Abstract
Introduction
Effective memory often requires recall of both specific information and the context in which the information was encountered. Total sleep deprivation (TSD) is known to impair memory for information items (e.g., words on a studied list), but the impact of TSD on binding, or associative linking, between items and context is not clear.
Methods
N=68 healthy adults (ages 22–40; 35 females) completed a 4-day (3-night) in-laboratory study. After a baseline night with 10h nighttime sleep opportunity, participants were randomly assigned to either 38h TSD (n=38) or a well-rested control (WRC) condition with 10h nighttime sleep opportunity (n=30). Both study arms concluded with a 10h nighttime recovery sleep opportunity. Participants completed a standardized recognition memory task at 14:50 on day 2 (baseline, session 1) and again 24h later (session 2). The memory task consisted of a study phase in which words with negative, positive, and neutral affective valence were spoken by a female or male speaker (50% each); followed immediately by a test phase, in which subjects made recognition judgments for the items (words) and their source (speaker).
Results
Mixed-effects ANOVA revealed significant interactions of session by condition for both word and speaker recognition (p<0.001). When sleep-deprived, TSD participants recognized fewer words and, for words that were correctly recognized, they were worse at recognizing the speaker, compared to baseline and to the WRC group. Negatively valenced words were associated with poorer word recognition (p<0.001), and in session 1 poorer source recognition (p = 0.032), but these valence effects did not interact with sleep deprivation.
Conclusion
TSD impaired memory for items, but more importantly, also impaired memory for the context in which items were presented, even if the items were themselves correctly recognized, and regardless of their affective valence. These results indicate that TSD may disrupt binding of information to its context, which could explain TSD deficits in decision-making tasks that require novel associative linking. Furthermore, our findings are important in real-world situations such as eyewitness accounts and perseveration of the influence of misinformation.
Support (if any)
NIH grant R21 CA167691 and CDMRP award W81XWH-20-1-0442
X-Vectors: New Quantitative Biomarkers for Early Parkinson's Disease Detection From Speech