Dissociable learning-dependent changes in REM and non-REM sleep in declarative and procedural memory systems

Sleep spindles and rapid eye movements have been found to increase following an intense period of learning on a combination of procedural memory tasks. It is not clear whether these changes are task specific, or the result of learning in general. The current study investigated changes in spindles, rapid eye movements, K-complexes and EEG spectral power following learning in good sleepers randomly assigned to one of four learning conditions: Pursuit Rotor (n = 9), Mirror Tracing (n = 9), Paired Associates (n = 9), and non-learning controls (n = 9). Following Pursuit Rotor learning, there was an increase in the duration of Stage 2 sleep, spindle density (number of spindles/min), average spindle duration, and an increase in low frequency sigma power (12-14 Hz) at occipital regions during SWS and at frontal regions during Stage 2 sleep in the second half of the night. These findings are consistent with previous findings that Pursuit Rotor learning is consolidated during Stage 2 sleep, and provide additional data to suggest that spindles across all non-REM stages may be a mechanism for brain plasticity. Following Paired Associates learning, theta power increased significantly at central regions during REM sleep. This study provides the first evidence that REM sleep theta activity is involved in declarative memory consolidation. Together, these findings support the hypothesis that brain plasticity during sleep does not involve a unitary process; that is, different types of learning have unique sleep-related memory consolidation mechanisms that act in dissociable brain regions at different times throughout the night. (c) 2007 Elsevier B.V. All rights reserved.