Bad sleepers’ night sleep quality improves after pre‐sleep cognitive training

Objectives/Introduction: Recently, we have reported that a complex ecological learning task (requiring the simultaneous activation of several cognitive functions), intensively administered at bedtime, improves daytime sleep continuity and stability, possibly as a result of ongoing memory processes1. Here, our aim is to extend these results to night time sleep in a sample of young adults with subjective sleep complaints. Methods: The sleep of 18 subjects (27.9 ± 8.0 years), identified as bad sleepers (PSQI global score ≤5) was recorded in three experimental conditions: (1) baseline undisturbed sleep (BL); (2) post “active‐ control” sleep (AC), i.e. a sleep episode preceded by a nonlearning control task, (tablet‐based, similar to the Psychomotor Vigilance Task); (3) post “training” sleep (TR), preceded by a complex learning task, i.e. a modified version of the word‐game Ruzzle, in which subjects have 1.5 min to form (by touching an iPad screen) as many words as possible with the 16 letters available on a 4 × 4 grid. At morning awakening from TR, subjects were retested at the cognitive task. Results: Post‐training sleep showed significantly decreased frequencies of arousals (F = 3.54, p = 0.04; BL>TR), state transitions (F = 5.05, p = 0.01; BL>TR and AC) and functional uncertainty (FU) periods (F = 6.24, p = 0.005; BL>TR and AC) as well as significantly less time spent in FU periods (F = 5.22, p = 0.01; BL>TR and AC). Awakening frequency was also reduced in TR relative to BL (total awakenings: F = 3.1; p = 0.05; BL>TR; short awakenings: F = 3.63, p = 0.04; BL>TR and AC). Finally, TR showed an increase in the number and mean duration of NREM‐REM cycles (respectively F = 9.25, p = 0.001; TR>BL and AC; F = 5.99, p = 0.006; TR>BL and AC), and longer time spent in cycles (F = 9.84, p < 0.001; TR>BL and AC). In TR, retest score at awakening was significantly higher than pre‐sleep score (t = 5.048, p < 0.001). Conclusions: In our sample of bad sleepers, night sleep in TR was notably more continuous, stable and organized than baseline sleep, confirming previous results obtained on a daytime nap1. These findings challenge the assumption that pre‐sleep cognitive activity impairs sleep quality and suggest to further explore cognitive training as a strategy to improve sleep in sleep‐disordered individuals. References: 1. Arzilli C. et al. Behav Sleep Med (2018) 25:1–9. Disclosure: Nothing to disclose.