Remembering specific features of emotional events across time: The role of REM sleep and prefrontal theta oscillations

Abstract

When an episode of emotional significance is encountered, it often results in the formation of a highly resistant memory representation that is easily retrieved for many succeeding years. Recent research shows that beyond generic consolidation processes, rapid eye movement (REM) sleep importantly contributes to this effect. However, the boundary conditions of consolidation processes during REM sleep, specifically whether these extend to source memory, have not been examined extensively. The current study tested the effects of putative consolidation processes emerging during REM sleep and slow wave sleep (SWS) on item and source memory of negative and neutral images, respectively. Results demonstrate superior emotional relative to neutral item memory retention after both late night REM sleep and early night SWS. Emotional source memory, on the other hand, exhibited an attenuated decline following late night REM sleep, whereas neutral source memory was selectively preserved across early night SWS. This pattern of results suggests a selective preservation of emotional source memory during REM sleep that is functionally dissociable from SWS-dependent reprocessing of neutral source memory. This was further substantiated by a neurophysiological dissociation: Postsleep emotional source memory was selectively correlated with frontal theta lateralization (REM sleep), whereas postsleep neutral item memory was correlated with SWS spindle power. As such, the present results contribute to a more comprehensive characterization of sleep-related consolidation mechanisms underlying emotional and neutral memory retention. Subsidiary analysis of emotional reactivity to previously encoded material revealed an enhancing rather than attenuating effect of late night REM sleep on emotional responses.

Sleep selectively stabilizes contextual aspects of negative memories

“…Considering the emotional dimension, the selective sleep benefit for negative contextual memory is consistent with sleep's negative vs. neutral advantages for associative memories of temporal order 33 , and spatial location 32 , although these splitnight studies found this effect to depend on opposing sleep stages (NREM-rich early vs. REM-rich late sleep, respectively). More broadly, these observations fit with the notion that the sleeping brain preferentially processes information deemed of future relevance 7 .…”

“…Given that these studies focused on pairs of items, one possibility is that sleep's effects depend on whether the type of association tested relates to links with background contextual components (in which case sleep may favor the stabilization of links with negative memories), or links between individual items (in which case the emotional advantage may be suppressed or even reversed). Complicating this view, the aforementioned split-night study that found superior picture-location memory for negative vs. neutral items across late REM-rich sleep observed the opposite pattern across early NREM-rich sleep 32 . Another split-night study observed a similar early-night benefit for neutral picture-color associations, but did not observe this finding for picture-location associations (for which no differential negative/neutral or early/late sleep effects were found) 25 .…”

“…A prominent theoretical account posits that sleep, and REM sleep in particular, serves to preserve the content of emotional memories while simultaneously removing their affective layer 55 . While the "emotional blunting" aspect of this model has received physiological support 35,36 , subjective ratings of arousal and/or valence have found inconsistent sleep-related modulation of emotional tone 18,19,23,24,32,37,38 , as reviewed in the introduction. Our findings provide some support for a role of sleep in reducing the emotional tone of memories, in that sleep reduced the perceived emotionality of both negative and mildly positive stimuli.…”

“…Some studies report no difference in emotional reactivity to emotional versus neutral items, with similar decreases 24 or increases 18 across sleep versus wake, or comparable ratings after NREM and REM sleep for both emotional categories 23 . Other investigators have described a REM-dependent preservation of negative ratings relative to wake 19 or REM deprivation 37 , while yet others have reported enhanced emotional reactivity after REM-rich sleep 32,38 .…”

“…These inconsistencies may be related to various experimental design factors (e.g., sleep vs. wake; full night vs. nap, split-night; sleep-deprivation; targeted memory reactivation), as well as the memory paradigm employed. Specifically, the clearer effect of sleep on hippocampus-dependent associative memories compared to item recognition could result in greater emotional benefits for memories with an associative component, but again, the available evidence is contradictory 25,[32][33][34] .…”