. Brain, 110, 1631—1646.
Experimental Psychology, 38A, 619—644. VICrOR, M. & ADAMS, R. D. (1953) The effect of alcohol on the
—¿ & SQUIRE, L. R. (1984) Paired-associate learning and nervous
system.InMetabolicandToxicDiseasesoftheNervous
priming effects in amnesia: a neuropsychological study. Journal System. Associationfor Researchin Nervous and Mental
of Experimental Psychology, 133, 556—570. Disease,
Vol.32.Baltimore: Williamsand Wilkins.
—¿ & —¿ (1986) Korsakoff's syndrome: a study of the —¿, —¿ & COLLINS, G. H. (1971) The Wernicke-Korsakoff

relation between anterograde amnesia and remote memory Syndrome.F. A. Davis.

impairment. Behavioral Neuroscience, 100, 165—170. —¿ & —¿ (1985) The alcoholic dementias. In Handbook of
—¿ & —¿ (1987) A neuropsychological study of fact memory Clinical Neurology. Vol 2 (46) (ed. J. A. M. Fredericks),
and source amnesia. Journal of Experimental Psychology: pp. 335—351.Amsterdam: Elsevier.
Learning.Memory, Cognition,13, 464-473. VONCRAMoN,D. Y., HEBEL,N. & SanJRI, U. (1985) A contribution
—¿, JERNIGAN, T. L. & SQUIRE, L. R. (1988) Korsakoff's to the anatomical basis of thalamic amnesia. Brain, 108,
syndrome: radiological (CT) findings and neuropsychological 997—1008.
correlates. Journal of Neuroscience, 8, 4400—4410. —¿ & SCHURI, U. (1992) The septo-hippocampal pathways and
—¿, JANOWSKY, J. S. & SQUIRE, L. R. (1990) Memory for the their relevance to human memory: a case report. Cortex, 28,
temporal order of events in patients with frontal lobe lesions and 411—422.
amnesicpatients.Neuropsychologia, 28, 803-813. WALLIS,W. E., WILLOUGHBY, E. & BAKER,P. (1978) Coma in the
SHOQEIRAT, M. A., MAYES, A., MACDONALD, C., ci a! (1990) Wernicke-Korsakoff syndrome. Lancet, 2, 400—401.
 THE KORSAKOFF SYNDROME 173

WARRINGTON, E. K. (1982) The double deviation of short- and deficiency: a comparative analysis. Akoho! andAkoholism, 20,
long-term memory deficits. In Human Memory andAmnesia (ed. 201—221.
L. S. Cermak), pp. 61—76.Hillsdale, NJ: Lawrence Eribaum. WooD, P., MURRAY,A., SINHA, B., et a! (1983) Wernicke's
—¿ & WEISKRANTZ, L. (1970) Amnesic syndrome: consolidation encephalopathy induced by hyperemesis gravidarum case
or retrieval? Nature, 228, 628—630. reports.BritishJournalof Obstetrics
and Gynaeco!ogy, 90,
—¿ & —¿ (1982) Amnesia —¿
a disconnection syndrome? 583—586.
Neuropsychologia, 20, 233-248. ZANGWILL, 0. L. (1946) Some qualitative observations on verbal
WECHSLER,D. (1945) A standard memory scale for clinical use. memory in cases of cerebral lesion. British Journal of
Journalof Psychology,19, 87-95. Psychology, 37, 8-19.
WEINGARTNER, H., RUDORFER, M. V., BUCHSBAUM, M. S., et a! —¿ (1953) Disorientation for age. Journal of Mental Science,

(I983a) Effectsofserotonin on memory impairments

produced 99, 698—701.
by ethanol. Science, 221, 472—473. ZOLA-MORGAN, S., SQUIRE, L. R. & AMARAL, D. 0. (1986) Human
—¿, GRAFMAN, J., BOUTELLE, W., ci a! (l983b) Forms of amnesia and the medial temporal region: enduring memory
memory failure. Science, 221, 380—382. impairment following a bilateral lesion limited to field CAl of
WEISKRANTZ, L. (1985) On issues and theories of the human the hippocampus. Journal of Neuroscience, 6, 2950-2967.
amnesic syndrome. In Memory Systems of the Brain (eds N. M. —¿ & —¿ (I989a) Lesions of the hippocampal formation
Weinberger, J. L. McGaugh & G. Lynch), pp. 380—415. but not lesions of the fornix or the mammillary nuclei produce
—¿ & WARRINGTON, E. K. (1979) Conditioning in amnesia long-lasting memory impairments in monkeys. Journal of
patients. Neuropsychologia, 17, 187-194. Neuroscience,9, 898—913.
WINOCUR, G., OXBURY, S., ROBERTS, R., ci a! (1984) Amnesia in a —¿, —¿ & —¿ (l989b) Lesions of the amygdala that

patient with bilateral lesions to the thalamus. Neuropsychologia, spare adjacentcorticalregionsdo not impair memory or
22, 123—143. exacerbate the impairment following lesions of the hippocampal
WI1-r, E. D. (1985) Neuroanatomical consequences of thiamine formation. Journal of Neuroscience, 9, 1922-1936.

Michael D. Kopelman, PhD, Division of Psychiatry and Psychology, UMDS, Guy's and St Thomas's
Hospitals, St Thomas's Campus, London SE1 7EH

(First received 20 August 1993, final revision 25 March 1994, accepted 26 April 1994)