How sleep affects memory for future-relevant information

Abstract

Research in three parts investigated sleep’s preferential consolidation of memories for experiences that are prioritized at encoding due to intrinsic characteristics (e.g., emotion), extrinsic characteristics (e.g., instructed learning, reward), or both. Results showed that sleep broadly strengthens memory for future-relevant information, with these prioritization cues at encoding aiding in the selection process for what is subsequently strengthened during sleep. Part I investigated the effects of sleep on the consolidation of information that was prioritized at encoding due to the intrinsic cue of emotion. Results showed that even once the emotionally salient aspect of the stimuli was removed (i.e., when memory was tested using a neutral cue), residual effects of emotion were reflected in enhanced visual activity following sleep, with this visual activity correlating with the percentage of rapid eye movement sleep obtained during consolidation and likely driven by enhanced occipital-hippocampal connectivity following sleep. This suggests that sleep prioritizes information that was salient due to the intrinsic cue of emotion at encoding, leading to changes in neural activity during retrieval even once that intrinsic cue is no longer present. As in Part I, most prior research has examined how sleep preferentially consolidates memory for information that is salient due to a single cue for future relevance. Part II investigated whether future relevance can be assigned to stimuli via top-down manipulations (i.e., extrinsic prioritization cues), as well as how sleep prioritizes memory for information when intrinsic and extrinsic cues for future relevance co-occur within the same stimuli. Results suggest that when multiple dimensions of future relevance co-occur, sleep prioritizes extrinsic cues (i.e., instructed learning, and to a lesser degree, reward) over intrinsic cues (i.e., emotion). Further, results suggest that additional cues for future relevance do not have additive effects on consolidation, but rather that sleep may binarize information based on whether it is future-relevant or not, preferentially consolidating memory for the former category. Lastly, Part III focused on a manipulation of extrinsic prioritization at encoding to investigate both how the effects of prioritization on memory differ minutes after encoding relative to after long-term consolidation processes take place, and also whether these effects depend on if a healthy versus restricted amount of nocturnal sleep is obtained during the consolidation interval. Results showed that a top-down manipulation of prioritization (i.e., typographical cueing) was effective in enhancing memory; highlighted relative to non-highlighted content was better remembered at multiple time points, with evidence suggesting that N3 (slow-wave) sleep may contribute to these memorial benefits. Together, findings across Parts I-III suggest that sleep selectively strengthens future-relevant information, elucidating which cues for future relevance at the time of encoding lead to enhanced consolidation following sleep, as well as how sleep acts on intrinsic and extrinsic prioritization cues when they co-occur. In identifying intrinsic targets of sleep’s selective consolidation effects, as well as extrinsic manipulations that can be applied to use sleep as a tool to enhance consolidation, these three studies have important implications for optimizing memory that are relevant across domains.