REVIEW ARTICLE
REVIEW ARTICLE
Memory suppression in Alzheimer’s disease
Mohamad El Haj1
Received: 5 October 2015 / Accepted: 9 December 2015 / Published online: 23 December 2015
� Springer-Verlag Italia 2015
Abstract An important challenge for memory is the
competition between appropriate and inappropriate infor-
mation during retrieval. This competition is normally
reduced thanks to controlled inhibitory processes that
suppress irrelevant memories. In Alzheimer’s disease
(AD), compromise of suppression ability may result in
strong competition between relevant and irrelevant mem-
ories during retrieval. The present review highlights this
issue by examining studies using the directed forgetting
method in AD. This method in which participants are
typically instructed to forget no longer relevant information
is argued to reflect suppression in memory. Studies using
the directed forgetting method suggest that AD participants
experience difficulties when they are asked to suppress no
longer relevant information in working, autobiographical,
source and destination memory. Difficulties in suppressing
irrelevant information, as may be observed in AD, may
hamper memory retrieval by activating irrelevant memo-
ries at the expense of relevant ones.
Keywords Alzheimer’s disease � Directed forgetting � Inhibition � Memory � Suppression
Suppressing interference in memory
The brain’s ability to store an unlimited amount of infor-
mation creates the problem of remembering the relevant
information from a large pool of interfering memories [1,
2]. This difficulty can be exacerbated by similarity between
the interfered memories since similarity between memories
induces strong competition during retrieval, leading to
difficulties in remembering the relevant memory [3]. For
example, when trying to remember what occurred during
one’s commute back home from work, difficulties with
recall are exacerbated if the journey regularly takes place at
the same time and on the same itinerary. On the other hand,
better recall can be expected if the targeted journey
involved an unusual event such as taking the subway rather
than driving.
An important challenge for memory retrieval is the
competition between appropriate and inappropriate infor-
mation, especially when such information is bound by
common contextual, semantic or emotional features.
Without effective control mechanisms that reduce this
competition, one may fail to remember the targeted episode
as the irrelevant information is likely to be retrieved.
Thanks to inhibitory mechanisms, such failure can be
avoided since these mechanisms are in charge of sup-
pressing competition during memory retrieval [4].
Retrieval inhibition, or the ability to suppress irrelevant
information during recall, is found to be compromised in
aging. According to May and Hasher [5], older adults
experience difficulties in suppressing the activation of
irrelevant thoughts and stimuli. As a result, their memory
becomes overloaded with task-irrelevant information,
which leaves few resources for processing task-relevant
information and induces cognitive decline. This account
has been supported by studies showing that older adults are
& Mohamad El Haj el.haj@hotmail.fr
1 Univ. Lille, CNRS, CHU Lille, UMR 9193 – SCALab –
Sciences Cognitives et Sciences Affectives, Lille F-59000,
France
123
Neurol Sci (2016) 37:337–343
DOI 10.1007/s10072-015-2441-5
prone to interference in verbal working memory, visu-
ospatial working memory, and even implicit memory [5].
As for episodic memory, older adults tend to maintain
information from tasks in the past even when that infor-
mation is no longer relevant for the current task [5]. This
sustained access to prior irrelevant information has been
attributed to reduced inhibitory mechanisms, which should
optimally operate to delete old, irrelevant information. In
this view, compromised inhibitory ability as observed in
aging may disrupt memory retrieval by introducing com-
petition between relevant and irrelevant responses.
Retrieval inhibition has also been found to be compro-
mised in Alzheimer’s disease (AD) and, as this review
suggests, this compromise may result in reduced ability to
suppress irrelevant information during memory retrieval in
AD. To illustrate this issue, the main cognitive character-
istics of AD, namely, memory and inhibitory compromise,
are discussed first.
AD, memory and inhibitory compromise
AD is a neurodegenerative disease characterized by accu-
mulation of b-amyloid peptides in the gray matter of the brain, an accumulation that forms b-amyloid peptide deposits and neurofibrillary tangles composed of hyper-
phosphorylated Tau protein [6]. On a cognitive level, the
main hallmark of AD is memory decline, i.e. an impaired
ability to acquire and remember new episodic information
[6]. This impairment is thought to be mediated by neu-
ropathology in the default mode network, a network that
includes nodes at the medial and lateral prefrontal cortex,
the posteromedial parietal lobe, angular gyrus, as well as
the medial temporal lobes [7]. AD-related memory decline
is mainly observed for episodic memory or the ability to
relive specific personal events that occurred at a particular
time and place. Individuals with AD experience difficulty
in retrieving the context in which a specific episode has
occurred, a difficulty that has been attributed to decline in
source memory or the ability to remember the conditions or
features in which a memory was acquired [8–10] (for a
review, see [11]). Similar to source memory decline,
research shows difficulties in destination memory in AD
[12–16]. Destination memory, a component of the episodic
system, refers to the ability to remember the destination to
remember to whom information has previously been
imparted (e.g. ‘‘did I tell you about the promotion?’’) (for a
review, see [17]). Another feature of episodic memory
decline in AD is impaired ability to relive past events
mentally [18, 19], a decline that has been attributed to
difficulties in retrieving contextual information [12, 20].
Taken together, episodic memory decline in AD hampers
retrieval of specific personal events.
Even though episodic memory decline has been con-
sidered as the main cognitive hallmark of AD, inhibitory
decline is also common in the disease. Studies suggest
that, in comparison to older adults, individuals with AD
systematically demonstrate a large interference effect as
assessed by the Stroop task (for a review, see, [21, 22]).
This effect refers to the increase in response time that
occurs when word meaning and stimulus do not match
(e.g., the word green presented in the color blue) relative
to when they correspond. As with the Stroop effect, AD-
related inhibitory decline can be observed on tasks
requiring controlled suppression, such as the Hayling task
in which participants are required to generate an appro-
priate word as quickly as possible to complete a sentence
[23, 24]. Another feature of AD-related inhibitory decline
is the negative priming effect [21], i.e. the increased
reaction time to respond to a stimulus that had been
recently inhibited. However, unlike their performance on
tasks requiring controlled inhibition, individuals with AD
tend to show some preservation on tasks requiring auto-
matic inhibition, the latter preservation can be observed
on inhibition of return paradigms, in which AD partici-
pants and controls tend to show similar reaction times to
detect a target if it appears in a location where it was
previously shown (for a review, see [21]). Dissociation
between controlled and automatic inhibitory performance
in AD fits with the theoretical model of Harnishfeger [25]
who made a distinction between effortful inhibitory con-
trol, which corresponds to suppression of the information
considered as irrelevant for the ongoing task, and auto-
matic interference from distracting. Besides being
observed in AD, compromise of controlled inhibition has
been observed in individuals with mild cognitive
impairment on the Stroop task [26, 27], the Hayling task
[28], and the Flanker task (a task on which participants
have to respond to a central target flanked by distractors)
[29]. Besides being associated with impaired frontal
activation, inhibitory compromise in mild cognitive
impairment are considered as a risk factor for the devel-
opment of AD [30]. In a related vein, Belanger et al. [27]
suggested that inhibitory compromise, as observed in mild
cognitive impairment, may develop well before cognitive
symptoms reach significant levels in AD.
Altogether, there is evidence to suggest a decline in
controlled inhibition or mechanisms allowing voluntary
suppression of irrelevant information in AD. This decline
can also be extended to controlled inhibition in memory, an
assumption that is supported by findings from studies using
the directed forgetting task.
338 Neurol Sci (2016) 37:337–343
123
AD and suppression in memory, the directed
forgetting task
Resistance to interference in memory can be assessed with
the directed forgetting task. In its conventional configura-
tion, this task requires the processing of two lists of words
(i.e., List 1 ? List 2) [4]. Subjects are typically instructed
to retain the words in List 1, after which they are asked
either to continue remembering or to forget the words in it.
Subsequently, they are instructed to retain the words in List
2. Finally, in a recall test, they are asked to remember all of
the words in both lists, regardless of the intermediary forget
or remember instructions. One main effect of the directed
forgetting method is the observation that participants with
the forget instruction tend to show poorer memory for the
items in List 1 than the remember participants. This effect
has been attributed to retrieval inhibition, whereby the
forget instruction induces a suppression of the List 1 words,
making them less accessible during retrieval [4]. The forget
instruction also reduces the proactive interference of these
words in List 2, thereby improving retrieval of List 2 in the
‘‘forget’’ participants [4]. More precisely, the suppression
effect of the forget instruction reduces the accessibility of
information for retrieval by deactivating the retrieval route
between the irrelevant information and the current cue
[31]. In other words, the suppression effect reflects route
deactivation rather than deletion of the irrelevant infor-
mation, an assumption supported by research showing that
recognition tests can reactivate to-be-forgotten items [4,
32].
A variant of the directed forgetting method is the
‘‘word’’ procedure in which the forget or remember
instruction is provided after the presentation of each word
rather than after List 1. However, both the list and word
procedures are thought to reflect intentional forgetting in
which the accessibility of irrelevant out-of-date informa-
tion is reduced and retrieval of relevant information is
selectively enhanced [33]. Both the list and word methods
have also been used in research assessing ability of
individuals with AD to suppress irrelevant information in
memory. In this area of research, one study administered a
directed forgetting working memory task by exposing AD
participants to two trigrams of three consonants each [24].
A ‘‘to be forgotten’’ indication followed the second tri-
gram, prompting participants to forget the trigram as they
would not be required to recall it later. Immediately after
the presentation of the trigrams, an interpolated activity
took place, consisting of reading strings of numbers aloud.
Afterwards, participants were asked to remember the
three letters of the trigrams, regardless of the ‘‘forget’’
instruction. Results showed difficulties in AD participants
to suppress the to-be-forgotten items from working
memory.
Difficulties to suppress irrelevant information from
memory in AD have also been observed for autobio-
graphical memory. This memory, or memory of personal
experiences and facts about the self, has been found to be
compromised in AD [19, 34–42]. One study assessed
autobiographical suppression by asking AD participants to
generate personal events (i.e., List 1). Then half of the
participants were told that these memories were no longer
relevant to the experiment (i.e., forget participants),
whereas the other half were asked to keep these memories
in mind (i.e., remember participants) [43]. Subsequently,
all the participants were instructed to generate other per-
sonal events (i.e., List 2) and in a subsequent recall test,
they were instructed to recall all the memories regardless of
the previously given forget or remember instructions (i.e.,
List 1 ? List 2). The results showed no effect of the forget
instruction; in other words, similar recall performances for
List 1 were observed in both the remember and the forget
AD participants. The absence of directed forgetting effect
was attributed by the authors to an impairment of autobi-
ographical memory suppression ability in AD.
Another insight into memory suppression in AD comes
from a study assessing directed forgetting in source
memory [44]. In this study, participants were instructed to
remember the source of presentation of items that were
presented by an experimenter with either a black-gloved
hand or a white-gloved hand (i.e., List 1). Afterwards, half
of the participants were asked to forget the source of List 1,
whereas the other half were asked to keep it in mind.
Subsequently, all the participants were asked to retain the
source of presentation of a second List of items (i.e., List 2)
and in a later recall test, they were asked to remember the
source of presentation of the items in List 1 and List 2,
regardless of the forget or remember instructions. These
procedures showed no effect of the forget instruction, as no
differences were observed between the remember and the
forget AD participants on retrieving the sources of List 1
items. Similar findings were observed in a study assessing
directed forgetting in destination memory [45]. In that
study, AD participants were asked to tell proverbs to
celebrities (i.e., List 1); then half of the participants were
asked to forget the destinations whereas the other half were
asked to keep them in mind. After telling other proverbs to
other celebrities (i.e., List 2), participants were asked to
remember the destinations of the List 1 and List 2,
regardless of the forget or remember instructions. These
procedures showed similar destination memory in the for-
get and remember AD participants, i.e. AD participants
showed difficulties in suppressing irrelevant information in
destination memory.
Taken together, empirical evidence suggests that AD
participants experience difficulties when they are asked to
suppress no longer relevant information in working,
Neurol Sci (2016) 37:337–343 339
123
autobiographical, source, and destination memory [24, 43–
45]. The compromise of memory suppression ability in AD
can be better understood by highlighting the neural basis of
this ability.
AD and suppression in memory: neural basis
With regard to memory processing in the directed forget-
ting method, a study observed greater activity in the
anterior cingulate (BA 32), the left inferior prefrontal
cortex (BA 9), and medial superior frontal gyrus (BA 6) for
to-be-remembered than for to-be-forgotten information at
encoding [46]. The same study also observed greater
activity in the left parahippocampal gyrus and right supe-
rior parietal gyrus (BA 7) for remembered than for for-
gotten information. These findings suggest that activity in
the ventral prefrontal and superior frontal regions is asso-
ciated with encoding effort in the directed forgetting
method, whereas the medial temporal and superior parietal
areas are involved in the success of encoding. Mirroring
this assumption, a study found that successful encoding and
retrieval of to-be-remembered information engaged the
hippocampus, the entorhinal cortex, the anterior medial
prefrontal cortex, the left inferior parietal cortex, the pos-
terior cingulate cortex and the precuneus, a network that is
well known to support associative encoding and retrieval
processes in episodic memory. With regard to suppression
processes in the directed forgetting method, a body of lit-
erature supports the role of the right prefrontal cortex in
inhibition of to-be-forgotten information [47, 48].
The neural basis of memory suppression in normal aging
was assessed in a study with the directed forgetting method
[49]. This study showed an increase in the recruitment of
frontal inhibitory control regions, especially the right
superior prefrontal cortex, but high activity in the parietal
cortex during intentional forgetting in healthy older adults.
According to the authors, the parietal activity may reflect
age-related compensation in older adults who show a
reduction in frontally mediated inhibition. Since the pari-
etal lobes are affected by the neuropathology of AD [50],
these inhibitory compensation mechanisms are likely to be
compromised, leaving individuals with AD with poor
suppression ability. Altogether, decline in both the right
prefrontal cortex and parietal cortex may underlie diffi-
culties in suppressing irrelevant information in memory, as
may be observed in AD.
Discussion and conclusion
Successful memory functioning depends on the intentional
ability to encode relevant information and to forget irrel-
evant information. AD-related memory compromise has
been widely attributed to difficulties in encoding and/or
general loss of available memories. The present review
contributes to understanding memory decline in AD by
highlighting difficulties in suppression of irrelevant mem-
ories during retrieval. With age and the passage of time,
memories and personal experiences tend to be more similar
than specific [36]. This similarity is likely to exacerbate
retrieval of a tar
