0062 Memory for Text Meaning Is Maintained During 24 Hours of Continuous Wakefulness

Abstract Introduction Total sleep deprivation (TSD) has been shown to impair associative memory, but little is known about the effects of TSD on memory for naturalistic text. Naturalistic text has different levels of representation: memory for surface form (verbatim text), which is short-lived; memory for textbase (ideas expressed in the text); and event model (representation of the situation described in the text). We investigated whether TSD differentially impacts memory by level of representation. Methods N=10 female active-duty dayshift (DS; n=5) and nightshift (NS; n=5) workers (ages 35.1±8.1y) matched by age participated in a 36h laboratory study. After waking at their habitual wake time (DS: 05:57±64min; NS: 12:52±62min) and working their normal shift, participants arrived at the laboratory at 08:00 (DS) or 20:00 (NS). Following a 2h acclimation period, participants completed a 24h constant routine protocol with continuous wakefulness, followed by recovery sleep 1h later. 1h into the constant routine, participants read three narrative texts (self-paced). Memory for text was tested immediately (3h awake), 1h later (4h awake), and 24h later (27h awake). Participants saw 64 recognition probes per text, which varied by whether they required memory for surface form, textbase, or event model. Discriminability indices were calculated for each level of representation. Results Mixed-effects ANOVA with fixed effects of test bout (0h, 1h, 24h after reading), level of representation, and their interaction, covariates for shift and age, and a random intercept over participants showed a significant effect for level of representation (F=32.17, p< 0.001). Memory for surface form, which was near chance levels, was worse (p< 0.05) than that for textbase and event model, which did not differ from each other (p=0.271). There was no significant effect for test bout (p>0.1). Conclusion Memory for naturalistic text was not significantly impaired by 24h TSD. Memory for surface form was at chance level throughout, as expected since it is known to be short-lived. However, memory for textbase and event model, which reflect text meaning, was retained despite the anticipated impact of TSD on associative memory. The widespread associations in texts may provide more cues to facilitate retrieval and maintain memory performance during TSD. Support (if any) CARE Fund FY22-POP-02; HSSA