Professional Documents
Culture Documents
2010-The Effects of Sleep On Episodic Memory in Older and Younger Adults PDF
2010-The Effects of Sleep On Episodic Memory in Older and Younger Adults PDF
Memory
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/pmem20
To cite this article: Mariam Aly & Morris Moscovitch (2010) The effects of sleep on episodic memory in older and
younger adults, Memory, 18:3, 327-334, DOI: 10.1080/09658211003601548
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)
contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy, completeness, or suitability
for any purpose of the Content. Any opinions and views expressed in this publication are the opinions
and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of
the Content should not be relied upon and should be independently verified with primary sources of
information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands,
costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or
systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution
in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions
MEMORY, 2010, 18 (3), 327334
Evidence on sleep-dependent benefits for episodic memory remains elusive. Furthermore we know little
Downloaded by [University of Connecticut] at 13:35 08 October 2014
about age-related changes on the effects of sleep on episodic memory. The study we report is the first to
compare the effects of sleep on episodic memories in younger and older adults. Memories of stories and
personal events were assessed following a retention interval that included sleep and following an equal
duration of wakefulness. Both older and younger adults have superior memory following sleep compared
to following wakefulness for both types of material. Amount of forgetting of personal events was less
during wakefulness in older adults than in younger adults, possibly due to spontaneous rehearsal.
Amount of time spent sleeping correlated highly with sleep benefit in older adults, suggesting that
quantity of total sleep, and/or time spent in some stages of sleep, are important contributors to age-
related differences in memory consolidation or protection from interference during sleep.
The notion that sleep may enhance our capacity Despite a plethora of research on the topic,
for alert mental functioning is a widespread one, knowledge about the effects of sleep on memory
reflected in the popularity of advice to ‘‘sleep is limited in two respects. First, while experiments
tight and wake bright’’. Far from mere folk investigating procedural memory improvements
wisdom, this idea has been given empirical have consistently shown benefits gained from
validation with regard to memory in particular. sleep, evidence of sleep-dependent episodic mem-
As early as 1924 the studies of Jenkins and ory consolidation has been more elusive (for
Dallenbach demonstrated superior memory for examples see Plihal & Born, 1997; Yaroush,
nonsense syllables following an interval contain- Sullivan, & Ekstrand, 1971; Walker, 2005). Sec-
ing sleep, compared to an equivalent duration of ond, age-related changes in the effects of sleep on
wakefulness. Since then many experiments em- memory have been largely ignored. Our paper
ploying different paradigms and assessing differ- aims to rectify this situation.
ent aspects of memory have been conducted, Sleep may have differential effects on memory
providing support for both episodic and proce- for older and younger adults, due to a multitude
dural memory consolidation during sleep (see of changes in sleep architecture that occur with
Gais & Born, 2004; Paller & Voss, 2004; Payne & age (Hornung, Danker-Hopfe, & Heuser, 2005;
Nadel, 2004; Walker, Brakefield, Morgan, Hobson, Van Cauter, Leproult, & Plat, 2000; Wolkove,
& Stickgold, 2002; Walker et al., 2003). In this Elkholy, Baltzan, & Palayew, 2007). Most impor-
paper we examine for the first time the effects tant for the present research is the finding that
of ageing on the benefit that sleep has for episodic there is a substantial decline in slow-wave sleep
memory. with age, and eventually stages 3 and 4 are
Address correspondence to: Mariam Aly, Department of Psychology, University of Toronto, 100 Saint George St., Toronto,
ON M5S 3G3, Canada. E-mail: mariam.aly@utoronto.ca
This work was supported by a CIHR Grant to Morris Moscovitch and Gordon Winocur.
# 2010 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business
http://www.psypress.com/memory DOI:10.1080/09658211003601548
328 ALY AND MOSCOVITCH
completely absent (Rosenzweig, Breedlove, & memory for the material in older and younger
Watson, 2005). Since these stages of sleep pre- adults, so that any age-related difference following
sumably underlie the consolidation of episodic periods of sleep or wakefulness would reflect
memories (Plihal & Born, 1997; Yaroush et al., only changes from a common baseline, rather
1971), deficits in SWS point to the possibility of than ones that would be confounded by baseline
less sleep-dependent episodic memory consolida- differences.
tion in older adults. Furthermore, older adults Although this is a strictly behavioural study,
show impairments on hippocampus-dependent finding equivalent benefits of sleep in older and
memory tasks (Lupien et al., 1998): changes in younger adults would suggest that those neurolo-
the structure and function of the hippocampus gical processes that confer a beneficial effect on
and prefrontal cortex may underlie memory memory during sleep are sufficiently intact, at
deficits in older adults and affect the quality of least in healthy older adults, to allow them to
sleep-dependent consolidation (Hornung et al., operate. If, however, even under conditions that
2005). promote encoding and retention, the age-related
Two studies that investigated the effects of decline in memory consolidation persists, it would
Downloaded by [University of Connecticut] at 13:35 08 October 2014
sleep on memory in older and younger adults suggest that it is related to neurological changes
confirmed that older adults benefited less from that accompany old age, such as the reduction of
sleep (Spencer, Gouw, & Ivry, 2007; Spencer, SWS or hippocampal deterioration.
Sumn, & Ivry, 2006). Explicit and implicit con-
textual sequence learning tasks were used. Both
tasks were presumed to engage hippocampal METHOD
activity. These tasks are therefore similar to
episodic memory tasks, which similarly depend Participants
on the hippocampus.
Younger adults showed greater memory im- A total of 10 younger (1929 years old) and 12
provements on these tasks following a 12-hour older (6980 years old) adults participated for $10
interval that included sleep, compared to an equal per hour. Younger adults were recruited through
duration spent awake. Older adults showed no flyers placed around the University of Toronto.
such sleep-dependent benefits, pointing to an age- Older adults were volunteers in the Adult Volun-
related decline of sleep-dependent memory con- teer Pool, and were recruited by phone. The data
solidation. Since this task presumably engaged of two older adults were discarded because they
hippocampal activity, the findings suggest the did not follow instructions, leaving a sample of 10
possibility that episodic memories, which simi- older adults. Participants were fluent in English,
larly depend on the hippocampus, may also be did not have any self-reported neurological,
differentially affected by sleep in older and psychological, or severe medical conditions (e.g.
younger adults. epilepsy, traumatic brain injury), and were not on
Research has suggested, however, that age- any medication that affects sleep. Ethics approval
related memory deficits may disappear when the was granted by the Department of Psychology
studied material engages the older adults’ interest Ethics Review Committee at the University of
rather than being neutral and indifferent (Car- Toronto.
stensen & Turk-Charles, 1994; Rahhal, May, &
Hasher, 2002). Because the material used in the
above studies*word lists and sequences of Materials
lights*resemble the latter more than the former,
it is possible that using material that is inherently Two measures of episodic memory were used.
more interesting for the older adults would elim- The first was the Logical Memory section of the
inate the age-related decline in sleep-dependent Wechsler Memory Scale III (WMS-III), which
memory consolidation. We decided, therefore, to consists of two short-paragraph-length stories
use stories and personally experienced events, (Story A and Story B). The stories were scored
both because such material would engage the in terms of the number of story units recalled, as
older adults’ interest, and because memory for specified in the WMS-III scoring protocol.
details of stories and personally experienced The second measure was a list of 12 questions
events is dependent on the hippocampus. We to assess personal episodic memories for the first
also thought it important to equate the initial (or last) conversation the participant had that
AGEING, SLEEP, AND EPISODIC MEMORY 329
morning (or the previous evening), as well as immediately, in as much detail as they could
memory for the first (or last) thing they read or remember. This was done to ensure that partici-
saw on the news, radio, or television (see Appendix). pants paid attention to the story, and also to
The questions were the same for each session with equate older and younger adults on initial mem-
the exception of the final two, of which one version ory. In cases where older adults’ memory was
was made to be appropriate for the morning and the significantly poorer than the average of younger
other for the evening. adults, the story was repeated until they approxi-
Finally, to gather information about partici- mated or passed the average (in all cases, only
pants’ sleep the previous night and over the one repetition was required).
preceding week, the St. Mary’s Hospital Sleep Participants were then asked the 12 questions
Questionnaire (Ellis et al., 1981) and the Pitts- about events that happened the previous evening.
burgh Sleep Quality Index (Buysse, Reynolds, They were told to answer them in as much detail
Monk, Berman, & Kupfer, 1989) were adminis- as they could, taking their time to think about the
tered. Questions assess sleep latency, sleep qual- answer if that was needed, but at the same time
ity, amount of sleep, and sleep satisfaction. This keeping their answer limited to what the question
Downloaded by [University of Connecticut] at 13:35 08 October 2014
older adult) mixed-model analysis of variance was interval than initially, and a score less than 1.0
conducted on the number of story units recalled indicates forgetting of the story over the delay.
at encoding. There was no main effect of encod- As Figure 1 illustrates, older adults performed
ing time or age group, and no interaction between worse than young adults overall, but they bene-
the two, thus ensuring that our encoding manip- fited as much as younger adults from sleep. This
ulations yielded equivalent initial memory for all latter impression was confirmed by a main effect
comparisons (morning encoding: 16.9 and 16.6; of time of test, F(1, 18)25.10, p.001. There
evening encoding: 15.2 and 16.3 story units for was no main effect of age group, or an age group
younger and older adults respectively). Further- by time of test interaction. The absence of a
more, a 2 (order of testing: AM-PM-AM or PM- significant main effect of age on memory, despite
AM-PM)2 (age group: younger or older adult) an apparent difference, may have arisen from low
analysis of variance on memory following sleep power due to the small number of participants.
Analysis of the simple main effect showed that
and following wakefulness revealed that order of
younger adults’ memories were significantly bet-
testing had no effect on memory performance,
ter following sleep than following wakefulness,
Downloaded by [University of Connecticut] at 13:35 08 October 2014
Figure 1. Mean proportion recall of WMS-III story following wakefulness and following sleep, for younger and older adults.
Proportion recall is number of story units recalled at test divided by number of story units recalled at encoding. Error bars indicate91
SEM *p .01, simple effect of time of test for older adults; **p.001, simple effect of time of test for younger adults; ***p.001, main
effect of time of test.
AGEING, SLEEP, AND EPISODIC MEMORY 331
group (younger or older adult) as the between- or anything out of the ordinary in their sleep
participants variable. There was a main effect of the night before, and none was very unsatisfied
time of test, F(1, 18)36.70, p.001, and a with his or her sleep. Thus we were ensured
significant time of test by age group interaction, that all participants slept reasonably well the
F(1, 18)5.70, p.028. night before.
Analysis of the simple main effect revealed that Younger and older adults did not differ in total
younger adults’ personal memories were signifi- number of hours slept during the sleep retention
cantly better following sleep compared to follow- interval (6.60 and 7.48 hours for older and younger
ing wakefulness, t(9)5.87, p.001. Younger adults, respectively). However, the difference
adults received an average score of 0.76 following between time spent asleep in the first, SWS-rich
wakefulness, compared to 0.94 following sleep, a half of the night was almost significant, with older
difference of 0.18. Older adults’ personal mem- adults having more early sleep than younger
ories were also significantly superior following adults, t(18)2.05, p.056 (3.26 and 2.24 hours
sleep compared to following wakefulness, t(9) for older and younger adults, respectively). Total
2.64, p.027. Older adults received an average number of hours slept during the sleep retention
score of 0.80 following wakefulness, compared to interval was significantly correlated with benefit
Figure 2. Mean proportion scores for details in personal event memories following wakefulness and following sleep, for younger
and older adults. Scores are number of points obtained divided by the maximum number of points possible. Error bars indicate91
SEM *p .03, simple effect of time of test for older adults; **p .0002, simple effect of time of test for younger adults; ***p.001,
main effect of time of test.
332 ALY AND MOSCOVITCH
from sleep (defined as WMS proportion recall structure, particularly those that are emotionally
following sleep minus WMS proportion recall engaging or personally relevant, may be a more
following wakefulness) for older adults, r.775, appropriate measure of the effects of ageing on
p.008, but not for younger adults. No other sleep and memory than lists of paired words.
significant correlations between measures of sleep Stories and personal events likely engage recol-
duration and measures of memory performance lective processes more than single words or word
were found. Finally, daytime naps did not affect pairs, memory for which may depend as much on
memory performance following wakefulness. familiarity as on recollection (Eichenbaum, Yo-
Thus, both younger and older adults’ memories nelinas, & Ranganath, 2007). Considering the
benefited from a night of sleep. This was true crucial role of the hippocampus in recollection
both for memories for stories and for personal and in sleep-related effects (Gais & Born, 2004;
memories. The interaction between age group Payne & Nadel, 2004), the results suggest that
and time of test for the personal memories could sleep confers benefits on recollection more than
have arisen because older adults’ memories did on familiarity.
not benefit as much from sleep as the memories
Downloaded by [University of Connecticut] at 13:35 08 October 2014
differentially affected following sleep and wake- may even be magnified when material is used that
fulness. has a narrative structure that engages the partici-
Whether the observed benefits from sleep are pant and his or her hippocampus. There seems,
due to slow-wave sleep in particular, as opposed after all, to be some truth to the cliché ‘‘sleep tight
to sleep in general, or a period of relative and wake bright’’.
quiescence and little interference, cannot be
Manuscript received 12 October 2009
determined from the present research, nor was
Manuscript accepted 22 December 2009
the study designed to answer that question. First published online 24 February 2010
Nonetheless, the results from the sleep question-
naires provide a clue. While there was no
difference between the age groups in total num-
ber of hours spent asleep before the morning test
session, older adults slept more during the early REFERENCES
part of the night than did younger adults. Back-
haus et al. (2007) found that when middle-aged Backhaus, J., Born, J., Hoeckesfeld, R., Fokuhl, S.,
Downloaded by [University of Connecticut] at 13:35 08 October 2014
and younger-adult groups had the same amount Hohagen, F., & Junghanns, K. (2007). Midlife
decline in declarative memory consolidation is
of slow-wave sleep (following early sleep in the correlated with a decline in slow wave sleep.
middle-aged group and late sleep in the younger- Learning & Memory, 14, 336341.
adult group), the effects of sleep on declarative Buysse, D. J., Reynolds, C. F., Monk, T. H., Berman,
memory did not differ between the two groups. S. R., & Kupfer, D. J. (1989). The Pittsburgh sleep
Importantly, initial performance was equivalent quality index: A new instrument for psychiatric
practice and research. Psychiatry Research, 28,
in the two age groups, which coheres with the 193213.
above-mentioned possibility that sleep-dependent Carstensen, L. L., & Turk-Charles, S. (1994). The
episodic memory benefits may be observed only salience of emotion across the adult life span.
with initial memories of sufficient strength. In the Psychology and Ageing, 10, 111122.
present experiments the memories of both age Eichenbaum, H., Yonelinas, A. P., & Ranganath, C.
(2007). The medial temporal lobe and recognition
groups may have benefited similarly from sleep if memory. Annual Review of Neuroscience, 30, 123
the older adults spent a comparable amount of 152.
time in the slow-wave stages of sleep as younger Ellis, B. W., Johns, M. W., Lancaster, R., Raptopoulos,
adults. For older adults, however, there was a P., Angelopoulos, N., & Priest, R. G. (1981). The St.
significant positive correlation between story Mary’s Hospital Sleep Questionnaire: A study of
reliability. Sleep, 4, 9397.
memory benefits following sleep and total num- Gais, S., & Born, J. (2004). Declarative memory
ber of hours spent asleep before the morning test consolidation: Mechanisms acting during human
session. This raises the possibility that the entire sleep. Learning & Memory, 11, 679685.
night of sleep, or a long period of quiescence and Hornung, O. P., Danker-Hopfe, H., & Heuser, I. (2005).
relatively little interference, may be an important Age-related changes in sleep and memory: Com-
monalities and interrelationships. Experimental Ger-
factor in promoting better memory following the ontology, 40, 279285.
sleep retention interval. Because sleep stages Jenkins, J. G., & Dallenbach, K. M. (1924). Oblivis-
were not monitored, no firm conclusions can be cence during sleep and waking. American Journal of
reached on the basis of our results about which Psychology, 35, 605612.
stages of sleep, if any, are important in episodic Lupien, S. J., de Leon, M., de Santi, S., Convit, A.,
Tarshish, C., Nair, N. P., et al. (1998). Cortisol levels
memory consolidation. during human ageing predict hippocampal atrophy
Nonetheless, our study provides a crucial start- and memory deficits. Nature Neuroscience, 1, 34.
ing point in data and methodology from which Mednick, S. C., Nakayama, K., & Stickgold, R. (2003).
future research can progress. It not only strength- Sleep-dependent learning: A nap is as good as a
ens support for the possible beneficial effects of night. Nature Neuroscience, 6, 697698.
Paller, K. A., & Voss, J. L. (2004). Memory reactivation
sleep on episodic memory, for which evidence has and consolidation during sleep. Learning & Mem-
been elusive, but also extends previous findings by ory, 11, 664670.
showing that these effects exist for young and old Payne, J. D., & Nadel, L. (2004). Sleep, dreams, and
alike. Moreover, this experiment was the first to memory consolidation: The role of the stress hor-
use memory for stories and for personally experi- mone cortisol. Learning & Memory, 11, 671678.
Plihal, W., & Born, J. (1997). Effects of early and late
enced events as the to-be-remembered items, nocturnal sleep on declarative and procedural
showing that declarative memory benefits follow- memory. Journal of Cognitive Neuroscience, 9, 534
ing sleep are not restricted to lists of words, and 547.
334 ALY AND MOSCOVITCH
Rahhal, T. A., May, C. P., & Hasher, L. (2002). Truth 3. What were they wearing? [Omit if phone conversation.]
and character: Sources that older adults can re- 4. What was the last thing you said to them, other than
member. Psychological Science, 13, 101105. something generic, like ‘‘goodbye’’?
Rosenzweig, M. R., Breedlove, S. M., & Watson, N. V. 5. What time did the conversation take place?
(2005). Biological rhythms, sleep, and dreaming. In 6. How long did the conversation last?
Biological psychology: An introduction to behavioral 7. Where did the conversation take place? Where were you
and cognitive neuroscience (pp. 424455). Sunder- situated in that location?
land, MA: Sinauer Associates Inc. 8. Describe two aspects of the surroundings specific to that
Spencer, R. M. C., Gouw, A. M., & Ivry, R. B. (2007). episode (i.e., do not say ‘‘refrigerator’’ if there is always a
Age-related decline of sleep-dependent consolida- fridge in the kitchen).
tion. Learning & Memory, 14, 480484. 9. What were you doing before the conversation?
Spencer, R. M. C., Sumn, M., & Ivry, R. B. (2006). 10. What did you do after the conversation was over?
Sleep-dependent consolidation of contextual learn- 11. What was the last thing you watched on TV or heard on
ing. Current Biology, 16, 10011005. the radio? [Omit from scoring if they did not watch TV or
Van Cauter, E., Leproult, R., & Plat, L. (2000). Age- listen to the radio.]
related changes in slow wave sleep and REM sleep 12. What was the last thing you read before bed? Describe one
and relationship with growth hormone and cortisol thing about it. [Omit from scoring if they did not read.]
levels in healthy men. The Journal of the American
Downloaded by [University of Connecticut] at 13:35 08 October 2014