Sleep Talking as Dream Enacting Behavior: a new perspective to study cognition during sleep

Introduction: Sleep Talking (ST) is characterized by the production of unaware linguistic vocalizations (STs) during sleep. On the one hand, ST could allow researchers the direct observation of cognitive processes during sleep. Interestingly, recent findings reported the replay of verbal material learned during an episode of ST. This phenomenon was consid- ered as an “overt replay” of high cognitive processes and may help sleep- related memory consolidation. However, data on this issue are still lacking, and the role of ST in memory consolidation is poorly understood. On the other hand, the vocal activations also showed the incorporations of waking experiences. Besides, some studies reported a high concordance between vocalizations and oneiric contents (Dream Enactment Behavior), providing a potential access to mental activity during sleep. It is known that the waking-life experiences could be incorporated according to the day-residue effect or dream-lag effect. Therefore, the incorporations of the verbal task or daily experiences could represent a reprocessing of semantic and/or autobiographical memories. Overall, the STs might permit the direct observation of these cognitive processes ongoing. Our explorative study had two independent aims: 1. The investigation of autobiographical incorporations into dream reports and STs. According to the “continuity hypothesis,” the incorporations into STs could represent “day-residue effect” or “the dream-lag effect”; 2. The assessment of ST's impact on memory consolidation, according to two alternative hypotheses: (A) the replay of verbal content on STs in- creases the sleep-dependent gain (defined as the difference between morning and evening recall) in the ST group, or (B) the sleep fragmentation due to STs is associated to a decreased gain. Materials and Methods: We recruited N¼28 participants with ST (F¼23; age mean: 23.71) and N¼27 controls (F¼21; age mean: 24.44). For eight days, participants performed home monitoring. They were instructed to complete daily logs (for seven evenings), sleep logs, and record their oneiric contents every morning. On the 8th day, a word-pair task was administered. ST subjects audio-recorded their vocal activations. Results: Results showed a higher gain in the control than ST group (t¼ 2.103; p¼0.04), but no significant correlation was observed between the number of STs and gain. Notably, one ST subject produced a word semantically related to the task and revealed an increase in the gain. The incorporations of wake-experiences in dream content revealed a dream lag effect of personally significant events in both groups (F¼ 3.510; p¼0.04). There was no correspondence between the daily activities, dream content, and STs, although some semantic correspondences were observed between STs and dream recalls (N¼4). Conclusions: In conclusion, our results supported the idea that STs could represent the overt expression of semantic memories consolidation. The worst performance of the ST group seems coherent with the hy- pothesis that sleep fragmentation due to STs is associated with a decreased sleep-dependent gain, although a polysomnographic assessment is needed. Consistently with literature, STs represent a window towards the dream activity, and multiple awaking protocols immediately after STs could provide more information about correspondence between STs and dream reports and the reprocessing of autobiographical memories