Vous êtes sur la page 1sur 11

European Journal of Neuroscience, Vol. 18, pp.

16601670, 2003

Federation of European Neuroscience Societies

Associative and recognition memory for novel objects in


dementia: implications for diagnosis
Andy C. H. Lee,1 Shibley Rahman,2 John R. Hodges,1,3 Barbara J. Sahakian2 and Kim S. Graham1
1

MRC Cognition and Brain Sciences Unit, 15 Chaucer Road, Cambridge CB2 2EF,UK
Department of Psychiatry, and
3
University Neurology Unit, University of Cambridge, Addenbrooke's Hospital, Hills Road, Cambridge, UK
2

Keywords: Alzheimer's disease, frontotemporal dementia, human, semantic dementia, perirhinal cortex, perception

Abstract
It has been demonstrated that patients with dementia of the Alzheimer's type show particular difculties with a task that measures
memory for object locations [R. Swainson et al. (2001) Dement. Geriatr. Cogn. Disord. 12, 26580]. The present study followed on from
this report by asking whether the decits seen in dementia of the Alzheimer's type were specic to this condition, or whether they would
also be seen in another common neurodegenerative syndrome, frontotemporal dementia. To investigate this important issue, we
examined memory for objectlocation pairs and visual recognition memory for novel patterns using two tests, the Paired Associates
Learning and Matching to Sample tasks, from the Cambridge Neuropsychological Testing Automated Battery. The performance of a
subset of the patients with dementia of the Alzheimer's type described by Swainson et al., selected on the basis of age and education,
was compared with matched groups of frontal variant frontotemporal dementia, semantic dementia and control subjects. In contrast to
the patients with dementia of the Alzheimer's type, who showed signicant impairment on both memory tests, the two frontotemporal
dementia groups did not perform signicantly poorer compared with control subjects on nearly all memory measures, other than
`memory score' from the paired associates learning task. These ndings conrm that tests of episodic memory, especially for the
location of objects in space, may be useful in the early diagnosis and differentiation of dementia of the Alzheimer's type.

Introduction
A recent study revealed that a visuo-spatial paired associates learning
task (PAL), in which subjects have to remember the location of objects,
accurately distinguished patients with dementia of the Alzheimer's
type (DAT) from depressed and control subjects (Swainson et al.,
2001). Furthermore, the task successfully identied a subgroup of
patients with questionable dementia who showed a similar prole to
DAT patients, a nding that suggests that these cases were already in
the early stages of Alzheimer's disease (see also Fowler et al., 1997).
Although this experiment revealed an exciting new paradigm for the
early detection of DAT, it is currently unclear whether a similar
impairment would be evident in patients with other forms of dementia,
such as frontotemporal dementia (FTD).
FTD is the second most prevalent cause of dementia in younger
populations (Hodges & Miller, 2001a,b; Ratnavalli et al., 2002), and
there are two clinically distinct forms: temporal variant FTD (or
semantic dementia, SD) and frontal variant FTD (fvFTD). In SD,
patients show a progressive, cross-modal, loss of semantic knowledge
in the context of focal atrophy to anterior and inferior temporal
lobe regions (Hodges et al., 1992). However, contrary to theories
that episodic memory is dependent upon intact semantic knowledge
(Tulving, 1995, 2001), SD patients typically demonstrate good pictorial associative and recognition memory, even for stimuli for which the

Correspondence: Dr A. Lee, as above.


E-mail: andy.lee@mrc-cbu.cam.ac.uk
Received 2 April 2003, revised 27 June 2003, accepted 1 July 2003
doi:10.1046/j.1460-9568.2003.02883.x

patients have lost semantic knowledge (Graham et al., 1997, 2000;


Simons et al., 2001, 2002).
In fvFTD, frontal pathology predominantly affects the orbitomedial
cortex causing changes in behaviour and personality, such as apathy,
loss of empathy, impulsivity, disinhibition, ritualistic behaviours and
increased risk-taking (Miller et al., 1995; Rahman et al., 1999; Bozeat
et al., 2000). With respect to episodic memory, Simons et al. (2002)
found that whereas a group of fvFTD patients possessed intact
recognition memory, they were profoundly impaired at recollecting
the time at which the studied stimuli had been presented (temporal
source monitoring). Spatial source monitoring (recollecting the spatial
location of visual stimuli) such as that assessed by the PAL has not
been examined in fvFTD, although, to our knowledge, frontal lobe
damage has not been previously associated with an impairment in
associative spatial memory.
In terms of contrasting DAT with FTD, Perry & Hodges (2000; also
Hodges et al., 1999) found that on episodic memory tests, such as
recall of a story and recognition memory for unfamiliar faces, the DAT
group showed a severe decit, with weaker impairments evident in
fvFTD and an even milder decit in SD. Studies in FTD suggest
therefore that although patients rarely show the profound amnesic
decits seen early in DAT, there can be signicant impairment
particularly in fvFTD patients in terms of encoding and/or retrieving
aspects of contextual information.
The rst aim of the present study was to determine whether the
decits seen in DAT patients on the PAL task from the Cambridge
Neuropsychological Test Automated Battery (CANTAB) could be
generalized to cases with FTD. Although the PAL task was the most

Memory in frontotemporal dementia


sensitive measure reported by Swainson et al. (2001), an additional
task from the CANTAB battery matching to sample (MTS) after a
delay for visual stimuli was also impaired in both DAT and a
subgroup of patients with questionable dementia, but not in depressed
patients. Furthermore, performance on the MTS and PAL tasks were
the only mnemonic measures utilized by Swainson et al. (2001) that
signicantly correlated with global cognitive function, as assessed by
the Mini Mental State Examination (MMSE). Given that both FTD
groups demonstrate relatively preserved episodic recognition memory,
the second aim of this study was to assess whether the CANTAB MTS
task may also be a useful tool in the early diagnosis and differentiation
of neurodegenerative diseases.
The MTS task is also interesting because it assesses recognition
memory using novel stimuli (items, in this case coloured rectangles, that
are not familiar from everyday life). Recent studies have used novel
visual items to demonstrate that lesions to the perirhinal cortex, which is
situated in the lateral bank of the collateral sulcus, can impair a monkey's
ability to process conjunctions of visual features (Buckley et al., 2001;
Bussey et al., 2002). There is accumulating evidence to suggest that
there is signicant perirhinal cortex atrophy in SD (Chan et al., 2001;
Galton et al., 2001; Davies et al., 2002) and, thus, SD patients may be
impaired on the MTS task, which requires the discrimination of novel
complex stimuli on the basis of both pattern and colour.
Considering the existing literature, it was difcult to make clear
predictions about the FTD patients. The SD patients were likely to
show good performance on both tasks, although the use of complex
novel stimuli could affect MTS task performance. It was uncertain
whether the fvFTD group would be impaired on the PAL: although
Simons et al. (2002) observed difculties with contextual recall in
fvFTD, there is little evidence for frontal lobe involvement in spatial
associative memory. However, the fvFTD patients were likely to
demonstrate good recognition memory on the MTS task.

1661

Table 1 provides a summary of the mean age and years of education of


the subject groups. This study received ethical approval from the
Cambridge Health Authority Local Research Ethics Committee and
informed consent from all participants.
All the patients tested in this study presented through the Memory
Disorders Clinic at Addenbrooke's Hospital, Cambridge, UK, and
have been longitudinally assessed on an extensive neuropsychological
battery. The diagnosis of probable DAT was based on inclusion and
exclusion criteria (McKahnn et al., 1984) developed by the National
Institute of Neurological and Communicative Disorders and Stroke
(NINCDS) and the Alzheimer's Disease and Related Disorders Association (ADRDA). All the FTD patients recruited for the study fullled
the LundManchester consensus criteria for frontotemporal lobar
degeneration (Neary et al., 1998). Patients diagnosed with the SD
subtype were characterized by a decit in semantic knowledge,
combined with relative preservation of syntax, phonology, perceptual
and visuospatial abilities, non-verbal problem solving and aspects of
episodic memory (Hodges et al., 1992). By contrast, patients diagnosed with fvFTD presented with a history of progressive change in
personality and behaviour with a number of possible dening features
including disinhibition, apathy, restlessness, poor planning, impulsiveness, poor self-care, reduced empathy, social withdrawal, reduced
verbal output and verbal stereotypes (Gregory & Hodges, 1996).
The semantic and episodic memory abilities of the patients were
quantied with a series of tests assessing word production, verbal and
non-verbal comprehension and recognition memory. These included
category uency, picture naming, the Pyramid and Palm Trees Test
(Howard & Patterson, 1992), and the Recognition Memory Test
(Warrington, 1984). The control subjects were not assessed on these
tasks and thus, for comparison, normative data for these tests have been
taken from a previous study (Hodges & Patterson, 1995). Table 1
provides the mean scores for each subject group on these tests.
Procedure

Methods
Subjects
Four different subject groups participated in this study, including 11
SD patients, 12 fvFTD patients, 10 probable DAT patients (original
data from a subset of patients assessed by Swainson et al., 2001) and 18
age-matched healthy control subjects (18 were assessed on MTS and
16 on PAL). The two FTD groups were selected to match the DAT
patients on the basis of MMSE scores (DAT range 1824), age and
years of education. None of the subjects had any previous experience
of the experimental tasks. Owing to behavioural difculties during
testing, the data from one fvFTD patient were not included in the
statistical analyses of the data (reducing the number of cases to 11).

The patients were either tested at their own home or at the Early
Dementia Clinic at Addenbrooke's Hospital, Cambridge, UK. The two
tests that were used were taken from the CANTAB (Cambridge
Cognition plc, Cambridge, UK). The CANTAB is a series of computerized tests that were developed from classical neuropsychological
tasks designed to assess memory and cognitive function. In the present
study, a portable Databrick 486 computer (Datalux, Winchester, VA,
USA) tted with a touch-sensitive colour monitor was used to run the
CANTAB. Subjects were seated in front of the computer so that they
could comfortably touch the monitor during the tasks and were given
an opportunity to familiarize themselves with the screen prior to the
start of testing.

Table 1. Mean (with standard error in parentheses) background neuropsychological data for the four subject groups

Group

Age
(years)

Education
(years)

MMSE
(30)

Category
fluency (total
score from
8 categories)

SD
fvFTD
DAT
Control 1y
Control 2
Control 3

61.64  2.11
62.50  2.75
62.50  2.75
60.25  1.45
60.39  1.29
69.67  1.60

12.10  1.02
11.57  0.78
11.50  0.60
12.31  2.60
12.28  2.47
10.78  0.46

23.18  1.68
25.70  2.33
20.60  0.75
N/A
N/A
29.17  0.21

28.55  7.50
81.00  9.15
54.10  2.62
N/A
N/A
113.71  3.96

Naming
(% correct
of 48 or 64)
0.41  0.20
0.96  0.01
0.90  0.03
N/A
N/A
0.930.02

Pyramids & Palm


Trees (% correct)

Recognition
memory (% correct)

Words

Pictures

Words

Faces

0.73  0.05
0.96  0.02
N/A
N/A
N/A
N/A

0.78  0.04
0.97  0.009
N/A
N/A
N/A
0.98  0.02

0.76  0.05
0.83  0.05
0.70  0.03
N/A
N/A
0.94  0.01

0.77  0.05
0.76  0.03
0.77  0.03
N/A
N/A
0.88  0.01

Control subjects tested on PAL. Control subjects tested on MTS. Control subjects for background neuropsychological tasks (from Hodges & Patterson, 1995).
Significantly impaired at a corrected level compared with control group (P 0.001). Significantly impaired at a corrected level compared with fvFTD and
control groups (P 0.0001). Significantly impaired at a corrected level compared with fvFTD, DAT and control groups (P 0.0001).


2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

1662 A. C. H. Lee et al.

Fig. 2. Screen-shots of the MTS simultaneous condition during (a) stimulus presentation, (b) the response stage and (c) the response feedback stage.

Fig. 1. Selected screen-shots from the PAL task during the encoding (a and b)
and retrieval (c and d) stages of the task.
2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

Memory in frontotemporal dementia


Paired associates learning (PAL)
This test assessed the ability to learn the spatial positions of up to eight
abstract visual patterns (see Fig. 1). The subject was initially presented with six white boxes in a circular conguration and these then
opened one at a time for 3 s each in a randomized order. On the rst
trial, one of these boxes contained a pattern and the subject was
required to remember the pattern and its spatial location. This
stimulus was then presented in the centre of the screen and the subject
was required to touch the box that contained this pattern. If the subject
made the correct response then the phrase `All correct' appeared in the
middle of the screen and the task progressed to the next problem. The
task gradually progressed in difculty, starting with two 1-pattern
problems (Stages 1 and 2), followed by two 2-pattern problems
(Stages 3 and 4), two 3-pattern problems (Stages 5 and 6), one 6pattern problem (Stage 7) and, nally, one 8-pattern problem (Stage
8), for which eight white boxes were presented. If an incorrect
response was given, in which at least one pattern was misplaced,
all the boxes were re-opened in a randomized order to remind the
subject of the locations of the patterns. The subject was then given
another attempt to locate successively all the patterns and in total was
allowed nine further attempts following the rst attempt. If these were
exceeded without success, then the task automatically terminated and
the subject was considered to have failed the task at that particular
stage. Performance was assessed according to three measures, two of
which (2 and 3) were designed to take into account the fact that the
subjects were allowed up to ten attempts to produce a correct answer:
(1) number of stages passed; (2) total errors across all response
attempts for Stage 7 only (6-object problem) and (3) memory score,
which was the sum of the number correct across all stages after the
rst presentation of the stimuli only (maximum score 26). The rst
two measures were previously adopted by Swainson et al. (2001),
with the second being the most accurate at distinguishing DAT from
depressed/control subjects and identifying a subgroup of patients with
questionable dementia who performed similarly to the DAT group.
Simultaneous and delayed matching to sample (MTS)
This test assessed the ability to perceive and learn abstract visual stimuli.
The subject was rst presented with a single rectangular pattern
comprising four quadrants of different colour and shape for 4.5 s and
was instructed to study this stimulus as a recognition trial would follow
immediately (see Fig. 2). One of four possible conditions then ensued: a
simultaneous condition or one of three delay conditions. In the simultaneous condition, the target stimulus remained on the screen while four
choice stimuli appeared directly below and the subject was required to
touch the stimulus that was exactly identical to the target stimulus. Only
one of the four choice stimuli matched the target stimulus exactly. One
of the other choice stimuli was a completely novel distractor in that it
differed in terms of both colour and shape. The other two choice stimuli
were partial distractors, with one possessing the same colour as the
target stimulus but a different shape and the other possessing the same
shape as the target stimulus but a different colour. In order to discourage
the subject from completing the task by simply learning the colour and
shape of a single quadrant, each of the choice stimuli possessed one
quadrant that was taken from the target stimulus. Feedback was
provided visually, in the form of a green tick or red cross and the word
`correct' or `incorrect', as well as aurally, via a high or low computergenerated tone. If an incorrect response was given, the subject had to
continue responding until the correct stimulus was selected. The delay
conditions were identical to the simultaneous condition except that the
target stimulus disappeared after the initial 4.5 s presentation period.
There was then a delay of 0 s, 4 s or 12 s, after which the four choice

1663

stimuli appeared and the subject was required to make a response.


Performance was assessed according to the proportion correct and the
average response time for each condition.
Statistical analyses
A series of parametric tests (analyses of variance, ANOVA) were used to
contrast the different subject groups' performance on the PAL and
MTS tasks. For planned comparisons, a statistical threshold of
P < 0.05 (uncorrected for multiple comparisons) was adopted. For
post-hoc analyses, a statistical threshold of P < 0.008 (corrected for
multiple comparisons) was used. This threshold was derived by
dividing the uncorrected threshold (P 0.05) by the number of
possible comparisons between the four subject groups (6).
The accuracy and response latency data for the simultaneous
condition and the three delay conditions of the MTS task were
analysed separately.

Results
Background neuropsychology
General characteristics
One-way ANOVAs revealed no signicant differences between all
groups assessed on the PAL and MTS tasks in terms of age
(F4,65 0.269, P > 0.9) or years of education (F4,58 0.238,
P > 0.9). There was also no signicant difference between the three
patient groups in terms of MMSE (F2,31 2.142, P > 0.1). Thus, any
signicant group effects in task performance were unlikely to be a
result of differences in age, education or the degree of disease severity
of the three patient groups.
Semantic memory function
As expected, the SD patient group exhibited impaired performance on
general tasks of semantic memory function compared with control data
taken from Hodges & Patterson (1995). One-way ANOVAs revealed a
signicant group effect for category uency (F3,55 44.43,
P < 0.001), picture naming (F3,55 30.80, P < 0.001) and the picture
version of the pyramid and palm trees test (PPT) (F2,44 41.44,
P < 0.001). Post-hoc tests showed that the SD group was signicantly
impaired in comparison with fvFTD and control groups on all these
tests (all P < 0.0001). Similarly, the SD group was found to perform
signicantly worse compared with the DAT group on picture naming
(P 0.0001), although there was only a trend towards a signicant
difference on category uency (P 0.046) at a corrected level.
Episodic memory function
On the recognition memory test (RMT) words version, there was a
signicant group effect (F3,51 12.37, P < 0.001). Whereas post-hoc
analyses revealed only a trend towards a signicant difference between
the fvFTD patients and the control group (P 0.034), both SD and
DAT patient groups were signicantly impaired compared with controls (both P < 0.001). On the RMT face version, there was also a
signicant group effect (F3,51 6.14, P 0.001), with all three patient
groups performing similarly. However, although there was a signicant
group effect, post-hoc analyses revealed that at a corrected level, there
were only trends towards a signicant difference between each of the
patient groups and the control group (SD, P 0.01; fvFTD, P 0.02;
DAT, P 0.02).
PAL
From Fig. 3, it can be seen that whereas all DAT patients failed to pass
the 6-object stage of the PAL task, all control subjects managed to pass

2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

1664 A. C. H. Lee et al.

Fig. 3. Percentage of subjects passing each stage of the PAL task. Stage 12, 1
object; Stage 34, 2 objects; Stage 56, 3 objects; Stage 7, 6 objects; Stage 8, 8
objects.

Fig. 4. (a) Mean memory score and stages completed for the four subject
groups ( standard error). (b) Mean memory score for the subject groups at each
stage of the PAL task ( standard error).

all stages. Of the FTD groups, 9% of fvFTD and 36% of SD cases


failed at the critical 6-object stage whereas 27% and 45% failed at the
most difcult 8-object stage. Figure 4a illustrates a signicant effect of
group on the number of stages passed (F3,47 40.60, P < 0.001) and
the memory score (F3,47 42.38, P < 0.001). Post-hoc analyses
revealed that on both these measures, the DAT group was signicantly
impaired compared with the SD, fvFTD and control groups
(P < 0.001). Whereas the SD and fvFTD groups showed a similar
level of performance to controls on the `number of stages completed'
(both, P > 0.1), the SD group's `memory score' was signicantly lower
than that obtained by the controls (P 0.001), and there was a trend
towards the fvFTD group's memory score being signicantly lower
than that of the controls (P 0.01). This nding implies that although
overall the FTD patients were able to solve the PAL task and complete
virtually all stages, they made more errors than the controls after the
rst presentation of the stimuli items. As seen in Fig. 4b, it is at the later
stages of the task that the FTD patients were making more errors than
the controls.
It is important to highlight that the memory score (sum of the
number correct across all stages after the rst presentation of the
stimuli only) is conated with the number of stages passed. Thus, the
higher mean memory score demonstrated by the control group is likely,
in part, to reect the nding that all the control subjects reached the 8object stage of the PAL, whereas a number of the patients, in particular
in the DAT group, failed to reach the later stages of the task. To account
for this, an additional one-way ANOVA was conducted to examine the
memory score up to and including the rst 2-object stage (stage 3 on
Fig. 3), the highest stage that was reached, but not necessarily passed,
by all subjects. This revealed a signicant effect of group (F3,47
15.80, P < 0.0001) and post-hoc analyses showed that although the
DAT group performed signicantly poorer than the control group and
the two FTD groups (all P 0.001) at this early stage of the PAL, there
was no signicant difference between the FTD groups and the control
group (both P > 0.8; see Fig. 4b).

Fig. 5. Mean number of errors on the 6-object stage for the individuals of each
subject group (`F' indicates that subject failed to reach this stage).

2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

Memory in frontotemporal dementia


To analyse whether the total number of errors they made at the 6object problem alone differed between the groups, only those patients
who reached that stage were included (SD, 11; fvFTD, 11; DAT, 3;
control, 16). Figure 5 shows the scores of individual subjects on the
PAL at the 6-pattern stage and, from this, it can be seen that there was a
signicant effect of group (F3,40 16.17, P < 0.001). Similar to the
other PAL measures, post-hoc analyses revealed that the DAT group
performed signicantly worse on this component than the three other
groups (P < 0.001). None of the other contrasts reached corrected
statistical signicance.
In terms of individual patients, it was noticeable that every DAT
patient was two standard deviations below the control mean regardless
of which PAL measure was considered. By contrast, only 24 of the SD
and fvFTD patients were impaired according to this criterion and,
moreover, there was no consistency in which patients were impaired on
the different PAL measures.
MTS
Figure 6 illustrates the mean performance accuracy and corresponding
standard errors of each subject group on the different conditions of the
MTS task. Given that performance was measured in terms of proportion correct, an arcsine transformation (2  arcsine(Hx)) was applied
to the data prior to statistical analysis (see Howell, 1997).
Although, on the simultaneous condition, the ANOVA analysis
showed a signicant effect of group (F3,49 3.25, P < 0.05), posthoc analyses found no differences that were signicant at the corrected
threshold. Numerically, however, the DAT group showed poorer
performance than all three groups and there was a trend towards a
signicant difference between the DAT patients and controls
(P > 0.02). On the delay conditions, a repeated measures ANOVA found
a signicant effect of both group (F3,46 9.41, P < 0.001) and delay
(F2,92 12.63, P < 0.001) but there was no evidence of a delay-bygroup interaction (F6,92 0.72, P > 0.6). Further analyses revealed
that the group effect could be attributed to poorer scores in the DAT
group compared with controls (P < 0.001). There was, however, no
signicant difference at a corrected level between the DAT patients and
both FTD groups (SD vs. DAT, P > 0.06; fvFTD vs. DAT, P > 0.02),
and between the FTD groups and controls (both P > 0.07).

1665

Figure 7 illustrates the performance of individual subjects in each


group on the different conditions of the MTS task. Only one SD
patient, two fvFTD patients and three DAT patients performed greater
than two standard deviations below the control mean at the simultaneous condition. As seen from Fig. 7, however, one of the fvFTD
patients and the SD patient were only marginally below this criterion.
At the delay conditions, there were no SD patients below this criterion
at 0 s, whereas three were impaired at the 4 s stage and two exceeded
this criterion at 12 s. Only 12 of the fvFTD patients were impaired at
each delay condition whereas 38 of the DAT patients were beyond
two standard deviations below the control mean at each delay.
To determine whether the different subject groups could be differentiated on the basis of the type of errors they made on the MTS task,
multivariate ANOVAs were conducted on the proportion of error types
that the three patient groups and the control group committed on the
delay and simultaneous conditions. There were three types of error that
were possible: (1) colour errors, in which a stimulus of the correct
colour but incorrect shape was chosen; (2) shape errors, in which a
stimulus of the correct shape but incorrect colour was chosen and (3)
distractor errors, in which a stimulus of incorrect colour and shape was
chosen. For the simultaneous condition, it was found that all subject
groups generally committed a greater proportion of colour (SD, 75%;
fvFTD, 67%; DAT, 19%; control, 29% of simultaneous errors) and
shape (SD, 25%; fvFTD, 33%; DAT, 63%; control, 71% of simultaneous errors) errors than distractor errors (SD, 0%; fvFTD, 0%; DAT,
18%; control, 0% of simultaneous errors). A multivariate ANOVA
revealed, however, no signicant difference between the groups (all
F3,23 < 2, P > 0.14), indicating that all four subject groups did not
commit signicantly different proportions of the three different error
types on the simultaneous condition. On the three delay conditions,
colour (SD, 54%; fvFTD, 55%; DAT, 28%; control, 21% of delay
errors) and shape errors (SD, 38%; fvFTD, 38%; DAT, 51%; control,
78% of delay errors) were also more frequent than distractor errors
(SD, 7%; fvFTD, 7%; DAT, 21%; control, 0.05% of delay errors).
There was, however, a signicant difference between groups on all
three error types (all F3,46 > 8, P < 0.0001), suggesting that the error
proles of the four subject groups were different from one another on
the delay conditions. Post-hoc analyses revealed that the DAT group
committed a signicantly greater proportion of distractor errors compared with controls and SD patients (both P < 0.008), with a trend
towards a signicant difference between the DAT group and the fvFTD
group (P 0.01). By contrast, both SD and fvFTD groups made a
signicantly greater proportion of colour and shape errors compared
with controls (all P < 0.002), whereas the DAT group was not signicantly different compared with the control group on colour errors
(P > 0.8), although there was a trend towards a signicant difference
on shape errors (P 0.02).
MTS response latency

Fig. 6. The proportion correct (arcsine transformed) on the simultaneous, 0 s,


4 s and 12 s conditions of the MTS task for each subject group ( standard
error).

Figure 8 illustrates the mean response latencies and corresponding


standard errors of each subject group on the different conditions of the
MTS task. A log10 transformation was applied to the response latency
data prior to statistical analysis.
In the simultaneous condition, there was a signicant effect of group
(F3,49 4.06, P < 0.02), although as for the accuracy measure none of
the post-hoc contrasts was signicant. Figure 8 suggests that all three
patient groups were marginally slower at this task than controls. A
repeated measures ANOVA based on the delay condition results found a
signicant effect of group (F3,48 3.145, P < 0.05) and delay
(F2,92 18.19, P < 0.001) but not a signicant group interaction
(F6,94 2.04, P > 0.06). Post-hoc analyses found no differences
between the subject groups that were signicant at the corrected

2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

1666 A. C. H. Lee et al.

Fig. 7. The proportion correct (arcsin transformed) at (a) the simultaneous condition, (b) the 0 s condition, (c) the 4 s condition and (d) the 12 s condition for the
individuals of each subject group. The dotted line represents the control mean and the dashed line indicates two standard deviations below this.

threshold (all P > 0.04). Figure 8 shows, however, that all three patient
groups were taking longer to respond than the neurological healthy
subjects.
Correlation with semantic function
Given recent suggestions that the perirhinal cortex may subserve
semantic memory processes in humans (Murray & Bussey, 1999),
the relationship between semantic function and performance on the
PAL and MTS tasks was examined. Multiple correlation analyses
between the SD patients' performance on tests of semantic function
and their level of impairment on the different measures of performance
on the PAL revealed only one signicant correlation between the
category uency score for living items and the PAL total memory score
(P 0.004, corrected for multiple comparisons). There were no signicant correlations between the semantic test scores and the SD
patients' performance on the different conditions of the MTS task.
Results summary
As predicted, the DAT group was signicantly impaired compared with
the control group on all three performance measures of the PAL task
(stages passed, memory score and errors at the 6-object stage). By
contrast, the SD groups was only signicantly impaired on the memory
score measure, whereas the fvFTD group was not signicantly impaired
at a corrected level on any measure, although there was a trend towards
a signicant difference on the memory score (see Table 2).

Fig. 8. The average latency of correct trials (log10 transformed) on the


simultaneous, 0 s, 4 s and 12 s conditions of the MTS for each subject group
( standard error).

2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

Memory in frontotemporal dementia


Table 2. Summary of overall group performances on PAL and MTS tasks
Patient group vs control group
Task

SD

fvFTD

DAT

PAL
Stages completed
Memory score
Errors at 6-object stage

Unimpaired
Impaired
Unimpaired

Unimpaired
Unimpaired
Unimpaired

Impaired
Impaired
Impaired

MTS simultaneous
Accuracy
Latency

Unimpaired
Unimpaired

Unimpaired
Unimpaired

Unimpaired
Unimpaired

MTS delay
Accuracy
Latency

Unimpaired
Unimpaired

Unimpaired
Unimpaired

Impaired
Unimpaired

Significantly impaired (at a corrected level P 0.008) compared with control


group.

On the simultaneous and delay conditions of the MTS task, all three
patient groups were numerically worse in terms of accuracy and
response latency in comparison with the control group. Only the
DAT group, however, was signicantly impaired at a corrected statistical threshold (P < 0.008) in terms of accuracy on the delay
conditions (see Table 2).

Discussion
Our experimental study was aimed at addressing issues about episodic
memory, in particular cross-modal association learning (PAL) and
recognition memory (MTS), in patients with neurodegenerative disease.
Paired associates learning in FTD
The rst question asked whether the decits previously documented in
DAT by Swainson et al. (2001) using the PAL task would generalize to
patients with FTD. Consistent with previous studies (Fowler et al.,
1997; Swainson et al., 2001), the PAL was found to be particularly
sensitive to DAT, with patients showing decits on all three performance measures (number of stages passed, number of errors at the 6pattern stage and memory score) compared with all other tested groups
(controls and FTD patients). By contrast, both groups of FTD patients
showed relatively good performance on the PAL task, although notably
the SD group was signicantly impaired on the `memory score'
measure (with a trend towards a signicant decit for the fvFTD
group) when the number of items correct at all stages after the rst
presentation of the stimuli was summed. Although this decit was
much milder than that seen in DAT (there was no signicant difference
between the two FTD groups and the control group on the memory
score up to the rst 2-object stage) this nding is important.
Although the majority of FTD patients were capable of performing
the PAL task at least in terms of accurately learning the association
between object and location the number of patients passing the
difcult 6- and 8-item stages was less than controls (see Fig. 3),
thereby reducing the total memory score obtained by these groups
compared with controls. This nding suggests poorer memory in some
patients with FTD, compared with age- and education-matched
healthy subjects, especially when larger sets of items are to be
remembered, and provides an interesting issue for further study.
At a clinical level, the most important nding was that at the critical
6-object stage, the majority of FTD patients performed well on the
PAL tasks, a result suggesting that PAL is indeed useful in the early
diagnosis of DAT. As noted by Swainson et al. (2001), not only does it

1667

accurately identify patients with questionable dementia who will go on


to develop full-blown DAT (Swainson et al., 2001), but it also, in many
cases, differentiates patients with fvFTD and SD from those with
established DAT. It is important to note, however, that the test was not
100% accurate: there was no condition that was failed by all DAT
patients, yet passed by all FTD cases (see Fig. 3). For example, in the 3item condition, all patients with FTD passed the stage, as did approximately 35% of DAT patients. In the next condition, a 6-objectlocation
trial, whereas all DAT patients failed, so did approximately 1040% of
fvFTD and SD cases, respectively. Further investigation into alternative performance measures (such as total errors adjusted for number
of stages passed and response attempts made) and combinations of
these scores also failed to identify a measure or combination of
measures that would be able to distinguish between the different
patient groups more effectively. Consequently the PAL task, as it is
presently designed, does not appear to be sensitive enough to differentiate these two diseases sufciently. Future designs should perhaps
include additional problems to circumvent this particular methodological problem. For example, a 4- and/or 5-item condition may bridge
the gap between the current 3- and 6-item conditions, and extra trials at
various stages of the task may provide more data at critical levels of
difculty.
The fact that the majority of SD cases showed good cross-modal
associative learning is consistent with the results from Simons et al.
(2002), in which many patients showed normal performance on an
associative task for door-sofa stimuli and a test requiring recollection
of the set ownership of studied pictures. By contrast, whereas Simons
et al. (2002) observed that ve patients with fvFTD had poor temporal
source memory but good item detection, a result they attributed to the
patients' prefrontal damage (see also ndings in normal subjects,
Schacter et al., 1984; Craik et al., 1990), our fvFTD group performed
well on the PAL, despite this task also requiring recollective memory
processes (i.e. the subject is required to recall the pairing of the object
and its presented location). The discrepancy in these ndings may be
explained by considering a study by Dimitrov et al. (1999). A group of
fvFTD patients were instructed to learn pairs of words and pictures and
then performed a cued-recall memory task. Although the majority of
patients were unimpaired, a small subgroup of fvFTD patients, who
had additional temporal lobe damage, showed a moderate memory
decit. This result, considered alongside our PAL ndings, suggests
that whereas source monitoring tasks similar to that used by Simons
et al. may be particularly dependent upon the prefrontal regions,
associative memory tasks may require more input from regions in
the temporal lobes. In particular, spatial tasks such as PAL may be
predominantly reliant upon medial temporal lobe (MTL), rather than
frontal lobe regions, as demonstrated by the sensitivity of the PAL to
DAT but not fvFTD. In support of this, to our knowledge, frontal
damage has only very rarely been reported in the literature to cause
associative spatial memory decits. Further support for the critical role
of MTL areas in associative spatial tasks comes from neuroimaging
studies that have demonstrated right parahippocampal/hippocampal
region activity during tasks of objectplace memory tests (Owen et al.,
1996; Maguire et al., 1996; Johnsrude et al., 1999), and from studies in
non-human primates that have found impairments in spatial scene
learning following both perirhinal and hippocampal lesions (Murray
et al., 1998).
Recognition memory for novel stimuli in DAT and FTD
The second question addressed by our study was whether an MTS task
based on novel stimuli would also serve as a sensitive tool for diagnosis
and differentiation between DAT and FTD. We found that the DAT
group, as reported by Swainson et al. (2001), showed the most

2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

1668 A. C. H. Lee et al.


profound decits on MTS whereas the SD and fvFTD groups were not
signicantly impaired on any conditions of the MTS task (see Fig. 7).
The scores of the FTD groups occupied an intermediate position and
were not signicantly different to either controls or DAT. So, although
this task may be sensitive to DAT, it may not be particularly useful in
discriminating between DAT and FTD. It is interesting to note,
however, that the FTD and DAT groups showed different error proles
on the delay conditions of the MTS task, with the latter committing a
greater proportion of distractor errors compared with both FTD groups,
who were more likely to make colour or shape errors. This suggests
that whereas the DAT group failed to remember any colour or shape
information on many of their error trials, the FTD groups were more
likely to retain at least one aspect of information correctly. Future
investigation into this measure of performance may determine its
suitability in the discrimination of FTD and DAT.
Investigating recognition memory for novel stimuli in SD is also of
important theoretical interest because studies in non-human primates
suggest that the perirhinal cortex (BA 35/36), which is situated in the
banks of the collateral sulcus in man, is critical for object recognition
memory (Aggleton & Brown, 1999; Murray & Bussey, 1999; Buckley
& Gaffan, 2000; Murray, 2000; Buckley et al., 2001). Volumetric
studies that have been carried out in SD have shown that patients with
this disease show signicant, albeit asymmetrical, atrophy to nonhippocampal MTL regions including the perirhinal cortex (Galton
et al., 2001; Chan et al., 2001). More recently, Davies et al. (2002)
demonstrated that a group of SD patients, a subset of which were
included in this study, exhibited signicant disproportionate atrophy to
the perirhinal and temporopolar cortices compared with a matched
group of DAT patients.
Evidence for a critical role of the perirhinal cortex in object
recognition memory comes largely from animal lesion studies that
have demonstrated that rhinal but not hippocampal lesions produce
severe stimulus recognition decits (Zola-Morgan et al., 1989; Meunier et al., 1993; Suzuki et al., 1993; Bachevalier, 1994; Bachevalier &
Mishkin, 1994; Eacott et al., 1994; Alvarez et al., 1995; Murray &
Mishkin, 1998). Recent studies, however, suggest that these decits
may be attributed to problems with higher order visual perception.
Eacott et al. (1994) reported that rhinal cortex lesioned monkeys were
impaired in both simultaneous and 0 s delay conditions of an MTS task,
despite neither imposing a signicant memory demand. It has also
been documented that perirhinal lesioned monkeys were impaired on
visual concurrent discrimination tasks when the stimulus set sizes were
large (Buckley & Gaffan, 1997a), contained a large number of distracting stimuli (Buckley & Gaffan, 1997a) or used different orientations of the stimuli across trials (Buckley & Gaffan, 1997b).
From these data it has been proposed that the perirhinal cortex may
form the nal stage of the ventral visual processing stream (Ungerleider & Mishkin, 1982; Murray & Bussey, 1999). Thus, although caudal
inferotemporal cortical regions, including higher visual areas V4 and
TE/TEO, are involved in the representation of simple stimulus features, rostral regions, including the perirhinal cortex, store conjunctions of stimulus features. In support of this, Buckley et al. (2001)
found that perirhinal cortex ablations impair a monkey's ability to
make visual discriminations on the basis of object perception but not
their ability to discriminate on the basis of simple features such as size,
shape and colour.
Despite a number of non-human primate studies providing evidence
for a role of the perirhinal cortex in visual processing, there has been
little support from human studies. Firstly, SD patients typically show
normal pictorial recognition memory early in the disease (Graham
et al., 1997, 2000; Simons et al., 2001, 2002). Moreover, even in the
small number of patients who show decient recognition memory, the

decit is typically milder than that seen in DAT and performance on


standard tests of perceptual processing is usually good (Simons et al.,
2002). Similarly, Buffalo et al. (1998) and Holdstock et al. (2000)
reported normal performance on MTS in patients with focal lesions
affecting the perirhinal cortex in both a simultaneous and 0-s delay
condition.
One suggested explanation for the discrepancy between the two
literatures has been that patients with SD were not tested on novel
stimuli, which are more typically adopted as test items in the nonhuman primate investigations. The present study provides tentative
evidence in support of this in that the SD group performed numerically
poorer than the control subjects on the MTS task, although this did not
reach statistical signicance. It is possible that a lack of statistical
power underlies this non-signicant result, and that a larger sample of
SD patients would push this difference towards signicance. Alternatively, there are a number of theoretical reasons to explain the
absence of a signicant impairment in this study. First, although a
subset of the SD patients from this study have been shown, via
volumetric analysis, to have disproportionate atrophy to the perirhinal
cortex compared with DAT patients (Davies et al., 2002), this damage
is incomplete and a proportion of the perirhinal cortex remains intact.
Furthermore, the fact that this volumetric analysis was carried out on
structural MRI volumes not concurrent with the period during which
testing was conducted for this study prevents any valid correlation
analyses between test scores and volumetric measurements. Second,
volumetric studies in SD patients reveal a pattern of atrophy that is
typically asymmetrical, affecting most often left temporal lobe regions
(Chan et al., 2001; Galton et al., 2001). It is possible therefore that a
number of the SD patients included in our study did not have sufcient
damage, either to the perirhinal cortex or involving this structure
bilaterally, to produce a signicant decit in recognition memory.
Further studies of aspects of memory function in SD, particularly those
that address perirhinal cortex function, will need to combine volumetric measurements of regions within the MTL with concurrent
performance on more cognitively clean neuropsychological measures
of perirhinal cortex function.
Lastly, it is possible that the lack of a signicant decit in the current
study and previous studies of novel recognition memory in perirhinal
lesion patients may be attributed to the use of stimuli of insufcient
visual complexity (e.g. Buckley et al., 2001; Bussey & Saksida, 2002;
Bussey et al., 2002). Relevant to this issue are the ndings of Stark &
Squire (2000) who observed that patients with perirhinal damage were
able to perform simple visual discriminations between stimuli of
differing shape, colour and size (thought to be dependent upon TE
regions), as well as on more difcult discriminations between photographs of different three-dimensional objects in a range of orientations
(demonstrated to be dependent upon perirhinal cortex in monkeys,
Buckley & Gaffan, 1998; Buckley et al., 2001). This nding implies
that the human perirhinal cortex may not subserve the same visual
functions as those suggested from the non-human primate studies (e.g.
Murray & Bussey, 1999; Buckley & Gaffan, 2000), although it is notable
that Stark & Squire (2000) used only a relatively small set of stimuli.

Summary
Patients with fvFTD and SD showed better episodic memory, as
measured by a paired associate learning task requiring retrieval of
objectlocation information, than a matched group of DAT patients.
Moreover, the two FTD groups performed better than the DAT group
on a delayed-matching-to-sample task comprising novel stimuli with
different patterns and colours, although this numerical difference did
not reach statistical signicance. These ndings support the accruing

2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

Memory in frontotemporal dementia


evidence that patients with FTD rarely demonstrate the profound
amnesic decits that are typically the rst cognitive symptom in
DAT, and that tests that require episodic memory, particularly
objectlocation association, provide a useful avenue for early differentiation of these two conditions. Importantly, however, our study
revealed that the inclusion of additional problems in the PAL, for
instance a 4- or 5-item problem or extra trials at critical stages of the
task, may increase its usefulness in the clinic.
Our study also addressed ongoing issues about the cognitive and
neural organization of aspects of long-term memory, and demonstrated
that whereas SD patients performed numerically poorer compared with
healthy controls on a difcult MTS task that utilized complex novel
stimuli, this difference was not statistically signicant. Considered in
isolation, this nding may suggest that SD patients, who have signicant atrophy to the perirhinal cortex, do indeed have difculties in
higher-order perception, as predicted by recent animal-informed models of MTL function. By contrast, when considered in conjunction with
the existing literature, our nding supports a recent stream of evidence
from studies in focal lesion cases (Buffalo et al., 1998; Holdstock et al.,
2000; Stark & Squire, 2000) that suggest that the human perirhinal
cortex does not subserve perceptual processes. To resolve this issue,
future studies will have to use more cognitively clean and demanding
measures of perirhinal cortex function to elucidate the function of this
region in humans.

Acknowledgements
We would like to thank all participating subjects, Andrew Blackwell and Robyn
Vessey (Department of Psychiatry, University of Cambridge, UK), Rachel
Swainson (School of Psychology, University of Nottingham, UK), Adrian
Owen (MRC-CBU, Cambridge, UK) and Eleanor Toye and Prisha Shah
(Cambridge Cognition plc, Cambridge, UK) for their assistance with the current
project. This work was funded by the MRC, UK and Alzheimer's Research
Trust, UK, and partly completed within an MRC Centre for Clinical and
Behavioural Neuroscience.

Abbreviations
CANTAB, Cambridge Neuropsychological Test Automated Battery; DAT,
dementia of the Alzheimer's Type; FTD, frontotemporal dementia; fvFTD,
frontal variant frontotemporal dementia; MMSE, Mini Mental State Examination; MTL, medial temporal lobe; MTS, Matching to Sample task; PAL, Paired
Associates Learning task; SD, semantic dementia.

References
Aggleton, J.P. & Brown, M.W. (1999) Episodic emmory, amnesia and the
hippocampalanterior thalamic axis. Behav. Brain Sci., 22, 425289.
Alvarez, P., Zola-Morgan, S. & Squire, L.R. (1995) Damage limited to the
hippocampal region produces long-lasting memory impairment in monkeys.
J. Neurosci., 15, 37963807.
Bachevalier, J. (1994) Medial temporal lobe structures and autism: a review of
clinical and experimental ndings. Neuropsychologia, 32, 627648.
Bachevalier, J. & Mishkin, M. (1994) Effects of selective neonatal temporal
lobe lesions on visual recognition memory in rhesus monkeys. J. Neurosci.,
14, 21282139.
Bozeat, S., Gregory, C.A., Ralph, M.A. & Hodges, J.R. (2000) Which neuropsychiatric and behavioural features distinguish frontal and temporal
variants of frontotemporal dementia from Alzheimer's disease? J. Neurol.
Neurosurg. Psychiatry, 69, 178186.
Buckley, M.J., Booth, M.C.A., Rolls, E.T. & Gaffan, D. (2001) Selective
perceptual impairments following perirhinal cortex ablation. J. Neurosci.,
21, 98249836.
Buckley, M.J. & Gaffan, D. (1997a) Impairment of visual object-discrimination
learning after perirhinal cortex ablation. Behav. Neurosci., 111, 467475.
Buckley, M.J. & Gaffan, D. (1997b) Learning and transfer of objectreward
associations and the role of the perirhinal cortex. Behav. Neurosci., 112, 19.

1669

Buckley, M.J. & Gaffan, D. (1998) Perirhinal cortex ablation impairs congural learning and paired-associate learning equally. Neuropsychologia, 36,
535546.
Buckley, M.J. & Gaffan, D. (2000) The hippocampus, perirhinal cortex and
memory in the monkey. In Bolhuis, J.J. (Ed.), Brain, Perception, Memory:
Advances in Cognitive Neuroscience. Oxford University Press, Oxford, pp.
279298.
Buffalo, E.A., Reber, P.J. & Squire, L.R. (1998) The human perirhinal cortex
and recognition memory. Hippocampus, 8, 330339.
Bussey, T.J. & Saksida, L.M. (2002) The organization of visual object representations: a connectionist model of effects of lesions in perirhinal cortex.
Eur. J. Neurosci., 15, 355364.
Bussey, T.J., Saksida, L.M. & Murray, E.A. (2002) Perirhinal cortex resolves
feature ambiguity in complex visual discriminations. Eur. J. Neurosci., 15,
365374.
Chan, D., Fox, N.C., Scahill, R.I., Crum, W.R., Whitwell, J.L., Leschziner, G.,
Rossor, A.M., Stevens, J.M., Cipolotti, L. & Rossor, M.N. (2001) Patterns of
temporal lobe atrophy in semantic dementia and Alzheimer's disease. Ann.
Neurol., 49, 433442.
Craik, F.I., Morris, L.W., Morris, R.G. & Loewen, E.R. (1990) Relations
between source amnesia and frontal lobe functioning in older adults. Psychol.
Aging, 5, 148151.
Davies, R.R., Xuereb, J.H. & Hodges, J.R. (2002) The human perirhinal cortex
in semantic memory: an in vivo and postmortem volume tric magnetic
resonance imaging study in semantic dementia, Alzheimer's disease and
matched controls. Neuropathol. Appl. Neurobiol., 28, 167168.
Dimitrov, M., Granetz, J., Peterson, M., Hollnagel, C., Alexander, G. &
Grafman, J. (1999) Associative learning impairments in patients with frontal
lobe damage. Brain Cogn., 41, 213230.
Eacott, M.J., Gaffan, D. & Murray, E.A. (1994) Preserved recognition memory
for small sets, and impaired stimulus identication for large sets, following
rhinal cortex ablations in monkeys. Eur. J. Neurosci., 6, 14661478.
Fowler, K.S., Saling, M.M., Conway, E.L., Semple, J.M. & Louis, W.J. (1997)
Computerized neuropsychological tests in the early detection of dementia:
prospective ndings. J. Int. Neuropsychol. Soc., 3, 139146.
Galton, C.J., Patterson, K., Graham, K., Lambon-Ralph, M.A., Williams, G.,
Antoun, N., Sahakian, B.J. & Hodges, J.R. (2001) Differing patterns of
temporal atrophy in Alzheimer's disease and semantic dementia. Neurology,
57, 216225.
Graham, K.S., Becker, J.T. & Hodges, J.R. (1997) On the relationship between
knowledge and memory for pictures: evidence from the study of patients with
semantic dementia and Alzheimer's disease. J. Int. Neuropsychol. Soc., 3,
534544.
Graham, K.S., Simons, J.S., Pratt, K.H., Patterson, K. & Hodges, J.R. (2000)
Insights from semantic dementia on the relationship between episodic and
semantic memory. Neuropsychologia, 38, 313324.
Gregory, C.A. & Hodges, J.R. (1996) Dementia of frontal type: use of consensus
criteria and prevalence of psychiatric features. Neuropsychiatr. Neuropsychol. Behav. Neurol., 9, 145153.
Hodges, J.R. & Miller, B. (2001a) The neuropsychology of frontal variant
frontotemporal dementia and semantic dementia. Introduction to the special
topic papers: Part II. Neurocase, 7, 113121.
Hodges, J.R. & Miller, B. (2001b) The classication, genetics and neuropathology of frontotemporal dementia. Introduction to the special topic papers: Part
I. Neurocase, 7, 3135.
Hodges, J.R. & Patterson, K. (1995) Is semantic memory consistently impaired
early in the course of Alzheimer's disease? Neuroanatomical and diagnostic
implications. Neuropsychologia, 33, 441459.
Hodges, J.R., Patterson, K., Oxbury, S. & Funnell, E. (1992) Semantic
dementia. Progressive uent aphasia with temporal lobe atrophy. Brain,
115, 17831806.
Hodges, J.R., Patterson, K., Ward, R., Garrard, P., Bak, T., Perry, R. & Gregory,
C. (1999) The differentiation of semantic dementia and frontal lobe dementia
(temporal and frontal variants of frontotemporal dementia) from early
Alzheimer's disease: a comparative neuropsychological study. Neuropsychology, 13, 3140.
Holdstock, J.S., Gutnikov, S.A., Gaffan, D. & Mayes, A.R. (2000) Perceptual
and mnemonic matching-to-sample in humans: contributions of the hippocampus, perirhinal and other medial temporal lobe cortices. Cortex, 36,
301322.
Howard, D. & Patterson, K. (1992) Pyramids and Palmtrees: a Test of Semantic
Access from Words and Pictures. Thames Valley Test Company, Bury St
Edmunds, UK.
Howell, D.C. (1997) Statistical Methods for Psychology. Duberry Press,
Belmont, CA, USA.

2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

1670 A. C. H. Lee et al.


Johnsrude, I.S., Owen, A.M., Crane, J., Milner, B. & Evans, A.C. (1999) A
cognitive activation study of memory for spatial relationships. Neuropsychologia, 37, 829841.
Maguire, E.A., Frackowiak, R.S. & Frith, C.D. (1996) Learning to nd your
way: a role for the human hippocampal formation. Proc. R. Soc. Lond. B Biol.
Sci., 263, 17451750.
McKahnn, G., Drachman, D., Folstein, M., Katzman, R., Price, D. & Stadlan,
E.M. (1984) Clinical diagnosis of Alzheimer's disease: report of the
NINCDS-ADRDA Work Group under the auspices of Department of Health
and Human Services Task Force on Alzheimer's Disease. Neurology, 34,
939944.
Meunier, M., Bachevalier, J., Mishkin, M. & Murray, E.A. (1993) Effects on
visual recognition of combined and separate ablations of the entorhinal and
perirhinal cortex in rhesus monkeys. J. Neurosci., 13, 54185432.
Miller, B.L., Darby, A.L., Swartz, J.R., Yener, G.G. & Mena, I. (1995) Dietary
changes, compulsions and sexual behavior in frontotemporal degeneration.
Dementia, 6, 195199.
Murray, E.A. (2000) Memory for objects in nonhuman primates. In Gazzaniga,
M. (Ed.), The Cognitive Neurosciences, 2nd edn. MIT Press, Cambridge,
MA, pp. 753763.
Murray, E.A., Baxter, M.G. & Gaffan, D. (1998) Monkeys with rhinal cortex
damage or neurotoxic hippocampal lesions are impaired on spatial scene
learning and object reversals. Behav. Neurosci., 112, 129112303.
Murray, E.A. & Bussey, T.J. (1999) Perceptualmnemonic functions of the
perirhinal cortex. Trends Cog. Sci., 3, 142151.
Murray, E.A. & Mishkin, M. (1998) Object recognition and location memory in
monkeys with excitotoxic lesions of the amygdala and hippocampus. J.
Neurosci., 18, 65686582.
Neary, D., Snowden, J.S., Gustafson, L., Passant, U., Stuss, D. & Black, S.
(1998) Frontotemporal lobar degeneration: a consensus on clinical diagnostic
criteria. Neurology, 51, 15461554.
Owen, A.M., Milner, B., Petrides, M. & Evans, A.C. (1996) Memory for object
features versus memory for object location: a positron-emission tomography
study of encoding and retrieval processes. Proc. Natl Acad. Sci. USA, 93,
92129217.
Perry, R.J. & Hodges, J.R. (2000) Differentiating frontal and temporal
variant frontotemporal dementia from Alzheimer's disease. Neurology, 54,
22772284.

Rahman, S., Sahakian, B.J., Hodges, J.R., Rogers, R.D. & Robbins, T.W. (1999)
Specic cognitive decits in mild frontal variant frontotemporal dementia.
Brain, 122, 14691493.
Ratnavalli, E., Brayne, C., Dawson, K. & Hodges, J.R. (2002) The prevalence of
frontotemporal dementia. Neurology, 58, 16151621.
Schacter, D., Harbluk, J. & McLachlan, D. (1984) Retrieval without recollection: an experimental analysis of source amnesia. J. Verb. Learn. Verb.
Behav., 23, 593611.
Simons, J.S., Graham, K.S., Galton, C.J., Patterson, K. & Hodges, J.R. (2001)
Semantic knowledge and episodic memory for faces in semantic dementia.
Neuropsychology, 15, 101114.
Simons, J.S., Verfaellie, M., Galton, C.J., Miller, B.L., Hodges, J.R.
& Graham, K.S. (2002) Recollection-based memory in frontotemporal
dementia: implications for theories of long-term memory. Brain, 125,
25232536.
Stark, C.E.L. & Squire, L.R. (2000) Intact visual perceptual discrimination in
humans in the absence of perirhinal cortex. Learn. Mem., 7, 273278.
Suzuki, W.A., Zola-Morgan, S., Squire, L.R. & Amaral, D.G. (1993) Lesions of
the perirhinal and parahippocampal cortices in the monkey produce longlasting memory impairment in the visual and tactual modalities. J. Neurosci.,
13, 24302451.
Swainson, R., Hodges, J.R., Galton, C.J., Semple, J., Michael, A., Dunn, B.D.,
Iddon, J.L., Robbins, T.W. & Sahakian, B.J. (2001) Early detection and
differential diagnosis of Alzheimer's disease and depression with neuropsychological tasks. Dement. Geriatr. Cogn. Disord, 12, 265280.
Tulving, E. (1995) Organization of memory: quo vadis? In Gazzaniga,
M. (Ed.), The Cognitive Neurosciences. MIT Press, Cambridge, MA,
pp. 839847.
Tulving, E. (2001) Episodic memory and common sense: how far apart? Phil.
Trans. R. Soc. Lond. B Biol. Sci., 356, 15051515.
Ungerleider, L. & Mishkin, M. (1982) Two cortical visual systems. In Ingle, D.,
Goodale, M. & Manseld, R. (Eds), Analysis of Visual Behavior. MIT
Press, Cambridge, MA, pp. 549586.
Warrington, E. (1984) Recognition Memory Test. NFER-Nelson, Windsor, UK.
Zola-Morgan, S., Squire, L.R., Amaral, D.G. & Suzuki, W.A. (1989) Lesions
of perirhinal and parahippocampal cortex that spare the amygdala and
hippocampal formation produce severe memory impairment. J. Neurosci.,
9, 43554370.

2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 16601670

Vous aimerez peut-être aussi