SLEEP 2011 Abstract Supplement

A. Basic Science IX. Learning, Memory and Cognition
0212
SLEEP’S (COMPLICATED AND UNEXPECTED)
INFLUENCES ON VERBAL MEMORY
Sheth B 1,2 , Varghese R 3 , Truong T 3
1
Electrical & Computer Engineering, University of Houston, Houston,
TX, USA, 2 Center for NeuroEngineering and Cognitive Science,
University of Houston, Houston, TX, USA, 3 University of Houston,
Houston, TX, USA
Introduction: Sleep protects verbal memories by boosting resistance
from subsequent (prospective) associative interference. We asked: (1)
Is sleep indispensable to verbal memory? (2) Does sleep protect verbal
memory from retroactive and non-associative forms of interference? (3)
Does sleep enhance verbal memory or help retain it better? The goal was
to delineate the influences and limits of sleep’s role in affecting recently
acquired verbal memories.
Methods: Each volunteer (n=160 total) participated in only one of our
experiments. Participants memorized unrelated pairs of words A i
-B i
.
Following an intervening period of 12-24 hours that did or did not include
sleep, participants memorized a second pair of interfering words
A i
-C i,
/C-D just before being tested on their memory of both.
Results: (1) Reducing required rehearsal during training increased the
benefits of sleep on test recall 12 hours later. Increasing required rehearsal
during training entirely compensated for the benefit of sleep
(sleep-85%, wake-90%). (2) When observers acquired non-associative
C-D pairs 12 hours following A-B pairs, sleep provided a statistically indistinguishable
benefit in test recall of A-B word pairs (13%) compared
with when observers acquired A-C pairs (19%). In another experiment,
participants learnt 60 A-B word pairs. Twelve hours later, they acquired
A-C word pairs. There was a significant (14%) sleep-dependent benefit
on test recall of A-C word pairs. (3) A-B recall 12 hours versus right
after training did not differ.
Conclusion: Sleep is one way, but not an exclusive one, of sheltering
verbal memory from associative prospective interference. Sleep can also
protect verbal memories from non-associative and retroactive interference.
Sleep does not help enhance verbal memory but shelters it from
fading. Additional analysis demonstrates that sleep prior to memory encoding
provides no benefit. Combined, our results are more consistent
with the idea that sleep benefits verbal memory by preventing its subsequent
re-exposure rather than by replaying it.
0213
NOVEL OBJECT LEARNING DEPENDS ON RAPID EYE
MOVEMENT SLEEP
Mednick SC 1 , McDevitt EA 1 , Brady M 2
1
Psychiatry, 9116a, UCSD, San Diego, CA, USA, 2 Psychology,
University of North Dakota, Grand Forks, ND, USA
Introduction: A fundamental function of biological vision is to detect
and recognize potential food items and predators from naturally cluttered
backgrounds when form and coloration of target objects are similar
to the background. Brady and Kersten (2003) previously showed that
subjects can do this via bootstrapped learning. Sleep, specifically rapid
eye movement (REM), has been shown to enhance perceptual learning.
Here, we examined the role of REM sleep in object learning in ambiguous
backgrounds.
Methods: At 9AM and 5PM, 82 subjects were tested on an object learning
task. At 1PM, subjects fitted with EEG electrodes and randomized
to a quiet rest condition (QR) or nap condition with or without REM.
Stimuli were artificial morphogenic objects that appear to be organic
forms but do not resemble a familiar class of organism and are camouflaged
by similar shapes. During AM training, subjects were shown
three objects in cluttered backgrounds and asked to trace each object in
the foreground (Trace condition). Next, subjects passively view these
camouflaged objects appearing with a characteristic sound. At PM test,
subjects were asked to recognize objects in the Test condition and retrace
objects in the Trace condition.
Results: REM group was significantly better than NREM and QR in
recognition (Test), both QR and NREM performed just above chance.
Both nap groups were significantly better than QR in the Trace condition.
Conclusion: We found an essential role for REM sleep in learning to
detect and recognize novel objects from camouflage. REM occupies a
large proportion of sleep during early development, and has been hypothesized
to be related to cortical plasticity. The following results suggest
that learning to recognize and segment objects that share the same
form and coloration as their background requires REM sleep in adulthood
and has implications for infant object learning.
Support (If Any): K01MH080992
0214
SPATIAL MEMORY PERFORMANCE SHIFTS ITS
ASSOCIATION FROM SLOW WAVE SLEEP TO REM SLEEP
WHEN AN EMOTIONAL COMPONENT IS ADDED
Stenstrom P 1,3 , Nielsen TA 2 , Solomonova E 3 , Bujwid V 4 , Amato A 4
1
Psychiatry, Harvard Medical School, Boston, MA, USA, 2 Psychiatry,
University of Montreal, Montreal, QC, Canada, 3 Psychology,
University of Montreal, Montreal, QC, Canada, 4 Psychology, McGill
University, Montreal, QC, Canada
Introduction: While many studies link slow wave sleep (SWS) to the
processing of declarative memory, a small but growing literature demonstrates
a role for rapid eye movement (REM) sleep when such memory
is emotionally charged. Here we examined if a SWS-associated task
shifts its dependence to REM sleep when this task is made emotional.
Methods: We had 16 participants undergo two matched spatial navigation
tasks that differed in emotional tone and observed how performance
in each task was affected by one night of selective REM deprivation.
Participants slept one night in a sleep laboratory and were randomly assigned
to either a REM sleep deprived (REMD; awakened after 5 minutes
of REM sleep) or a control (CTL; awakened after 25 minutes of
REM sleep). Ten to 12 days post-task participants were given surface
maps of the two versions of the spatial task (fig. S1) and instructed to
mark the houses they had visited.
Results: The REMD group was successfully REM sleep deprived,
showing lower %REM sleep (M=9.51, SD=1.08) than the CTL group
(M=16.71, SD=5.57; t12=3.36, p=.005). The REMD group had more
missed targets than the CTL group on the emotional version of the task
(REMD: M=2.86, SD=.38; CTL: M=2.00, SD=.58; t12=-3.29, p=.01)
but fewer wrong targets on the neutral version (CTL: M=2.67, SD=.79;
REMD: M=1.43, SD=.78: t12=2.76, p=.02). Missed targets on the emotional
version of the task were negatively correlated with min in REM
sleep (r16=-0.71, p=.005), and %REM sleep (r16=-0.71, p=.004). Wrong
targets on the neutral version of the task were negatively correlated with
min in Stage 4 sleep (r16=-0.62, p=.01) and %Stage 4 sleep (r16=-0.65,
p=.006) and positively correlated with %REM sleep (r16=0.55, p=.03).
Conclusion: Together, these findings suggest that imbuing a spatial
memory task with dysphoric emotions switches its association from
SWS to REM sleep.
Support (If Any): This research was supported by a Natural Sciences
and Engineering Research Council of Canada grant and a Natural Sciences
and Engineering Research Council of Canada Graduate Scholarship.
SLEEP, Volume 34, Abstract Supplement, 2011
A76