Abstract:
Sleep spindles have been found to increase following an intense period of learning
on a combination of motor tasks. It is not clear whether these changes are task specific, or
a result of learning in general. The current study investigated changes in sleep spindles
and spectral power following learning on cognitive procedural (C-PM), simple procedural
(S-PM) or declarative (DM) learning tasks. It was hypothesized that S-PM learning
would result in increases in Sigma power during Non-REM sleep, whereas C-PM and
DM learning would not affect Sigma power. It was also hypothesized that DM learning
would increase Theta power during REM sleep, whereas S-PM and C-PM learning would
not affect Theta power.
Thirty-six participants spent three consecutive nights in the sleep laboratory.
Baseline polysomnographic recordings were collected on night 2. Participants were
randomly assigned to one of four conditions: C-PM, S-PM, DM or control (C). Memory
task training occurred on night 3 followed by polysomnographic recording. Re-testing on
respective memory tasks occurred one-week following training. EEG was sampled at
256Hz from 16 sites during sleep. Artifact-free EEG from each sleep stage was submitted
to power spectral analysis.
The C-PM group made significantly fewer errors, the DM group recalled more,
and the S-PM improved on performance from test to re-test. There was a significant night
by group interaction for the duration of Stage 2 sleep. Independent t-tests revealed that
the S-PM group had significantly more Stage 2 sleep on the test night than the C group.
The C-PM and the DM group did not differ from controls in the duration of Stage 2 sleep
on test night. There was no significant change in the duration of slow wave sleep (SWS)
or REM sleep.
Sleep spindle density (spindles/minute) increased significantly from baseline to
test night following S-PM learning, but not for C-PM, DM or C groups. This is the first
study to have shown that the same pattern of results was found for spindles in SWS. Low
Sigma power (12-14Hz) increased significantly during SWS following S-PM learning but
not for C-PM, DM or C groups. This effect was maximal at Cz, and the largest increase
in Sigma power was at Oz. It was also found that Theta power increased significantly
during REM sleep following DM learning, but not for S-PM, C-PM or C groups. This
effect was maximal at Cz and the largest change in Theta power was observed at Cz.
These findings are consistent with the previous research that simple procedural
learning is consolidated during Stage 2 sleep, and provide additional data to suggest that
sleep spindles across all non-REM stages and not just Stage 2 sleep may be a mechanism
for brain plasticity. This study also provides the first evidence to suggest that Theta
activity during REM sleep is involved in memory consolidation.
