Neuropsychologia 50 (2012) 592602

Contents lists available at SciVerse ScienceDirect

Neuropsychologia
journal homepage: www.elsevier.com/locate/neuropsychologia

Using virtual reality to characterize episodic memory proles in amnestic mild

cognitive impairment and Alzheimers disease: Inuence of active and passive
encoding
G. Plancher, A. Tirard, V. Gyselinck, S. Nicolas, P. Piolino
Memory and Cognition Lab, Paris Descartes University, Paris, France

a r t i c l e

i n f o

Article history:
Received 9 April 2011
Received in revised form
18 December 2011
Accepted 20 December 2011
Available online 13 January 2012
Keywords:
Episodic memory
Long-term binding
Encoding
Virtual reality
Aging
Dementia
Mild cognitive impairment

a b s t r a c t
Most neuropsychological assessments of episodic memory bear little similarity to the events that patients
actually experience as memories in daily life. The rst aim of this study was to use a virtual environment
to characterize episodic memory proles in an ecological fashion, which includes memory for central
and perceptual details, spatiotemporal contextual elements, and binding. This study included subjects
from three different populations: healthy older adults, patients with amnestic mild cognitive impairment
(aMCI) and patients with early to moderate Alzheimers disease (AD). Second, we sought to determine
whether environmental factors that can affect encoding (active vs. passive exploration) inuence memory
performance in pathological aging. Third, we benchmarked the results of our virtual reality episodic
memory test against a classical memory test and a subjective daily memory complaint scale. Here, the
participants were successively immersed in two virtual environments; the rst, as the driver of a virtual
car (active exploration) and the second, as the passenger of that car (passive exploration). Subjects were
instructed to encode all elements of the environment as well as the associated spatiotemporal contexts.
Following each immersion, we assessed the patients recall and recognition of central information (i.e., the
elements of the environment), contextual information (i.e., temporal, egocentric and allocentric spatial
information) and lastly, the quality of binding. We found that the AD patients performances were inferior
to that of the aMCI and even more to that of the healthy aged groups, in line with the progression of
hippocampal atrophy reported in the literature. Spatial allocentric memory assessments were found
to be particularly useful for distinguishing aMCI patients from healthy older adults. Active exploration
yielded enhanced recall of central and allocentric spatial information, as well as binding in all groups. This
led aMCI patients to achieve better performance scores on immediate temporal memory tasks. Finally,
the patients daily memory complaints were more highly correlated with the performances on the virtual
test than with their performances on the classical memory test. Taken together, these results highlight
specic cognitive differences found between these three populations that may provide additional insight
into the early diagnosis and rehabilitation of pathological aging. In particular, neuropsychological studies
would benet to use virtual tests and a multi-component approach to assess episodic memory, and
encourage active encoding of information in patients suffering from mild or severe age-related memory
impairment. The benecial effect of active encoding on episodic memory in aMCI and early to moderate
AD is discussed in the context of relatively preserved frontal and motor brain functions implicated in
self-referential effects and procedural abilities.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
Episodic memory may be described as the conscious recollection of personal events combined with their phenomenological

Corresponding author at: Memory and Cognition Lab, 71 avenue Edouard Vaillant, 92774, Boulogne-Billancourt, France. Tel.: +33 1 55 20 59 22;
fax: +33 1 55 20 58 53.
E-mail address: pascale.piolino@parisdescartes.fr (P. Piolino).
0028-3932/$ see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.neuropsychologia.2011.12.013

and spatiotemporal encoding contexts (Tulving, 2002). The multiple components of memory (central and contextual information)
that together form a complete episodic memory are thought to be
linked through a process known as binding (Kessels, Hobbel, &
Postma, 2007), mainly supported by the hippocampus (Slotnick,
2010). Numerous studies have demonstrated that episodic memory
impairment is one of the hallmarks of early clinical manifestations of Alzheimers disease (AD) (Hodges, 2006) and amnesic
mild cognitive impairment (aMCI) (Petersen et al., 2001, 1999). In
parallel, some researchers have reported reduced volume in the

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

hippocampus and in other brain regions such as parahippocampal

gyrus and cingulate cortex in both AD and aMCI patients (Chtelat
et al., 2002; De Leon et al., 2006; Dickerson et al., 2001; Fox and
Schott, 2004; Gerardin et al., 2009; Risacher et al., 2009). Typically,
MCI patients have intermediate brain volumes between AD patients
and healthy controls (Evans et al., 2010), with an atrophy of the hippocampus that is less spatially extended compared to AD (Gerardin
et al., 2009). In addition, their gray matter density in the posterior association cortex seems preserved compare to AD (Chtelat,
Villain, Desgranges, Eustache, & Baron, 2009).
Classical neuropsychological tools used to assess episodic memory are far from encompassing the complexity of this kind of
memory (for review, Piolino, Desgranges, & Eustache, 2009).
Episodic memory is generally assessed with verbal tasks, although
much of what people remember in everyday life refers to visual
information and to actions that they have performed. The objective
of our study was to demonstrate that virtual environments can be
used as tools to characterize, in detail, the episodic memory proles
of AD and aMCI patients through tasks that parallel requirements
of daily life.
So far, the state of central memory in AD and aMCI, such as memory for items, has been mainly evaluated with words, indicating
decits of free and cued recalls and recognition during AD (Salmon
& Bondi, 2009; Spaan, Raaijmakers, & Jonker, 2003, for review),
and to a lesser extent, aMCI (Perri, Carlesimo, Serra, & Caltagirone,
2005). Several studies have focused on the retrieval of contextual information in both populations and have observed spatial
memory decits (for review, Iachini, Iavarone, Senese, Ruotolo, &
Ruggiero, 2009), including egocentric and allocentric impairments
(Hort et al., 2007; Lacz et al., 2009), as well as temporal memory
decits (Putzke, Rickert, Duke, Marson, & Harrell, 2000). However, DeIpolyi, Rankin, Mucke, Miller, and Gorno-Tempini (2007)
found that after immersion in a real-life contextual environment,
aMCI patients recognized landmarks just as effectively as controls
but could not nd their locations on maps or recall the order in
which they had been encountered. In addition, decreases in binding quality have been observed in visuospatial (item with location)
and verbal associative learning tasks (placing two words together)
(Fowler, Saling, Conway, Semple, & Louis, 2002; Swainson et al.,
2001; Troyer et al., 2008). Overall, previous studies have generally
led researchers to assume that AD patients suffer from a general
episodic memory decit and aMCI patients from a less profound
one corresponding to the progression of the atrophy (Chtelat et al.,
2009; Evans et al., 2010).
However, various clinical memory tests used in dementia generally measure only one aspect of episodic memory in isolation, rather
than offering a complete measure of its components such as memory for what, where, and when, and binding between components.
This is in sharp contrast with the requirements of daily life. Moreover, the conditions of these tests are generally far from ecological.
Many studies have argued that neuropsychological assessments
should bear some degree of similarity to the demands of daily life
(Farias, Harrell, Neumann, & Houtz, 2003; Schultheis, Himelstein,
& Rizzo, 2002; Wojtasik et al., 2010). Indeed, it has been demonstrated that standard neuropsychological memory assessments are
only moderately correlated with behavior-based measures of daily
living skills (e.g., dialing a telephone, selecting shopping items with
a written list, etc.) (Farias et al., 2003).
To improve the diagnosis and the rehabilitation of aMCI and
AD patients, studies can take advantage of new technologies used
to develop paradigms that are able to detect functional changes
in cognitive abilities and present conditions that resemble daily
life. By assessing the various components of an episodic memory
(factual, spatial, temporal, and binding) both simultaneously and
multimodally, virtual reality (VR) enables the performance of an
integrated assessment of episodic memory. To date, only a few

593

neuropsychological studies have used VR to test memory in pathological aging. Previous studies have mainly focused on navigational
processes. Some studies with aMCI patients (Cushman, Stein, &
Duffy, 2008) have found a close relationship between performance
in virtual and real environments. Moreover, other studies with
AD patients (Burgess, Trinkler, King, Kennedy, & Cipolotti, 2006;
Drzezga et al., 2005; Zakzanis, Quintin, Graham, & Mraz, 2009)
have specically found allocentric spatial impairments. Widmann,
Beinhoof and Riepe (2012) immersed AD patients and healthy participants in a virtual environment to assess the learning of verbal
material in situations that imitate natural conditions. AD patients
were found to be impaired in free memory recall of shop names
compared to healthy participants, and the impairment was more
marked than that observed with classical list-learning. The study
argued that list-learning paradigms wrongly estimated the memory capacities of patients in every day situations. Recently, we
developed a virtual reality test that seeks to reect the denition of
episodic memory more closely than other standard neuropsychological tests (for review, Plancher, Nicolas, & Piolino, 2008). Various
aspects of episodic memory (what, where, when, and binding)
were evaluated in a population of young adults and healthy older
adults after they drove a car in a virtual environment representing
a city with different, but specic, areas and elements (Plancher,
Gyselinck, Nicolas, & Piolino, 2010). We observed a difference
between the healthy aged controls and younger participants in
memory for spatiotemporal context and binding. Moreover, the virtual test was sensitive to the older subjects memory complaints,
which was contrary to the standard verbal episodic memory test.
One aim of the present study, which follows-up on results
obtained by Plancher et al. (2010), was to use this new VR test to
determine changes in the various components of episodic memory
in pathological aging by comparing aMCI and AD patients with a
new population of healthy aged controls. In addition, we aimed to
test whether environmental inuences on encoding could improve
memory performance. Precisely, we sought to determine if active
exploration of the virtual environment improves memory compared to passive exploration. It has been previously demonstrated
that presenting information in the memory environment can serve
as compensatory support for decient self-initiated processing
and can enhance memory performance (Craik, 1983, 1986; Luo &
Craik, 2008; Naveh-Benjamin, Craik, & Ben-Shaul, 2002, for review).
For example, positive effects of environmental support have been
observed when older participants beneted from the provision of
more elaborate pictorial information (Park, Puglisi, & Smith, 1986;
Schacter, Israel, & Racine, 1999) or context and semantic organizational structure of materials (Park, Smith, Morrell, Puglisi, & Dudley,
1990). An active exploration involves cognitive processes that are
different than those engaged during passive exploration. In particular, personal involvement is higher in active exploration, which
may potentially result in a self-reference effect that is known to
improve memory, even in aging (Gutchess, Kensinger, & Schacter,
2010; Lalanne et al., 2010; Ruby et al., 2009). In addition, encoding information through motion may solicit, even if it is far from
real movements, some procedural skills. Self-referential effect and
procedural skills mainly depend on medial prefrontal cortex (Craik
et al., 1999; Kelley et al., 2002; Martinelli, Sperduti, & Piolino,
in press), striatum and motor cortex respectively (Nilsson et al.,
2000; Nyberg et al., 2001; Pennartz, Ito, Verschure, Battaglia, &
Robbins, 2011; Squire, 2004). These processes and the underlying
cortical structures would be partially preserved in mild AD (Chase
et al., 1984; Grady et al., 2003). Learning based on implicit procedural skills appears to be one of the best preserved skills found
in AD patients (Deweer et al., 1994; Deweer, Pillon, Michon, &
Dubois, 1993; Gabrieli, Corkin, Mickel, & Growdon, 1993; Hirono
et al., 1997; Lipinska & Backman, 1997; Van Halteren-van Tilborg,
Scherder, & Hulstijn, 2007). Further, some studies have previously

594

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

shown that using implicit procedural encoding, instead of explicit

verbal encoding, can improve the AD patients recall of action
sentences (e.g., peel a banana) at all stages of the disease (early,
moderate and severe) (Cohen, 1981; Karlsson, Backman l. Herlitz,
Nilsson l., Winblad, & Osterlind, 1989; Lekeu, Van der Linden,
Moonen, & Salmon, 2002). Cohen (1989) suggested that the superiority of action could be explained partly by the non-strategic nature
of encoding during motor action and by the addition of a motor
component to the memory trace.
In this study, our rst objective was to differentiate between
the episodic memory proles of aMCI and mild AD patients compared to normal aging, using a virtual reality environment to assess
central and contextual episodic memory as well as binding functions with tasks that involve immediate and delayed recalls and
immediate recognition. We expected that episodic memory performance would be inferior in mild AD than in aMCI and even more
in healthy aged groups, in line with the established progression of
atrophy in the hippocampus. Our second objective was to test the
effects of a facilitating factor on the pathological populations with
reference to encoding. We expected that active encoding would
improve memory performance compared to passive encoding in
aMCI and mild Alzheimers disease since it may rely on cerebral
regions which are still partially preserved. Finally, we sought to
demonstrate that complex virtual environments may provide tools
to reect subjective cognitive decits in pathological aging, by
calculating and comparing the correlations of both of our virtual
reality and verbal classical memory tests with a daily memory complaint scale.

2. Methods
2.1. Participants and experimental design
Twenty-one healthy older participants (17 female and 4 male; mean age = 76.95;
SD = 5.8; min = 67, max = 88), 15 aMCI patients (8 female and 7 male; mean
age = 78.06; SD = 5.76; min = 70, max = 88), and 15 patients with AD (13 female and
2 male; mean age = 76.53; SD = 5.51; min = 69, max = 88) voluntarily participated in
this study. Early to moderate AD and aMCI patients were recruited from hospitals
in Paris, France, according to the criteria of the National Institute of Neurological
Disorders and Stroke and the Alzheimers Disease and Related Disorders Association (McKahnn et al., 1984). MCI participants met the criteria for an amnestic
single-domain form of MCI as dened by Petersen et al. (2001). The clinical diagnostic criteria for aMCI included the presence of a memory complaint, impaired
memory function (recall performance on a standard episodic memory test 1.5
standard deviations below age- and education-matched norms), preserved general
cognitive function, intact activities of daily living, and the absence of dementia.
All participants gave their informed consent before beginning the study. The study
was performed in compliance with the Declaration of Helsinki. The healthy participants were recruited from a local panel of volunteers and also lled in a medical
questionnaire to verify that they had no previous neurological or psychiatric disorders and were not undergoing any treatments that would affect their cognition.
All participants had to meet several criteria prior to being included in this study:
no decit in color recognition (color blindness test), no hemi-neglect (clock test)
and a visual eld that covers a minimum angle of 46 (visual eld test). In addition,
the Mini Mental State Exam (MMSE) was used to evaluate the level of dementia
of each group (Folstein, Folstein, & McHugh, 1975). Healthy older adults scored
at least 27 (max = 30) (indicating an absence of dementia), AD patients scored at
least 14 (indicating an early to moderate stage of dementia) and aMCI patients
scored at least 24. Participants also received the minimal version of the Geriatric
Depression Scale (Clment, Nassif, Lger, & Marchan, 1997; Yesavage, 1988) to
ensure that depressive conditions would not affect the participants memory performance. Given that experiences with driving in real life could be related to abilities
used to drive a virtual car, all participants were required to have a drivers license.
Finally, the level of education and gender distribution was different among the
groups, as the aMCI patients level of education and number of men was signicantly greater than that of the other groups (p < .01, 8 female participants and 7
male participants). Thus, we included both variables as covariates in all statistical
analyses.
The present study manipulated three variables: population (healthy older adults,
aMCI patients, AD patients), exploration (active and passive) and recall (immediate
and delayed) within participants. Each exploration was associated with one environment. The assignment of the exploration to the environment and the order in
which participants encountered them were counterbalanced.

2.2. Neuropsychological assessment

Each participant completed several neuropsychological tests. Demographic data
and neuropsychological scores are presented for each population in Table 1. The subjective Cognitive Difculties Scale (CDS, McNair & Kahn, 1983) was used to assess
cognitive complaints in daily life. The CDS is a 39-item self-report measure that
uses a Likert-type scale. For each item (e.g., Do you have difculty remembering the
names of the people you know?), the participants have to choose among a range of ve
responses: 1 corresponds to never and 5 corresponds to very often. Semantic
abilities were evaluated with the 44-item Mill Hill test (French translation; Deltour,
1993), which is comprised of 44 vocabulary questions. We investigated executive
functions and episodic memory using well-established tests. First, exibility was
measured with the Trail-Making Test (TMT) (Lezak, Howieson, & Loring, 2004). Second, inhibition abilities were evaluated using the interference score from the Stroop
test (Stroop, 1935). Short-term memory was assessed with the forward digit span,
and working memory (the ability to simultaneously maintain and process information) was evaluated with the backward digit span of the Wechsler Adult Intelligence
Scale (WAIS) (Wechsler, 2001). To test verbal episodic memory, participants were
given the word-learning test described by Grober and Buschke (1987).
2.3. Materials
2.3.1. Virtual apparatus
The virtual environment was created with Virtools Dev 3.0 software
(www.virtools.com). A PC laptop computer was used to present a three-dimensional
view of the environment. In the active condition, the environment was explored with
a virtual car using a real steering wheel, a gas pedal and a brake pedal. A video projector was used to project the environment onto a screen that was 85 cm high and
110 cm wide. The participants were seated in a comfortable chair, and the virtual
environment was projected 150 cm in front of them.
2.3.2. Virtual environments
Two urban environments inspired by photos of Paris were created, one for each
exploration (see Fig. 1 for an example of one environment). In each town, one route
connected nine specic scenes. Each specic scene was comprised of different elements: one central element (e.g., a newsstand) and two or three secondary elements
(e.g., a man or a bench). In the rst town, called city, the order in which the nine
scenes were encountered (identied by the main element in each scene) was the
following: a train station, a newsstand, a post ofce, a road work zone, a fountain,
an old building, a parking lot, a kebab shop, and a set of shops. In the second town,
called town, the nine scenes were as follows: a park, an area of tall buildings, a
set of restaurants, some parked cars, a grocery store, bus stops, a bar-tabac (a bar
that also sells tobacco), a town hall, and a set of billboards. In both environments,
specic areas were located at a turn (see Fig. 1). In addition, a soundtrack of typical
city noises (cars, people, etc.) helped enable the participants to feel immersed in
each environment.
2.4. Procedure
2.4.1. Training sessions
All participants underwent two training sessions in an empty environment, one
before each immersion. If the rst immersion was active, then the training session
was active as well. Similarly, if the rst immersion was passive, then the training
session was passive as well. The spatial layout of the training environment was
different from those that had been used subsequently for the test. This allowed
the participants the opportunity to familiarize themselves with virtual reality and
with the control of the virtual car. The training session lasted until the participants
felt familiar with the equipment. They were then free to navigate anywhere on the
training track. Subsequently, the participants performed the test session.
2.4.2. Test session
First, the participants were immersed in one environment (city or town), as
either active or passive explorers. In each virtual town, active participants navigated
through a single route composed of ten turns (see Fig. 1). The active participants were
instructed to drive into the town, not stop, and to memorize all the elements of the
scenes that they encountered within the environment. In addition, they were asked
to remember details, specically, the temporal and spatial context associated with
the elements so that they would be able to recall them at the end of the presentation. The passive participants received the same instructions except for the fact that
they would be driven in the town. To match the active and passive conditions, each
active path was recorded and presented to another participant of the same group
as a passive path. Passive participants only watched the video and did not drive or
decide the itinerary. Because the driving of an active participant was recorded and
played back to the passive subject, the visual information displayed was the same
as that of the active subject except that the passive subject had the point of view of
a passenger (rather than the driver).
The immersion began when participants had no further questions regarding the
task and ended when participants reached the edge of the town. The immersion
time depended on the speed of the participants. Our virtual apparatus was built so
that the range between the slow and fast speed was very small. Further, all of the

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

595

Table 1
Means and SDs of demographics and general neuropsychological abilities.
Older adults (n = 21)

aMCI patients (n = 15)

AD patients (n = 15)

ANOVA/ANCOVAS

Age (years)
Level of educationd
Gender (male:female)

76.9 (5.8)
3.5 (1.5)
(4:17)

78 (5.7)
5 (1.7)
(7:8)

76.5 (5.5)
3.2 (1.5) b , **
(2:13)

F(2, 48) = 5.6**

Depression
Mini GDS (cut-off <2/5)

.4 (.9)

.1 (.3)

.8 (1.3)

F<1

Global cognition
MMSE
CDS

28.7 (.8)
22.7 (12.6)

26.5 (1.6)c , *
43.5 (14.5)c , *

19.3 (3.5)a , *** , b , ***

74.7 (32.7)a , *** ,b***

F(2, 47) = 84.2***

F(2, 47) = 27.2***

Semantic abilities
Mill Hill

34.37 (7.1)

34.06 (7.6)

25 (7.3)a , *** , b , **

F(2, 47) = 5.3**

Executive functions
TMTA (s)
TMTB-A (s)
Inhibition
Forward span
Backward span

48.8 (13.4)
160.4 (119)
16.4 (23.8)
5.8 (.3)
3.4 (.9)

52.3 (24.1)
65 (46)
34.2 (17.5)
5.2 (1.1)c , *
2.8 (.6)c , *

128.8 (80.6)a , *** , b , *

198 (87.4)b , ***
17.2 (23.7)
4.4 (.8)a , *** , b , *
2.3 (.5)a , ***

F(2, 47) = 12.8***

F(2, 47) = 3.85*
F(2, 47) = 1.16, n.s.
F(2, 47) = 15.4***
F(2, 47) = 10.1***

Verbal memory
Immediate recall
Total recall (3 trials)
Delayed total recall
Intrusions
Perseverations
Recognition: hits/false recognitions

15.9 (.21)
46.8 (1.1)
15.9 (.2)
0
0
15.7 (.6)/(0)

13.6 (1.8)c , **
32.2 (8.4)c , ***
11.8 (3)
1.6 (1.4)c , **
.7 (1.5)
13.5 (1)/1(1)

10.6 (3.1)a , *** , b , ***

19.8 (8.3)a , *** , b , ***
5 (3.2)
3.4 (3.2)a , *** , b , **
3.4 (3.2)a , *** , b , **
11.7 (5)a , *** /2 (3.3)a , *

F(2, 47) = 32.5***

F(2, 47) = 78.2***
F(2, 47) = 90.9***
F(2, 47) = 25.3***
F(2, 47) = 9.5***
F(2, 47) = 8.85*** F(2, 47) = 4.7*

Scores show impairment in Alzheimer disease (AD) relative to the mean scores of matched control older adults.
Scores show impairment in AD relative to the mean scores of amnesic Mild Cognitive Impairment (aMCI) patients.
c
Scores show impairments in aMCI relative to the mean scores of matched control older adults.
d
According to Poitrenaud. Level 1 = illiterate; 2 = able to read, write, count; 3 = 5 years of education, corresponding to the end of the elementary school; 4 = 8 years of
education; 5 = 10 years of education; 6 = Bachelors degree; 7 = postgraduate studies (Masters, Phd).
*
p < .05.
**
p < .01.
***
p < .001.
b

participants received instructions to not stop within the environment. Given that
each passive participant was matched with an active participant, the test times were
also globally matched across the conditions.
Immediately after the immersion, the participants performed recall and recognition tests that assessed their episodic memory of the environment. We used a series
of oral memory tests that were previously validated by a VR study on memory and
aging (Plancher et al., 2010). After 20 min, the participants performed the delayed
recall in a similar manner to the rst immediate recall. Likewise, the participants
also performed the neuropsychological tests during an interval of 20 min. Following, the participants were also immersed in a second environment, either as active
or passive explorers (the counterpart to the rst immersion), and were instructed

to follow the same procedure. After the second immediate recall, the participants
performed other neuropsychological tests. Subjects were assessed on two different
days to avoid overly long sessions.
2.4.3. Immediate recall test
When participants remembered an element, they had to spell out any remembered details as well as the associated spatiotemporal context. The experimenter
would note all recalls on a structured grid of responses. There was no specic order in
the recall of the components. Once the element was recalled, the participants could
then provide contextual recall in any order. The participants were never probed. For
each recall test, the instructions were associated with an example as follows:

Fig. 1. Overview of the city environment and picture of a specic area (newsstand with a man and a bench on the right).

596

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

- Try to remember all the elements you saw in the town (e.g., a girl, the post ofce,
the restaurants, a bench) (what, maximum = 35)
- Give as many perceptual details about the elements as possible (e.g., the girl was
wearing an orange t-shirt) (details)
- Situate the elements in time: were they at the beginning, the middle, or at the
end of the town? (when)
- Situate the elements of the scene to each other (e.g., the girl was to the left of
the post ofce) (allocentric where)
- Try to remember if you turned left or right after the element (egocentric where).
For example, one possible answer could be: at the beginning of the town (when),
there was a park (what) and a girl (what); the girl was wearing an orange tee-shirt
(detail); she was in front of the park (allocentric where); I turned left after this scene
(egocentric where). In total, 5 min were allowed for each verbal recall. Two experts
rated each memory and any difference of opinion between the experts was discussed
until a consensus was reached.
In addition, we computed a binding score. For each element recalled, we noted
whether the participants recalled the associated components (details, when, allocentric where and egocentric where). For example, if they recalled the shops, did they
recall where, when and the details associated with this element? The binding score
for a subject is the sum of all the contextual recalls associated with the elements
recalled. For example, if one participant recalled 5 elements with 1 piece of temporal information, 2 details and 4 allocentric spatial recalls in total, they had a binding
score of 7.
2.4.4. Recognition
The recognition task was composed of written sentences presented on the computer screen that described the presence of elements (e.g., there was a church in
the town), spatial relationships between elements (e.g., a man was at the right of
the newsstand), and temporal relationships between elements (e.g., the train station was before the fountain). There were eight sentences per category: the correct
response to half of the sentences was yes and to the other half was no. The test
included a total of 24 sentences. Participants gave their answers orally. The correct answers were split into what recognitions, where recognitions, and when
recognitions and corresponded to the number of correct hits and correct rejections. A recognition can rely on an episodic recollection (autonoetic consciousness)
as well as on a feeling of familiarity (noetic consciousness). To learn about the quality
of the recognitions in more detail, we complemented the recognition task with the
Remember/Know/Guess (recollection/familiarity/uncertainty) paradigm after each
recognition test (Gardiner, 1988, 2001; Tulving, 1983, 1985). In this paradigm, the
subjects were asked to verbalize aloud a judgment about the quality of their state
of mind associated with each of their recognitions. A recollection judgment means
that the subject based his/her recognition on the recollection of a specic event and
was re-experiencing it with details and the source of acquisition (e.g., I can mentally
view the train station in the town). A familiarity judgment corresponds to a feeling
of familiarity in the absence of any such recollection (e.g., I know the train station
was at the beginning of the town, but I cannot view it mentally). A guess judgment indicates that a subject was not sure of the response (e.g., I assume that there
was a train station). This category was added so that Know (K) responses would
not depend on the degree of uncertainty (Mntyl, 1993). To take independent
scores into account, we calculated the recollection (autonoetic consciousness) and
familiarity (noetic consciousness) scores as the mean with the following formula:
Recollection = R/n (n = maximum number of correct recognitions, i.e., = 8 by category) and Familiarity = (K/n)/(1 recollection score) recommended by Yonelinas,
Kroll, Dobbins, Lazzara, and Knight, 1998).
2.4.5. Delayed recall test
The delayed recall task was identical to the immediate recall task, with the
exception that it was administered 20 min after immersion. The instructions and
scores were documented similarly to those of the immediate recall task (please see
above). There was no delayed recognition test.

3. Results
The levels of education and gender were included as controlled
variables in all of the statistical analyses.
3.1. Neuropsychological assessment
ANCOVAs were conducted on each score to test for differences
between populations on neuropsychological tests (see Table 1 for a
summary of the results). As classically observed, the MMSE scores of
AD patients were less than those of aMCI patients, and those of aMCI
patients were less than those of the control group. Importantly,
the reverse pattern was observed for CDS scores. Furthermore, AD
patients were impaired on the Mill Hill, TMTA, and TMTB-A tests

Fig. 2. Recall performance (raw score) for what score according to population (older
adults, aMCI patients and AD patients) and exploration (active and passive).

compared to aMCI patients, while the aMCI patients performance

was similar to the control group. Finally, the scores of both patient
groups were impaired compared with baseline level assessments
of the forward and backward digit span.
An ANCOVA was performed on all scores from the Grober and
Buschke verbal memory assessment, with the population as a
between-subjects factor. Overall, the results (see Table 1) conrmed what has classically been observed in the literature; a decit
in free recall, cued recall, and recognition for AD patients (Salmon &
Bondi, 2009; Spaan et al., 2003) and memory impairment in recall
but not in recognition for aMCI patients (Fischer et al., 2007; Perri,
Serra, Carlesimo, & Caltagirone, 2007). These results attest that our
three groups were in fact different.
3.2. VR episodic memory assessment
Analyses were performed on the raw scores for what, details,
when, egocentric where, allocentric where, binding, and correct
recognitions with associated recollection and familiarity judgments
through a series of ANCOVAs with population (older adults, aMCI or
AD patients) as a between-subjects factor and exploration (active
or passive) and recall (immediate or delayed) as within-subjects
factors. To determine the direction of the differences, we carried
out post hoc Tukey tests. The effect sizes were represented with
partial eta squared (2p ).
3.2.1. Recall scores
The ANCOVA on what scores (see Fig. 2) revealed a strong effect
of population on the quantity of elements correctly recalled (F(2,
47) = 79.4, p < .001, 2p = .77), indicating that both the AD (p < .001)
and aMCI (p < .001) patients recalls were impaired compared to
healthy control participants. However, the aMCI group was signicantly less impaired than the AD group (p < .001). We also observed
an effect of exploration (F(1, 47) = 4.8, p < .05, 2p = .09), indicating
that active exploration led to a better performance than passive
exploration, although there was no effect of delay. No interaction
was found to be signicant.
An effect of population on the recall of details (see Fig. 3) (F(2,
47) = 22.9, p < .001, 2p = .50) was observed, indicating that both AD
(p < .001) and aMCI (p < .05) patients recall of details was impaired.
aMCI patients were found to be less impaired than AD patients
(p < .001). An effect of exploration (F(1, 47) = 4.9, p < .05, 2p = .10)
indicated that performance was better after passive exploration.
No other signicant result for details was found.
An effect of population was also observed for when scores (F(2,
47) = 31.3, p < .001, 2p = .58) indicating that AD patients (p < .001)
recalled less temporal information compared with healthy older
participants and the aMCI group, who were in turn less impaired
than the AD group (p < .001) (means of 4.0, 3.7 and 1.2 for the

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

597

Fig. 3. Recall performance (raw score) for detail score according to population and
exploration.

Fig. 5. Recall performance (raw score) for binding score according to population,
exploration and recall.

healthy older, aMCI and AD participants respectively). No significant differences appeared between the aMCI and healthy groups.
In addition, an interaction between population and recall (F(2,
47) = 6.5, p < .01, 2p = .22) showed that only AD (means of .8 and
1.6 for immediate and delayed recalls respectively) and aMCI
patients (means of 4.2 and 3.3) presented a difference between
immediate and delayed recalls (p < .001). Finally, a triple interaction (F(2, 46) = 6.5, p < .01, 2p = .23) indicated that aMCI patients
(p < .01) (means of 4.8 and 3.5 for active and passive exploration
respectively) and the control group (p < .01) (means of 4.7 and 3.6)
performed better after an active exploration but only at immediate
recall.
An effect of population on the egocentric where score (F(2,
47) = 38.7, p < .001, 2p = .63), was observed, indicating that AD
patients (mean of 1.1) (p < .001) and aMCI patients (mean of 3.2)
(p < .01) recalls were decient compared with healthy older adults
(mean of 4.3). Again, aMCI patients presented a smaller decit than
AD patients (p < .001). No signicant interaction was found.
An effect of population on the allocentric where score (F(2,
47) = 28.4, p < .001, 2p = .55), was observed (see Fig. 4), indicating
that AD (p < .001) and aMCI patients (p < .001) recalled signicantly
less allocentric spatial information than healthy older adults, with
aMCI patients exhibiting less impairment than AD patients (p < .01).
In addition, an effect of exploration (F(1, 47) = 5.3, p < .05, 2p = .10)
showed that active exploration led to lesser impairment in allocentric recall for all groups. No signicant interaction was found.
An effect of population on the binding performance (F(2,
47) = 39.4, p < .001, 2p = .63), was observed (see Fig. 5), indicating
that AD patients (p < .001) linked together less information than
older adults and aMCI patients (p < .001). The effect of population
interacted with recall (F(1, 47) = 3.5, p < .05, 2p = .13), indicating
that the aMCI patients delayed recall, but not immediate recall,

was impaired compared to the control group (p < .001). Finally, an

exploration effect (F(1, 47) = 7.5, p < .05, 2p = .13) showed better
performance after an active exploration for all groups.

Fig. 4. Recall performance (raw score) for allocentric where score according to
population and exploration.

3.2.2. Recognition scores

3.2.2.1. Correct recognitions. We performed an ANCOVA on the
correct recognitions with population (older adults, aMCI or AD
patients) as a between-subjects factor and kind of recognitions
(what, when and where) and exploration (active or passive) as
within-subjects factors. An effect of population was observed on
correct recognitions (F(2, 46) = 15.1, p < .001, 2p = .40), indicating that AD patients (mean of 3.6) recognized signicantly fewer
information than older adults (mean of 5.1) (p < .001) and that
aMCI patients (mean of 4.5) recognized more information than
AD patients (p < .05) but less than healthy older adults (p < .05). An
effect of the kind of recognitions appeared (F(2, 92) = 3.2, p < .05,
2p = .06), showing that recognition of the elements (mean of 5.3)
was better accomplished than recognition of temporal (mean of
3.8) and spatial information (mean of 4.1) (p < .001). No interaction
between populations and the kind of recognitions was observed
and no effect of exploration appeared.
3.2.2.2. Recollection and familiarity judgments. Similarly, we conducted an ANCOVA on recollection judgments with population as a
between-subjects factor and kind of recognitions and exploration
as within-subjects factors. An effect of population was observed on
recollection (R) judgments in AD patients (mean of .61) (p < .001),
and aMCI patients (mean of .32) (p < .01) had fewer recollections
than the older participants (mean of .25) (F(2, 46) = 18.5, p < .001,
2p = .44). No effect of the kind of recognition or of exploration
was observed on the quantity of R judgments. Regarding the
ANCOVA on familiarity (F) judgments, no signicant differences
between groups were found, F < 1. However, an effect of the kind of
recognitions was observed (F(2, 92) = 3.1, p < .05, 2p = .06), but no
signicant differences appeared using post hoc tests.
In summary, our results (Table 2) revealed that AD patients
were impaired compared to the healthy older group in all recalls,
which includes central information, its multi-component context,
and binding. This pattern was observed in immediate and delayed
recalls for both recognition and recollection judgments. Familiarity judgment was the only score preserved in Alzheimers disease
pathology. In addition, aMCI patients impairment was limited
compared to the global impairment observed in AD patients, and
their performance on temporal recall and immediate binding was
similar to healthy controls. Their recognition of elements, temporal
and spatial information was impaired as well as their recollection
judgments, but not their familiarity judgments. Further, all participants beneted from active exploration, including patients, which
improved the recall of elements, spatial allocentric information

598

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

Table 2
Synthesis of signicant results (effects of population, exploration and recall) observing in the VR episodic memory assessment.
Effects

Population

Exploration

What
Details
When

AD < aMCI < OA

AD < aMCI < OA
AD < aMCI; OA; OA = aMCI

All groups benet of active exploration

All groups benet of passive exploration

Recall

Immediate recall > delayed recall for aMCI and AD

OA and aMCI benet of active exploration at immediate recall

Where egocentric
Where allocentric
Binding
Recognition

AD < aMCI < OA

AD < aMCI < OA
AD < aMCI; OA; OA = aMCI
AD < aMCI < OA
What > when and where

All groups benet of active exploration

All groups benet of active exploration

aMCI = OA at immediate but not at delayed recall

AD: Alzheimer disease; aMCI: amnesic mild cognitive impairment; OA: older adult.

and binding. However, although there was no interaction between

exploration and population, the effect of action was observable
mainly for aMCI patients. In addition, the healthy and aMCI groups
beneted from active exploration for their immediate recall of temporal information. Finally, recall of details improved after passive
exploration compared to active exploration. This was observed
regardless of the group but the descriptive results indicated it was
essentially true for healthy older participants.
3.2.3. Correlations between episodic scores and subjective daily
complaint
We conducted analyses of BravaisPearson correlations to
determine the features that a daily complaint scale such as the
CDS, shares with both of our VR and classical verbal memory tests
(Grober and Buschke) (Table 3). These correlations were computed
separately for each group and were also controlled for the level of
education and gender. All scores used to calculate correlations were
raw scores. For each episodic memory component of the VR test, we
took an average score of all conditions (passive, active, immediate,
and delayed). We observed that many VR scores in the AD group
(what, when and egocentric where recalls, binding, and where recognitions) correlated negatively with the CDS. We found that the more
complaints that the patients had about their cognitive experiences
in daily life, the more impaired they were on the VR test. Only one
correlation was observed in the aMCI group. Moreover, when we
looked at the correlations of the CDS score with each VR condition
(active, passive, immediate, and delayed, for example, recall of what
in immediate recall after passive exploration), we observed that
aMCI patients VR scores in the passive condition were negatively
correlated with the CDS (recall of detail and of central and spatial
information). Regarding the AD patients and the older adult group,
the negative correlations appeared in all conditions (active, passive
and immediate, and delayed recalls). Moreover, we observed that
scores on the Grober and Buschke test were not correlated with the
CDS except for the healthy group, where recognition of the verbal
test was negatively correlated with the CDS.
4. Discussion
We will rst discuss the ndings from our episodic memory
test in both pathological populations. Second, we will discuss some potential explanations for the enhanced retrieval of

information following active exploration. Third, we will discuss

what the sensitivity of our VR test implies for future neuropsychological applications.
4.1. Differentiating dementia from normal aging using a VR
assessment
After having demonstrated the tests sensitivity to benchmark
episodic memory measures in healthy younger and older groups
(Plancher et al., 2010), our VR test seems to be well-adapted to
characterize amnesic differences between normal and pathological
aging.
Our study conrms, with an ecological test of memory, the
presence of a profound genuine decit of episodic memory in AD
involving encoding, consolidation and retrieval processes (Chtelat,
Desgranges, & Eustache, 2006; Greene, Baddeley, & Hodges, 1996;
Mayes, 1988; Perri et al., 2007; Plancher, Guyard, Nicolas, & Piolino,
2009; Putzke et al., 2000). In addition, as classically observed
with verbal memory tests, aMCI patients were impaired compared
with the control group at recalling central information (Belleville,
Sylvain-Roy, de Boysson, & Mnard, 2008; Chtelat et al., 2006;
Dubois & Albert, 2004; Perri et al., 2005) but performed better on
recalling central information than AD patients.
Concerning their spatial memory scores, the aMCI and AD
patients were impaired in comparison with the control population
for both the immediate and delayed recall of egocentric memory
(memory for the position of elements regarding the position of our
body) and allocentric spatial information (memory for spatial relationships between objects independent of the position of the body).
Spatial allocentric memory assessments were found to be particularly useful for distinguishing aMCI patients from healthy older
adults (Hort et al., 2007). Thus, this provides a potential diagnostic cue for aMCI pathology. This cognitive ability is considered to
depend mainly on the hippocampal areas, which suffer more damage in aMCI than in healthy aging (e.g., Wolf et al., 2003). More
specically, a virtual reality study showed that reduced volume
of the hippocampus, the precuneus and parietal regions predicts,
respectively, the allocentric and egocentric memory impairment of
patients with aMCI (Weniger, Ruhlede, Lange, Wolf, & Irle, 2011).
Moreover, aMCI patients who converted to dementia 6 years later
showed initially reduced hippocampal volume and impaired allocentric memory compared with aMCI patients who did not convert

Table 3
Intercorrelations between the VR test and CDS and between the Grober and Buchke memory test and CDS. Only signicant correlations are presented (p < .05).
Recall
What
CDS
OA
aMCI
AD

Recognition
Details

When

Ego where

Allo where

Binding

.50
.71

.81

What

Grober and Buchke

When

Where

.52
.82

.81

Recall

Recognition
.59

.67
.56

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

to dementia, whereas reduced parietal volume was not associated

with the probability of conversion (Weniger et al., 2011). These
results suggest that allocentric assessment is a better cue for early
prediction of AD. However, different allocentric tests are required
to consolidate knowledge of the allocentric impairment demonstrated in aMCI, for example, a mapping recall where participants
were asked to draw a map of the environment or recognition of
snapshots from a novel perspective compared with the encoding
perspective.
In addition, the retrieval quality of some contextual information
was preserved in aMCI when temporal information was requested
at immediate recall. To our knowledge, this is the rst time that
temporal memory has been assessed in aMCI patients. Recall of
temporal memory was evaluated in this study by asking participants to specify whether the item occurred at the beginning, the
middle, or at the end of the town. A recent study demonstrated
that differences in the precision of the temporal demands of a task
led to the activation of different neuroanatomical regions (Jenkins
& Ranganath, 2010). On the nest-grained temporal memory test
(recalling the order in which stimuli were originally presented),
accuracy was predicted by activity in the parahippocampal cortex,
whereas on a coarse temporal test (which approximately indicates when during the course of the experiment the stimulus
appeared), accuracy was predicted by activity in several regions
of the prefrontal cortex and in the hippocampus. Our VR temporal assessment is similar to a coarse temporal test, where aMCI
patients retain some capacities that depend on relatively preserved
frontal regions. In future studies, we will need to rene our temporal assessment by asking a precise sequence of elements to conrm
aMCI preservation at immediate recall. The recall of sequential
events is supposed to be strongly related to hippocampal function (Igli, Zaoui, Berthoz, & Rondi-Reig, 2009) and would thus be
sensitive to aMCI.
Overall, we observed that the quality of binding depended on
the delay in aMCI. Unlike after a delay, the patients ability to
bind central and contextual information immediately after immersion was similar to controls. Taken together, this nding suggests
that all contextual recalls were preserved at immediate recall in
aMCI, indicating problems with memory consolidation. However,
their lower performance on recognition (especially on recollection)
also indicated some limitations with encoding. Indeed, the aMCI
and AD patients demonstrated impaired encoding on recognition
tests, which is consistent with results obtained in previous studies
(Clment, Belleville, & Mellah, 2010; Lenzi et al., 2011). Accordingly,
these studies found reduced activation in the medial temporal
structures during encoding compared with healthy participants.
However, increase of brain activation was also observed in frontal
regions for aMCI patients during recognition tasks (Clment et al.,
2010; Lenzi et al., 2011), suggesting compensatory mechanisms
aiding to reduce their cognitive decits compared to AD. In future
studies, improved recognition scores should distinguish correct hits
from rejections, and omissions from false alarms. Interestingly, we
conrmed, using a VR test, the preservation of familiarity, but not of
recollection in aMCI and AD patients (Serra et al., 2010; Westerberg,
Paller, Holdstock, Mayes, & Reber, 2006), supporting the idea that
recollection and familiarity are associated with distinct anatomical
substrates. Indeed, damage to the hippocampus is known to disrupt recollection but to spare familiarity (Aggleton & Brown, 1999;
Yonelinas, 2002; Yonelinas, Aly, Wang, & Koen, 2010; Yonelinas &
Jacoby, 1995).
4.2. Environmental support: active vs. passive exploration
Previous studies have shown that active exploration of a virtual environment enhances recall of spatial information in young
participants (e.g., Brooks, Attree, Rose, Clifford, & Leadbetter, 1999;

599

Carassa, Geminiani, Morganti, & Varotto, 2002). Our data indicated

that all participants, including aMCI and AD patients, were more
likely to retrieve central and allocentric spatial information and
binding after active exploration of the environment.
Why did this enhancement appear even as all memory traces
seem to be in rapid decline? This type of encoding based on procedural skills has been documented to be one of the most preserved in
AD patients (Lipinska & Backman, 1997; Van Halteren-van Tilborg
et al., 2007). While explicit or declarative learning methods are the
starting point of most rehabilitation programs for dementia sufferers, AD patients have been shown to benet more from implicit
or procedural learning methods (Rsler et al., 2002; Zanetti et al.,
2001). For example, patients were able to re-learn to dance or to
use a telephone through an implicit approach. This implicit learning
occurs without awareness, often simply through repeated exposure, and can be unconsciously revived (Buchner & Wippich, 1998).
In our VR study, encoding of spatial information through driving
appears to enhance spatial retrieval of the participants. The memory benets of action accrue information that requires action during
encoding (i.e., spatial memory), which is in accordance with the
theory of Tulving and Thomson (1973). Explications of the benecial effect of action in patients might be provided by researches
showing that procedural memory, allowing automatic execution
of skills, can operate independently of hippocampal structures
recruiting other structures as the striatum (Pennartz et al., 2011;
Squire, 2004). In line with this hypothesis, some studies reported
that the striatum and the cerebellum, engaged during the execution of well established procedure (e.g., driving), are relatively
well preserved in early to moderate AD (Perani et al., 1993). In
addition, it has been observed that retrieval of prototypic action
(e.g., brush your teeth) recruits the motor cortex (Nilsson et al.,
2000; Nyberg et al., 2001) that is also preserved in AD (Chase
et al., 1984). Consequently, it could be suggested that patients can
enhance their memory performance when memory recruits well
preserved motor brain areas. This question could be further investigated using our VR test by studying patients with motor symptoms
as observed in Parkinsons disease. However, we also found benets
of active exploration for the participants recall of central information (i.e., not spatial information but rather a declarative one) and
the relationship between pieces of information (binding). If action
enhances binding, it seems logical that the central and contextual
information should have increased together. We can assume that
after an active exploration, memory traces of the spatial environment are enhanced and as a result, all related information is more
likely to be retrieved. Other theories have also suggested that the
hippocampus is the structure that underpins the establishment of
links between different pieces of information (Atienza et al., 2011;
Eichenbaum, 2000). However, enhancement of central memory
may be through another process, which may underlie this action
effect.
Indeed, it seems clear that actively exploring an environment
does not elicit the same cognitive procedures as viewing that environment as a video. During active exploration, patients navigate
themselves. In line with the generation effect observed with words
(Rabinowitz & Craik, 1986); a word is better recalled when it was
generated with encoding rather than through reading. We suggest that the central and spatial information encountered in the
course of navigation is subsequently better remembered because
its perception was generated by the actual actions of the participants. In addition, because participants are more engaged in the
task and/or may be more attentive, this should enable them to
encode information more strongly. However, the results provide
hints that it is probably not the attention per se that enables better
encoding in this situation. In fact we saw in healthy participants
that recall of details was improved after passive exploration (see
Fig. 2). Other previous studies (Plancher et al., 2008) have shown

600

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

that driving in a virtual environment while encoding information

may be considered as a double task. In an active condition, participants do not have enough resources to pay attention to every
detail of the town. Nevertheless, another cognitive mechanism that
may explain active enhancement is self-referential processing.
Although self-referential processing is a prerequisite for episodic
memory retrieval (Tulving, 2002), it tends to be more pronounced
in VR studies using active exploration than in typical laboratory
conditions. Self-referential processing has been typically associated
with the medial prefrontal cortex (Craik et al., 1999; Kelley et al.,
2002; Martinelli et al., in press), an anterior area that is better preserved than more posterior areas in early dementia (Grady et al.,
2003). Accordingly, some studies have reported a benet in normal
aging and in early AD patients who self-reference both encoding
and retrieval of episodic memories (Gutchess et al., 2010; Lalanne,
Grolleau, & Piolino, 2010; Ruby et al., 2009).
In total, we suggest that the benet of the active condition
would mainly rely on the implication of procedural skills and selfinvolvement during encoding. However, it is important to point out
that although the interactions between exploration and population
were not signicant, some descriptive results (see Figs. 2 and 3)
indicated that the benet of the active condition was essentially
true for aMCI patients. In summary, even if these rst results are
promising, future research should investigate in more detail this
issue in aMCI and early to moderate AD. The extent of the damage
of the neocortex in the severe-stage of AD is expected to lessen the
positive impact of action on episodic memory.
4.3. Sensitivity of our VR test to daily memory complaints
This study also aimed to benchmark our virtual test against a
classical verbal test. Although both the Grober and Buschke test and
our VR test revealed episodic memory differences between normal
and pathological groups as well as between the two pathological
groups, their relationship to CDS scores was starkly different. Interestingly, on the VR-test, the AD patients performance on the recall
of elements, temporal and spatial information, binding and spatial recognition correlated with their CDS scores. Conversely, their
scores on the verbal test entirely failed to correlate with the CDS.
In addition, when we look into detailed correlations using separate
scores for each condition, we observed that the aMCI patients VR
scores in the passive condition also negatively correlated with the
CDS (recall of detail and of central and spatial information). These
results have extended what we observed in a previous VR study
with a group of normal aging subjects (Plancher et al., 2010), and
were re-conrmed in the present study. We offer evidence that our
new VR test can also be used to evaluate cognitive functions in a
way that is relevant to the patients subjective decits in pathological aging. This difference suggests that our VR test can be used to
test memory in a way that is more closely related to the memory
requirements of daily life compared with the classical verbal test.
Overall, our results are in accordance with the notion that virtual reality is better adapted for early diagnosis and rehabilitation
of patients with suspected AD than current traditional (verbal)
memory tools (Farias et al., 2003; Schultheis et al., 2002). Our
results conrm that the impairment of allocentric spatial memory in particular, typically reecting hippocampal damage, can be
used as a diagnostic cue for aMCI. In addition, active involvement
enhanced the patients episodic memory performance, including
those affected by AD, potentially reecting the relative preservation of some automatic processes related to motor and frontal
areas, suggesting that future studies should focus on the role of
active encoding in episodic memory to better understand the cognitive and neural processes that underlie this enhancement. From
a clinical point of view, it will be important to incorporate active
encoding into episodic memory training, and to encourage patients

to use their own actions to encode information in their daily life.

Overall, VR studies are likely to create more multi-sensorial and
self-relevant situations than typical laboratory conditions. VR can
be used to build or reproduce an innite variety of environments,
such as the patients place of living, creating highly ergonomic conditions for rehabilitation. Overall, our study clearly demonstrates
the feasibility of using VR technology to study the episodic memory
decits of patients with aMCI and AD.
Acknowledgments
The authors would like to thank all the volunteers and patients,
the neuropsychologists V. Lamidey and E. Lima Leite, Dr F.
Mahieux and the clinical staff involved, E. Orriols for the virtual
environments and P. Reeve and especially language editor of Neuropsychologia for English and style corrections.
References
Aggleton, J. P., & Brown, M. (1999). Episodic memory, amnesia, and the hippocampalanterior thalamic axis. Behavioural Brain Science, 22, 425489.
Atienza, M., Atalaia-Silva, K. C., Gonzalez-Escamilla, G., Gil-Neciga, E., SuarezGonzalez, A., & Cantero, J. L. (2011). Associative memory decits in mild cognitive
impairment: The role of hippocampal formation. NeuroImage, 57, 13311342.
Belleville, S., Sylvain-Roy, S., de Boysson, C., & Mnard, M. C. (2008). Characterizing
the memory changes in persons with mild cognitive impairment. Progress in
Brain Research, 169, 365375.
Brooks, B. M., Attree, E. A., Rose, F. D., Clifford, B. R., & Leadbetter, A. G. (1999). The
specicity of memory enhancement during interaction with a virtual environment. Memory, 7, 6578.
Buchner, A., & Wippich, W. (1998). Differences and commonalities between implicit
learning and implicit memory. In M. A. Stadler, & P. A. Frensch (Eds.), Handbook
of implicit learning (pp. 346). Thousand Oaks, CA: Sage Publications.
Burgess, N., Trinkler, I., King, J., Kennedy, A., & Cipolotti, L. (2006). Impaired allocentric spatial memory underlying topographical disorientation. Reviews in the
Neurosciences, 17, 239251.
Carassa, A., Geminiani, G., Morganti, F., & Varotto, D. (2002). Active and passive spatial learning in a complex virtual environment: The effect of efcient exploration.
Cognitive Processing, 34, 6581.
Chase, T. N., Foster, N. L., Fedio, P., Brooks, R., Mansi, L., & Di Chiro, G. (1984). Regional
cortical dysfunction in Alzheimers disease as determined by positron emission
tomography. Annals of Neurology, 15, 170174.
Clment, F., Belleville, S., & Mellah, S. (2010). Functional neuroanatomy of the
encoding and retrieval processes of verbal episodic memory in MCI. Cortex, 46,
10051015.
Chtelat, G., Desgranges, B., de la Sayette, V., Viader, F., Eustache, F., & Baron, J.
C. (2002). Mapping gray matter loss with voxel-based morphometry in mild
cognitive impairment. Neuroreport, 13, 19391943.
Chtelat, G., Desgranges, B., & Eustache, F. (2006). Brain prole of hypometabolism
in early Alzheimers disease: Relationships with cognitive decits and atrophy.
Revue de Neurologie, 162, 945951.
Chtelat, G., Villain, N., Desgranges, B., Eustache, F., & Baron, J. C. (2009). Posterior cingulate hypometabolism in early Alzheimers disease: What is the contribution
of local atrophy versus disconnection? Brain, 132, e133.
Clment, J. P., Nassif, R., Lger, J. M., & Marchan, F. (1997). Mise au point et contribution la validation dune version francaise brve de la Geriatric depression
scale de Yesavage. LEncphale, 23, 9199.
Cohen, G. (1989). Memory in the real world. Hillsdall: Erlbaum.
Cohen, R. L. (1981). On the generality of some memory laws. Scandinavian Journal of
Psychology, 22, 267281.
Craik, F. I. M. (1983). On the transfer of information from temporary to permanent
memory. Philosophical Transactions of the Royal Society of London, 302, 341359.
Craik, F. I. M. (1986). A functional account of age differences in memory. In F. Klix, &
H. Hagendorf (Eds.), Human memory and cognitive capabilities, mechanisms and
performance (pp. 409422). New York: Elsevier.
Craik, F. I. M., Moroz, T. M., Moscovitch, M., Stuss, D. T., Winocur, G., Tulving, E., et al.
(1999). In search of the self: a positron emission tomography study. Psychological
Science, 10, 2634.
(2008). Detecting navigational decits in cognitive aging and Alzheimer disease
using virtual reality. Neurology, 71, 888895.
DeIpolyi, R., Rankin, K. P., Mucke, L., Miller, B. L., & Gorno-Tempini, M. L. (2007).
Spatial cognition and the human navigation network in AD and MCI. Neurology,
69, 986997.
De Leon, M., DeSanti, S., Zinkowski, R., Mehta, P., Pratico, D., Segal, S., et al. (2006).
Longitudinal CSF and MRI biomarkers improve the diagnosis of mild cognitive
impairment. Neurobiology of Aging, 27, 394401.
Deltour, J. J. (1993). (Mill Hill vocabulary scale of J. C. Raven. French adaptation and
comparative norms of the Mill Hill and of Standard progressive matrices (PM 38))
Echelle de vocabulaire de Mill Hill de J. C. Raven: Adaptation francaise et normes

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

compares du Mill Hill et du Standard progressive matrices (PM 38). In Manuel
Editions lApplication des Techniques Modernes, Braine-le-Chteau.
Deweer, B., Ergis, A. M., Fossati, P., Pillon, B., Boller, F., Agid, Y., et al. (1994). Explicit
memory, procedural learning and lexical priming in Alzheimers disease. Cortex,
30, 113126.
Deweer, B., Pillon, B., Michon, A., & Dubois, B. (1993). Mirror reading in Alzheimers
disease: Normal skill learning and acquisition of item-specic information. Journal of Clinical Experimental Neuropsychology, 15, 789804.
Dickerson, B. C., Goncharova, I., Sullivan, M. P., Forchetti, C., Wilson, R. S., Bennett,
D. A., et al. (2001). MRIderived entorhinal and hippocampal atrophy in incipient
and very mild Alzheimers disease. Neurobiology of Aging, 22, 747754.
Drzezga, A., Grimmer, T., Peller, M., Wermke, M., Siebner, H., Rauschecker, J. P., et al.
(2005). Impaired cross-modal inhibition in Alzheimer disease. PLoS Medicine, 2,
e288.
Dubois, B., & Albert, M. L. (2004). Amnestic MCI or prodromal Alzheimers disease?
Lancet Neurology, 3, 246248.
Eichenbaum, H. (2000). A cortical-hippocampal system for declarative memory.
Nature Review Neuroscience, 1, 4150.
Evans, M. C., Barnes, J., Nielsen, C., Kim, L. G., Clegg, S. L., Blair, M., et al. (2010). Volume changes in Alzheimers disease and mild cognitive impairment: Cognitive
associations. European Radiology, 20, 674682.
Farias, S. T., Harrell, E., Neumann, C., & Houtz, A. (2003). The relationship between
neuropsychological performance and daily functioning in individuals with
Alzheimers disease: Ecological validity of neuropsychological tests. Archives of
Clinical Neuropsychology, 18, 655672.
Fischer, P., Jungwirth, S., Zehetmayer, S., Weissgram, S., Hoenigschnabl, S., Gelpi,
E., et al. (2007). Conversion from subtypes of mild cognitive impairment to
Alzheimer dementia. Neurology, 23, 288291.
Folstein, M. F., Folstein, S., & McHugh, P. R. (1975). Mini Mental State: A practical
method for grading the cognitive state of patients for the clinician. Journal of
Psychiatric Research, 121, 189198.
Fowler, K. S., Saling, M. M., Conway, E. L., Semple, J. M., & Louis, W. J. (2002). Paired
associate performance in the early detection of DAT. Journal of International
Neuropsychology Society, 8, 5871.
Fox, N., & Schott, J. (2004). Imaging cerebral atrophy: Normal ageing to Alzheimers
disease. Lancet, 363, 392394.
Gabrieli, J. D., Corkin, S., Mickel, S. F., & Growdon, J. H. (1993). Intact acquisition and
long-term retention of mirror-tracing skill in Alzheimers disease and in global
amnesia. Behavioral Neuroscience, 107, 899910.
Gardiner, J. M. (1988). Functional aspects of recollective experience. Memory and
Cognition, 16, 309313.
Gardiner, J. M. (2001). Episodic memory and autonoetic consciousness: A rstperson approach. Philosophical Transactions of the Royal Society B, 356,
13511361.
Gerardin, E., Chtelat, G., Chupin, M., Cuingnet, R., Desgranges, B., Kim, H. S., et al.
(2009). Multidimensional classication of hippocampal shape features discriminates Alzheimers disease and mild cognitive impairment from normal aging.
NeuroImage, 47, 14761486.
Grady, C. L., McIntosh, A. R., Beig, S., Keightley, M. L., Burian, H., & Black, S. E. (2003).
Evidence from functional neuroimaging of a compensatory prefrontal network
in Alzheimers disease. The Journal of Neuroscience, 23, 986993.
Greene, J. D. W., Baddeley, A. D., & Hodges, J. R. (1996). Analysis of the episodic
memory decit in early Alzheimers disease: Evidence from the doors and people
test. Neuropsychologia, 34, 537551.
Grober, E., & Buschke, H. (1987). Genuine memory decits in dementia. Developmental Neuropsychology, 3, 1336.
Gutchess, A. H., Kensinger, E. A., & Schacter, D. L. (2010). Functional neuroimaging
of self-referential encoding with age. Neuropsychologia, 48, 211219.
Hirono, N., Mori, E., Ikejiri, Y., Imamura, T., Shimomura, T., Ikeda, M., et al. (1997).
Procedural memory in patients with mild Alzheimers disease. Dementia and
Geriatric Cognitive Disorders, 8, 210216.
Hodges, J. R. (2006). Alzheimers centennial legacy: Origins, landmarks and the current status of knowledge concerning cognitive aspects. Brain, 129, 28112822.
Hort, J., Lacz, J., Vyhnalek, M., Bojar, M., Bures, J., & Vlcek, K. (2007). Spatial navigation decit in amnestic mild cognitive impairment. Proceedings of the National
Academy of Sciences of the United States of America, 104, 40424047.
Iachini, T., Iavarone, A., Senese, V. P., Ruotolo, F., & Ruggiero, G. (2009). Visuospatial memory in healthy elderly, AD and MCI: A review. Current Aging Science, 2,
4359.
Igli, K., Zaoui, M., Berthoz, A., & Rondi-Reig, L. (2009). Sequential egocentric strategy is acquired as early as allocentric strategy: Parallel acquisition of these two
navigation strategies. Hippocampus, 19, 11991211.
Jenkins, L. J., & Ranganath, C. (2010). Prefrontal and medial temporal lobe activity at encoding predicts temporal context memory. Journal of Neuroscience, 17,
1555815565.
Karlsson, T., Backman l. Herlitz, A., Nilsson l., G., Winblad, B., & Osterlind, P. O. (1989).
Memory improvement at different stages of Alzheimers disease. Neuropsychologia, 27, 737742.
Kelley, W. M., Macrae, C. N., Wyland, C., Caglar, S., Inati, S., & Heatherton, T. F. (2002).
Finding the self? An event-related fMRI study. Journal of Cognitive Neuroscience,
14, 785794.
Kessels, R., Hobbel, D., & Postma, A. (2007). Aging, context memory and binding: A
comparison of what, where and when in young and older adults. International
Journal of Neurosciences, 117, 795810.

601

Lalanne, J., Grolleau, P., & Piolino, P. (2010). Self-reference effect and episodic memory in normal aging and Alzheimers disease: Myth or reality? Psychologie &
Neuropsychiatrie du Vieillissement, 8, 277294.
Lacz, J., Vlcek, K., Vyhnlek, M., Vajnerov, O., Ort, M., Holmerov, I., et al. (2009).
Spatial navigation testing discriminates two types of amnestic mild cognitive
impairment. Behavioural Brain Research, 202, 252259.
Lekeu, F., Van der Linden, M., Moonen, G., & Salmon, E. (2002). Exploring the effect
of action familiarity on SPTs recall performance in Alzheimers disease. Journal
of Clinical Experimental Neuropsychology, 24, 10571069.
Lenzi, D., Serra, L., Perri, R., Pantano, P., Lenzi, G. L., Paulesu, E., et al. (2011). Single domain amnestic MCI: A multiple cognitive domains fMRI investigation.
Neurobiology of Aging, 32, 15421557.
Lezak, M. D., Howieson, D. B., & Loring, D. W. (2004). Neuropsychological assessment
(4th ed.). Oxford, England: Oxford University Press.
Lipinska, B., & Backman, L. (1997). Encoding-retrieval interactions in mild
Alzheimers disease: The role of access to categorical information. Brain and
Cognition, 34, 274286.
Luo, L., & Craik, F. I. M. (2008). Aging and memory: A cognitive approach. Canadian
Journal of Psychiatry, 53, 346353.
Mntyl, T. (1993). Knowing but not remembering: Adult age differences in recollective experience. Memory & Cognition, 21, 379388.
Martinelli, P., Sperduti, M., & Piolino, P. (in press). Neural substrates of the
self-memory system: New insights from a meta-analysis. Human Brain
Mapping, doi:10.1002/hbm.22008.
Mayes, A. R. (1988). Human organic memory disorders. Cambridge: Cambridge University Press.
McKahnn, G., Drachman, D., Folstein, M., Katzman, R., Price, D., & Stadlan, E. M.
(1984). Clinical diagnosis of Alzheimers disease: Report of the NINCDS-ADRDA
work group under the auspices of the Department of Health and Human Services
Task Force on Alzheimers disease. Neurology, 34, 934939.
McNair, D., & Kahn, R. (1983). Self-assessment of cognitive decits. In T. Crook, S. Ferris, & R. Bartus (Eds.), Assessment in geriatric psychopharmacology (pp. 137143).
New Canaan, CT: Powley.
Naveh-Benjamin, M., Craik, F. I. M., & Ben-Shaul, L. (2002). Age-related differences in
cued recall: Effects of support at encoding and retrieval. Aging, Neuropsychology,
and Cognition, 9, 276287.
Nilsson, L. G., Nyberg, L., Klingberg, T., berg, C., Persson, J., & Roland, P. (2000).
Activity in motor areas while remembering action events. NeuroReport, 11,
21992201.
Nyberg, L., Petersson, K. M., Nilsson, L. G., Sandblom, J., Aberg, C., & Ingvar, M.
(2001). Reactivation of motor brain areas during explicit memory for actions.
NeuroImage, 14, 521528.
Park, D. C., Puglisi, J. T., & Smith, A. D. (1986). Memory for pictures: Does an agerelated decline exist? Psychology and Aging, 1, 1117.
Park, D. C., Smith, A. D., Morell, R. W., Puglisi, J. T., & Dudley, W. N. (1990). Effects of
contextual integration on recall of pictures by older adults. Journal of Gerontology: Psychological Sciences, 45, 5257.
Pennartz, C. M., Ito, R., Verschure, P. F., Battaglia, F. P., & Robbins, T. W. (2011). The
hippocampal-striatal axis in learning, prediction and goal-directed behavior.
Trends in Neurosciences, 34, 548559.
Perani, D., Bressi, S., Cappa, S. F., Vallar, G., Alberoni, M., Grassi, F., et al. (1993).
Evidence of multiple memory systems in the human brain. A [18F] FDG PET
metabolic study. Brain, 116, 903919.
Perri, R., Carlesimo, G. A., Serra, L., & Caltagirone, C. (2005). Early diagnosis group of
the Italian interdisciplinary network on Alzheimers disease. Characterisation of
memory prole in subjects with mild cognitive impairment. Journal of Clinical
and Experimental Neuropsychology, 27, 10331055.
Perri, R., Serra, L., Carlesimo, G. A., & Caltagirone, C. (2007). Italian interdisciplinary
network on Alzheimers disease, early diagnosis group. Neuropsychology, 21,
549558.
Petersen, R. C., Doody, R., Kurz, A., Mohs, R. C., Morris, J. C., Rabins, P. V., et al.
(2001). Current concepts in mild cognitive impairment. Archives of Neurology,
58, 19851992.
Petersen, R. C., Smith, G. E., Waring, S. C., Ivnik, R. J., Tangalos, E. G., & Kokmen,
E. (1999). Mild cognitive impairment: Clinical characterization and outcome.
Archives of Neurology, 56, 303308.
Piolino, P., Desgranges, B., & Eustache, F. (2009). Episodic autobiographical memory over the course of time: Cognitive, neuropsychological and neuroimaging
ndings. Neuropsychologia, 47, 23142329.
Plancher, G., Guyard, A., Nicolas, S., & Piolino, P. (2009). Mechanisms underlying
the production of false memories for famous peoples names in aging and
Alzheimers disease. Neuropsychologia, 47, 25272536.
Plancher, G., Gyselinck, V., Nicolas, S., & Piolino, P. (2010). Age effect on components
of episodic memory and feature binding: A virtual reality study. Neuropsychology, 24, 379390.
Plancher, G., Nicolas, S., & Piolino, P. (2008). Contribution of virtual reality for neuropsychology of memory: Study in ageing. Psychologie et NeuroPsychiatrie du
vieillissement, 6, 722.
Putzke, J. D., Rickert, E. J., Duke, L. W., Marson, D., & Harrell, L. (2000). Differential automatic processing decits in early stage Alzheimers disease. Aging,
Neuropsychology and Cognition, 7, 112118.
Rabinowitz, J. C., & Craik, F. I. M. (1986). Specic enhancement effects
associated with word generation. Journal of Memory and Language, 25,
226237.

602

G. Plancher et al. / Neuropsychologia 50 (2012) 592602

Risacher, S. L., Saykin, A. J., West, J. D., Shen, L., Firpi, H. A., & McDonald, B. C. (2009).
Baseline MRI predictors of conversion from MCI to probable AD in the ADNI
cohort. Current Alzheimer Research, 6, 347361.
Rsler, A., Seifritz, E., Kruchi, K., Spoerl, D., Brokuslaus, I., Proserpi, S-M., et al.
(2002). Skill learning in patients with moderate Alzheimers disease: A prospective pilot-study of waltz-lessons. International Journal of Geriatric Psychiatry, 17,
11551156.
Ruby, P., Collette, F., DArgembeau, A. D., Pters, F., Degueldre, C., Balteau, E.,
et al. (2009). Perspective taking to assess self-personality: Whats modied in
Alzheimers disease? Neurobiology of Aging, 30, 16371651.
Salmon, D. P., & Bondi, M. W. (2009). Neuropsychological assessment of dementia.
Annual Review of Psychology, 60, 257282.
Schacter, D. L., Israel, L., & Racine, C. (1999). Suppressing false recognition in younger
and older adults: The distinctiveness heuristic. Journal of Memory and Language,
40, 124.
Schultheis, M. T., Himelstein, J., & Rizzo, A. A. (2002). Virtual reality and neuropsychology: Upgrading the current tools. Journal of Head Trauma Rehabilitation, 17,
379394.
Serra, L., Bozzali, M., Cercignani, M., Perri, R., Fadda, L., Caltagirone, C., et al. (2010).
Recollection and familiarity in amnesic mild cognitive impairment. Neuropsychology, 24, 316326.
Slotnick, S. D. (2010). Does the hippocampus mediate objective binding or subjective
remembering? NeuroImage, 49, 17691776.
Spaan, P. E., Raaijmakers, J. G., & Jonker, C. (2003). Alzheimers disease versus normal ageing: A review of the efciency of clinical and experimental memory
measures. Journal of Clinical Experimental Neuropsychology, 25, 216233.
Squire, L. R. (2004). Memory systems of the brain: A brief history and current perspective. Neurobiology of Learning and Memory, 82, 171177.
Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18, 643662.
Swainson, R., Hodges, J. R., Galton, C. J., Semple, J., Michael, A., Dunn, B. D., et al. (2001).
Early detection and differential diagnosis of Alzheimers disease and depression
with neuropsychological tasks. Dementia and Geriatric Cognitive Disorders, 12,
265280.
Troyer, A. K., Murphy, K. J., Anderson, N. D., Hayman-Abello, B. A., Craik, F. I., &
Moscovitch, M. (2008). Item and associative memory in amnestic mild cognitive
impairment: Performance on standardized memory tests. Neuropsychology, 22,
1016.
Tulving, E. (1983). Elements of episodic memory. New York: Oxford University Press.
Tulving, E. (1985). Memory and consciousness. Canadian Psychologist, 26, 112.
Tulving, E. (2002). Episodic memory: From mind to brain. Annual Review of Psychology, 53, 125.

Tulving, E., & Thomson, D. M. (1973). Encoding specicity and retrieval processes in
semantic memory. Psychological Review, 80, 352373.
Van Halteren-van Tilborg, I. A., Scherder, E. J., & Hulstijn, W. (2007). Motor-skill
learning in Alzheimers disease: A review with an eye to the clinical practice.
Neuropsychology Review, 3, 203212.
Wechsler, D. (2001). Wechsler Adult Intelligence Scale Third Edition (WAIS III). Paris
(French version): EAP.
Weniger, G., Ruhleder, M., Lange, C., Wolf, S., & Irle, E. (2011). Egocentric and allocentric memory as assessed by virtual reality in individuals with amnestic mild
cognitive impairment. Neuropsychologia, 49, 518527.
Westerberg, C. E., Paller, K. E., Holdstock, J. S., Mayes, A. R., & Reber, P. J. (2006).
When memory does not fail: Familiarity-based recognition in mild cognitive
impairment and Alzheimers disease. Neuropsychology, 20, 193205.
Widmann, C. N., Beinhoff, U., & Riepe, M. W. (2012). Everyday memory decits in
very mild Alzheimers disease. Neurobiology of Aging, 33, 297303.
Wojtasik, V., Olivier, C., Lekeu, F., Quittre, A., Adam, S., & Salmon, E. (2010). A grid
for a precise analysis of daily activities. Neuropsychological Rehabilitation, 20,
120136.
Wolf, H., Jelic, V., Gertz, H. J., Nordberg, A., Julin, P., & Wahlund, L. O. (2003). A critical discussion of the role of neuroimaging in mild cognitive impairment. Acta
Neurologica Scandinavica, 107, 5276.
Yesavage, J. A. (1988). Geriatric depression scale. Psychopharmacology Bulletin, 24,
709711.
Yonelinas, A. P. (2002). The nature of recollection and familiarity: A review of 30
years of research. Journal of Memory and Language, 46, 441517.
Yonelinas, A. P., Aly, M., Wang, W. C., & Koen, J. D. (2010). Recollection and familiarity: Examining controversial assumptions and new directions. Hippocampus,
20, 11781194.
Yonelinas, A. P., & Jacoby, L. L. (1995). The relation between remembering and knowing as bases for recognition: Effects of size congruency. Journal of Memory and
Language, 34, 622643.
Yonelinas, A. P., Kroll, N. E., Dobbins, I., Lazzara, M., & Knight, R. T. (1998). Recollection
and familiarity decits in amnesia: Convergence of remember-know, process
dissociation, and receiver operating characteristic data. Neuropsychology, 12,
323339.
Zakzanis, K. K., Quintin, G., Graham, S. J., & Mraz, R. (2009). Age and dementia related
differences in spatial navigation within an immersive virtual environment. Medical Science Monitor: International Medical Journal of Experimental and Clinical
Research, 15, 140150.
Zanetti, O., Zanieri, G., Di Giovanni, G., De Vreese, L. P., Pezzini, A., Metitieri, T., et al.
(2001). Effectiveness of procedural memory stimulation in mild Alzheimers disease patients: A controlled study. Neuropsychological Rehabilitation, 11, 263272.