Assessing large spontaneous and sensory-evoked sleep slow waves and their relation to dreaming

This project addresses the challenging and longstanding question of how the
brain generates dreams. Although dreams have interested humankind since the
earliest ages, the precise mechanisms underlying dreaming and their function
are still unknown. Understanding dreaming not only serves to satisfy scientific
curiosity. It will also pave the way for a better understanding of sleep
disorders that are related to abnormal forms of mental activity during sleep.
Indeed, while most people hardly remember any mental experiences upon
awakening, and besides the occasional nightmare, do not consider dreams as
problematic, for up to 5% of the population, dreams can become a real burden
Complaints related to mental activity during sleep include, amongst others,
recurrent nightmares; parasomnias, representing abnormal, often dream-related
behavior; epic dreaming disorder, a condition in which dreams are perceived as
excessive and tiring, and paradoxical insomnia, characterized by the subjective
impression of feeling awake during sleep and relentless ruminations (recurrent
thoughts). In addition, a better understanding of dreaming may shed light on
related phenomena including hallucinations and delusions encountered in
psychiatric disorders.

As a highly subjective phenomenon that is confined to the realm of sleep1,
dreaming has remained inaccessible to direct scientific investigation for a
long time. It was initially assumed that dreams occurred almost exclusively
during rapid eye movement (REM) sleep, a sleep stage that owes its name to the
brisk eye movements that occur under closed eye lids, and in which brain
activity, measured with electroencephalography (EEG), is fast and surprisingly
wake-like. However, later studies established that dreams can also occur in
Non-REM (NREM) sleep, a sleep stage with very different EEG patterns,
consisting in prominent low-frequency oscillations called slow waves. In a
recent study, we investigated the neural correlates of dreaming using
high-density (hd-)EEG, combined with over 800 serial awakenings to collect
dream reports6. We discovered a neural signature of dreaming shared by both REM
and NREM sleep, characterized by a local activation of posterior brain regions
(grouped under the name *posterior hot zone*). We were also able, for the first
time, to image brain activations corresponding to specific dream contents in
full-fledged sleep, including faces, movement, speech, and the spatial setting
of dreams. Although this landmark study provided a potential answer to the
longstanding question of why dreaming can occur in behavioral states with
globally different EEG patterns (by showing that dreaming requires a local
activation, irrespective of the EEG in the rest of the cortex), it remains
unknown which neurophysiological processes underlie this neural signature of
dreaming, and how the contents of dreams are generated.

Intriguingly, in a series of subsequent studies, we found that dreaming is
closely related to two types of brain waves, which will be referred to here as
type I and type II potentials. More specifically, we found that a dream was
particularly likely to be reported when in NREM sleep, a large and widespread
slow wave (type I potential) appeared in the EEG recording shortly before the
awakening, and when, at the same time, slow waves in the posterior hot zone of
the dreams (type II potentials) were particularly small. The two types of slow
waves display distinct variations across the night, occur on different EEG
backgrounds, induce specific EEG changes, are differentially affected by
development and experience. Recent studies have documented two similar types of
slow waves also in rodent NREM sleep and even suggested that they serve
distinct sleep-related functions (memory consolidation vs. forgetting).
Although slow waves are a typically hallmark of NREM sleep, they may also occur
in REM sleep, albeit with a smaller amplitude. Interestingly, we recently
discovered that also in REM sleep there are two types of slow waves (called
sawtooth waves and medio-occipital slow waves), with similar properties to
those in NREM sleep. Finally, brain waves similar to type I potentials can be
seen during wakefulness, in response to unexpected sensory stimuli. Taken
together, these observations suggest that the neurophysiological processes
underlying the two types of slow waves could be present in different
behavioural states (REM and NREM sleep).

The overarching aims of this proposal are 1) to identify the neurophysiological
processes underlying type I slow waves and 2) to manipulate them in order to
determine whether and how they relate to dreams.
The specific aims are outlined below:
Primary Objectives
1) First part of the study: provide a detailed cortical mapping and comparison
of spontaneous and sensory-induced type I brain potentials
2) Second part of the study: to determine how spontaneous type I brain
potentials relate to dream contents
Secondary Objective(s):
1) To compare type I potential *equivalents* across behavioural states (wake,
NREM and REM sleep)
2) To determine whether type I brain potentials are associated with specific
patterns of muscle activations and autonomic nervous system changes

This is an interventional study with a within-subject design:
In the first part of the study, stimulus-induced type I brain potentials
(resulting from the administration of sensory stimuli, i.e. the intervention)
will be compared with naturally occurring type I brain potentials during sleep
within the same subject and night of sleep.
In the second part of the study, the intervention will consist in awakening
study participants at different times of the night and asking them about dreams
and their characteristics. The characteristics of type I potentials preceding
the awakenings will be compared between different *outcomes* (presence of
dreaming, sensory modality of the dream, surprise in the dream, etc.).

Sensory stimuli will be applied during wake and sleep - including visual,
auditory, and tactile stimuli. These stimuli are well within the comfortable
range, and can be applied during sleep without disturbing the participant.

The risks associated with this study protocol are minimal for the study
participants and essentially consist in unlikely allergic reactions to the EEG
net or electrode gel, fatigue and sleepiness the day after the study due to
slight sleep fragmentation, and in susceptible individuals, occurrence of a
migraine, seizure or hypomanic episode. Measures have been taken to prevent or
minimize these risks, including thorough screening of participants, adequate
information and appropriate instructions. There is not direct benefit for
participants.