http://journal.frontiersin.org/article/10.3389/fnagi.2015.00074/full?
Introduction
Dementia is a degeneration of the brain and therefore of
many cognitive processes, including memory. Memory deficits are often
evident before any other signs of dementia are obvious (Masur et al., 1994; Bäckman et al., 2001; Jorm et al., 2005).
Monitoring memory function can therefore be useful for early diagnosis
of dementia, which in turn can help with the management of the disorder,
potentially therapeutically slowing down the progression. For example,
it has been shown that early deficits in episodic memory abilities can
be indicative of the likelihood of a person developing Alzheimer's
disease later on in life (Bäckman et al., 2001).
Episodic memory is our memory for personally experienced episodes from
our own past, which we typically experience as “Mental Time Travel”: a
mentally re-experiencing of the episode in question (Suddendorf and Corballis, 1997).
One of the problems with using cognitive indicators as
potential early-warning signals for dementia is that many cognitive
capacities diminish as we get older. Processing speed, working memory,
and long-term memory are all known to decline steadily as we age,
although aspects of verbal short-term memory (e.g., digit span) and
vocabulary may decline rapidly in later-life (Hedden and Gabrieli, 2004).
With regard to long-term memory, while semantic processes are
relatively unaffected, episodic memory exhibits a much greater degree of
decline (Nyberg et al., 2012).
Numerous studies have shown performance impairments in episodic-like
memory tests in older people, even if there is no evidence of dementia
or Mild Cognitive Impairment (MCI) (Harris et al., 2002). For example, Kessels et al. (2007)
demonstrated broad performance decrements in older adults on a
visuo-spatial episodic memory task which were especially pronounced in
conditions requiring contextual binding. Tasks requiring the learning
and recall of word lists (e.g., Rey-Auditory or California Verbal
Learning Tests; R-AVLT/CVLT) have been found to be impaired in aging (Lundervold et al., 2014), with particular deficits in temporal order indices (Blachstein et al., 2012). There is also some suggestion that the age-related decline in verbal episodic memory may be greater in males than females (Lundervold et al., 2014).
Because of these changes, it is sometimes difficult to distinguish the
early signs of dementia from natural declines in cognitive capacity with
old age. However, it has been suggested that measures such as the
Rey-AVLT may be useful in delineating different dementias (Tierney et al., 1994; Ricci et al., 2012).
One potential criticism of many of the clinical tests of
episodic memory is that they do not have very high ecological validity (Sbordone and Long, 1996).
Everyday episodic memory typically has a number of characteristics that
are not easily captured in most clinical tests: it is made up of
long-term memories for unique events in their spatiotemporal context
(what happened, where it was, when it was). The information is usually
encoded in an incidental manner, and freely recalled, without any cues
relating to the original event (Pause et al., 2013).
Laboratory tests usually match some of these features, but rarely all
of them. For example, some tests, like the R-AVLT, are about free recall
of long-term (30-min) memory (in this case of a list of words), but the
information is just a list of words (no spatiotemporal context needs to
be remembered, although the optional temporal-order trial can be
administered; Vakil and Blachstein, 1994); and it is learned in an intentional manner and rehearsed several times. Other tests (e.g., the Object Relocation task; Kessels et al., 1999)
capture the binding between objects (what happened) and spatial
locations (where it was); they typically do this over short retention
intervals, using recognition processes for the items (though not for the
locations) and again include intentional encoding of the information.
The advantage of all these tests is that the experimenter/clinician
knows exactly which answers are correct and which are wrong, because
they control the information to be retained. When more ecologically
valid measures of episodic memory are used, such as having people freely
recall real events from their own lives, the scoring of these memories
necessarily has to rely on the amount of detail recalled, rather than on
the accuracy of these memories, as no objective record usually exists
of the original event (e.g., Irish et al., 2011).
In addition, episodes that are recalled are often ones that have been
recounted many times in the past, and may therefore contain more
semantic information than actual episodic recall (Pause et al., 2013). Despite these criticisms, existing tests of episodic memory clearly have been useful (e.g., Bäckman et al., 2001), but they may miss aspects of real-world episodic memory.
Recently, a number of new tests have been developed to
try and overcome some of the drawbacks of the traditional tests and gain
more ecological validity. Some of these tests are based on a
reconceptualization of episodic memory which was originally adapted for
use with non-human animals. In the absence of language, the tests are
based on the animal experiencing two unique episodes, and then
demonstrating through their behavior what is remembered about these two
episodes. These tests emphasize the long-term retention of unique
information about events in their spatiotemporal context. In the first
study to do so, food-hoarding California scrub jays (Aphelocoma californica)
hid two types of food on each of two separate occasions. Having been
trained to know that the preferred food type degrades after several
days, but the non-preferred one does not, they were then tested shortly
after the second hiding episode. They only recovered the preferred food
in the locations where they had hidden it in the second hiding episode,
showing that they remembered which food (what) they had hidden in which
locations (where) and on which occasion (when) (Clayton and Dickinson, 1998). Since then, several variations on this task have been developed for other animals, including other birds (Feeney et al., 2009; Zinkivskay et al., 2009; Gould et al., 2012), as well as rats and mice (Dere et al., 2005; Eacott et al., 2005; Babb and Crystal, 2006; Kart-Teke et al., 2006; Roberts et al., 2008).
More recently, adaptations of these tasks have been
developed for humans. In a typical task, participants experience one or
two unique events, and then have to recall what happened where, and when
(Pause et al., 2010; Plancher et al., 2010; Hayne and Imuta, 2011; Holland and Smulders, 2011; Russell et al., 2011; Easton et al., 2012; Russell and Hanna, 2012; Cheke and Clayton, 2013; Inostroza et al., 2013; Saive et al., 2013, 2014; Newcombe et al., 2014; Weber et al., 2014).
This is either in terms of “in which of the two episodes,” or “when in
the episode,” asking about the sequence in which things happened. Some
of these methods require an explicit response from the participants, but
some try to assess memory purely based on behavioral responses (e.g.,
exploration behavior; Pause et al., 2010; Weber et al., 2014). All of these novel methodologies (with the exception of Holland and Smulders, 2011)
use displays on a computer as the information to be remembered.
However, we believe that this lacks the richness and complexity of
real-world situations, which are part of natural episodic memories, and
may therefore be less natural for older people to interact with. A
real-world task may also have better real-world predictive value.
In the current study, we use a further adapted version of the task first reported by Holland and Smulders (2011).
In this task, participants hide eight different objects in eight
different locations (indicated by the experimenter) in a real-world room
on each of two occasions on the same day (see Materials and Methods for
details). After another 2 h, participants are then taken back into the
room and asked to recall which objects they have hidden where, and on
which occasion. The participants are told a cover story about the study,
so that they would encode the information incidentally, rather than
intentionally. Therefore, this task tests relatively long-term memory
(>60 min; Pause et al., 2013)
for incidentally-encoded information about unique and somewhat unusual
(and hence arousing) episodes in their spatiotemporal contexts, hence
fulfilling all seven of Pause et al.'s (2013)
criteria for a good test of episodic-like memory. We also ask them
about their subjective experience of the recall, fulfilling the
criterion for real episodic memory as well. Part of the memory retrieval
is based on free recall, although the spatial locations are in view of
the participant and could therefore be solved using a familiarity
mechanism. Because the participants move around a real environment and
interact with real objects and locations, the task has added ecological
validity over computer-based or paper-based tests. Because the objects
are all unique, the task also allows us to test object memory and
spatial memory independently of the memory for how different features of
the episodes are bound together. The goal of the study was to
investigate whether older participants would show a deficit in this
novel test of episodic memory, and to compare their performance to other
cognitive tasks in which age differences are well established.
More at link.
No comments:
Post a Comment