All Memories Are Equal at Night: Sleep Protects Without Preference for Strength
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Objectives
Sleep benefits memory, but whether this benefit depends on initial memory strength remains heavily debated. We asked whether overnight sleep preferentially stabilizes weakly vs. strongly encoded declarative items, and whether classical sleep oscillations account for any benefit.
Methods
Ninety healthy adults (n=30/group) learned 80 unrelated word pairs; half were presented twice (S−, weak), while the other half was studied three times (S+, strong). Memory retention was tested after (i) 9hrs containing nocturnal sleep (n= 30), (ii) a comparable daytime waking period (n= 30), or (iii) immediately (∼40-min) after encoding. Accuracy (cued recall) and reaction times (RT) indexed memory accessibility. In the sleep group, polysomnography was used to quantify slow oscillation (SO) density as well as fast spindle intensity to study potential relations to memory performance.
Results
Strongly encoded items (S+) were recalled better than weak items (S−) in all three groups. An interaction indicated that WAKE disproportionately forgot more weak items S− relative to strong S+ items, whereas SLEEP showed no differential benefit for weak versus strong items, with performance more closely resembling that of the immediate recall group. Reaction-time analyses echoed this pattern with WAKE responding more slowly than SLEEP and IMMEDIATE, and there is no RT difference between SLEEP and IMMEDIATE. S+ were retrieved faster than S− (with no group interaction). Finally, classical sleep markers did not account for overnight retention: spindle and slow-oscillation metrics showed no stronger associations on the learning night than on the adaptation night and did not differ by item strength, suggesting largely trait-like rather than state-dependent relationships between these oscillations and declarative memory.
Conclusions
Sleep protects declarative memories irrespective of initial strength and matches immediate recall performance. Traditional sleep metrics, such as slow oscillation density and spindle intensity, did not predict overnight retention performance but instead reflected more trait-like relationships.
