Sleep-dependent facilitation of episodic memory details

Abstract

While a role for sleep in declarative memory processing is established, the qualitative nature of this consolidation benefit, and the physiological mechanisms mediating it, remain debated. Here, we investigate the impact of sleep physiology on characteristics of episodic memory using an item- (memory elements) and context- (contextual details associated with those elements) learning paradigm; the latter being especially dependent on the hippocampus. Following back-to-back encoding of two word lists, each associated with a different context, participants were assigned to either a Nap-group, who obtained a 120-min nap, or a No Nap-group. Six hours post-encoding, participants performed a recognition test involving item-memory and context-memory judgments. In contrast to item-memory, which demonstrated no between-group differences, a significant benefit in context-memory developed in the Nap-group, the extent of which correlated both with the amount of stage-2 NREM sleep and frontal fast sleep-spindles. Furthermore, a difference was observed on the basis of word-list order, with the sleep benefit and associated physiological correlations being selective for the second word-list, learned last (most proximal to sleep). These findings suggest that sleep may preferentially benefit contextual (hippocampal-dependent) aspects of memory, supported by sleep-spindle oscillations, and that the temporal order of initial learning differentially determines subsequent offline consolidation.

Figure 1. Experimental design.

Participants studied two lists of 50 words, one after the other. The two lists were each associated with a different set of contextual cues, making each list distinct. Subjects were then randomly assigned to the Nap- or No Nap-group, with participants in the Nap-group obtaining a 120 min sleep opportunity, while the No Nap-group performed standard daily activities. After 6 hours of offline consolidation time, both groups returned for a memory recognition test, where the 100 old items were presented together with a 100 intermixed foils (new items).

Figure 2. Memory performance and sleep association.

A) Item-memory performance and B) context-memory performance for the No Nap- (clear bar) and Nap-group (filled bar). C) Within the Nap-group (filled bar in 2B) the extent of context-memory retention significantly correlated with total time spent in stage-2 sleep. Error bar represents s.e.m. * P<0.05 n.s. = non-significant.

Figure 3. The association in the Nap-group between context-memory retention and fast sleep spindles across the four electrode derivations (left top corner box label), with corresponding r- and P-values provided.

Figure 4. Context memory and sleep association.

A) Context-memory performance in the No Nap- (clear bar) and Nap-group (filled bar) for List1, and B) for List2, C) Within the Nap-group, the association between the amount of Stage-2 NREM sleep and the extent of List1 context-memory retention, and similarly D) the corresponding association for List2 retention. Error bar represents s.e.m. *P = 0.028, n.s. = non-significant.

Figure 5. The association in the Nap-group between context-memory retention for List2 and fast sleep spindles across the four electrode derivations (left top corner box label), with corresponding r- and P-values provided.