Out-of-body experiences in relation to lucid dreaming and sleep paralysis: A theoretical review and conceptual model

Out-of-body experiences (OBEs) are characterized by the subjective experience of being located outside the physical body. Little is known about the neurophysiology of spontaneous OBEs, which are often reported by healthy individuals as occurring during states of reduced vigilance, particularly in proximity to or during sleep (sleep-related OBEs). In this paper, we review the current state of research on sleep-related OBEs and hypothesize that maintaining consciousness during transitions from wakefulness to REM sleep (sleep-onset REM periods) may facilitate sleep-related OBEs. Based on this hypothesis, we propose a new conceptual model that potentially describes the relationship between OBEs and sleep states. The model sheds light on the phenomenological differences between sleep-related OBEs and similar states of consciousness, such as lucid dreaming (the realization of being in a dream state) and sleep paralysis (feeling paralyzed while falling asleep or waking up), and explores the potential polysomnographic features underlying sleep-related OBEs. Additionally, we apply the predictive coding framework and suggest a connecting link between sleep-related OBEs and OBEs reported during wakefulness.


Introduction
Out-of-body experiences (OBEs) are a type of multisensory experience in which individuals feel as if they were outside of their physical body, occasionally seeming to observe the world from an elevated and extracorporeal perspective (Blackmore, 1988;Irwin, 1988;Levitan et al., 1999;Blanke et al., 2004).These experiences have been documented in various cultures throughout history and are estimated to occur in up to 5-10 % of the healthy population (Blanke et al., 2015).Despite their prevalence, OBEs remain a poorly understood and minimally explored phenomenon.
OBEs tend to appear spontaneously, although some healthy individuals have reported the ability to induce them volitionally, and particularly during low-vigilance periods that occur near the onset of sleep or during dreaming (Fox, 1962;Tart, 1967;Monroe, 1971;Muldoon and Carrington, 1974;LaBerge, 1986).Within these states, OBEs have been associated with two specific sleep experiences: lucid dreaming episodes, where individuals are aware of being in a dream state (Noreika et al., 2010); and sleep paralysis episodes, where individuals feel unable to move while falling asleep or waking up (Denis et al., 2018).
The connection between OBEs and lucid dreams or sleep paralysis has been acknowledged in various sources, including the scientific literature (Blackmore, 1988;Irwin, 1988;Levitan et al., 1999;Cheyne and Girard, 2009;Raduga et al., 2020), self-reports made by OBE experiencers (Fox, 1962;Monroe, 1971;Muldoon and Carrington, 1974;LaBerge, 1986) and descriptions found in online guides or forums.Some individuals have even suggested that passing through bodily paralysis may be necessary or at least contribute to achieve OBEs during sleep (Fox, 1962;Monroe, 1971;Muldoon and Carrington, 1974), or that sleep paralysis could become an effective doorway to induce lucid dreams or OBEs (Conesa, 2002;Emslie, 2014;Mallett et al., 2022).Therefore, exploring the occurrence of OBEs during sleep and dreaming may be critical in our journey towards understanding the overall OBE phenomenon.
The main objective of this paper is to establish a theoretical connection between OBEs and sleep/dreaming states.To achieve this aim, we will first address one major issue found in OBE research (Section 2): the wide-ranging phenomenological variability of OBE reports (for reviews see Bünning and Blanke, 2005;Carruthers, 2018).This variability may impact research conclusions and OBE classifications, as both different experiences and varying degrees of the same experience may be grouped together as the same phenomenon.Consequently, we will propose a definition of OBEs that considers only those phenomenological traits common to all OBEs: the subjective experience of being located outside one's physical body (Section 2).We will subsequently present a new conceptual model to describe OBEs happening during sleep/dreaming (Sections 3-5).
Our primary working hypothesis is that maintaining consciousness during transitions either from wakefulness to dreaming-a state of "conscious falling sleep" already acknowledged by Levitan et al. (1999)-or from dreaming to wakefulness, may facilitate sleep-related OBEs, sleep paralysis episodes, and lucid dreams.With this hypothesis in mind, our initial focus will be on examining the body of research connecting OBEs to dreams and wake-sleep transitions (Section 3).Next, we will offer a descriptive analysis of the different phenomenological dimensions that can be used to discriminate between OBEs triggered during sleep and similar states of consciousness such as sleep paralysis and lucid dreaming (Section 4).Once this distinction is made, we will define the theoretical framework underlying the Sleep-onset REM mental events continuum and Sleep-related OBE spectrum (Section 5).The former continuum investigates the sequential occurrence of sleep paralysis, OBEs and lucid dreams during transitions from wakefulness to dreaming.The latter spectrum proposes that sleep paralysis, OBEs and other unusual bodily sensations reported during sleep may collectively constitute a singular phenomenon, manifested at different degrees of intensity.While these two spectra are highly speculative and require empirical validation, they have the potential to help future research studies to identify and predict sleep-related OBEs.

A common definition and classification of out-of-body experiences
The study of OBEs has always been challenging due to the uncommon and spontaneous nature of the phenomenon, as well as its mystical, religious, and parapsychological aspects (Irwin, 1988;Bünning and Blanke, 2005).Research during the 20th century was mainly based on surveys and lists of compiled and mostly retrospective experiences, with only a few authors carrying out electrophysiological (EEG) and induction studies (Tart, 1967(Tart, , 1968;;Palmer and Lieberman, 1975;McCreery andClaridge, 1996a, 1996b).In 2002, Blanke et al. showed that OBEs could be induced by electrical stimulation of the right temporo-parietal junction (TPJ) in a patient undergoing presurgical epilepsy evaluation (Blanke et al., 2002).Among other brain functions, the right TPJ is known to be involved in attentional reorienting, theory of mind and cognitive empathy (Chang et al., 2013;Krall et al., 2015;Mai et al., 2016).The TPJ is also considered a key region for vestibular processing, multisensory integration and body perception (Blanke and Mohr, 2005).Later studies confirmed the association of OBEs with the TPJ (Blanke et al., 2004(Blanke et al., , 2005;;De Ridder et al., 2007;Smith and Messier, 2014) and additionally linked OBEs to motor imagery, cortical hyperexcitability, cervical spinal cord damage (Overney et al., 2009;Braithwaite et al., 2013;Smith and Messier, 2014;Milne et al., 2019), as well as to other brain regions such as the cerebellum, the anterior insular cortex and the posterior cingulate cortex (De Ridder et al., 2007;Smith and Messier, 2014; K. Yu et al., 2018;Hiromitsu et al., 2020).
However, this recent neuroscience research presents some limitations.One is the wide range of variability found in OBE reports, which makes it difficult to establish clear-cut defining criteria for OBEs.For instance, OBE reports can be accompanied by emotions ranging from fear to joy (Twemlow et al., 1982;Blackmore, 1984), and may or may not involve visual hallucinations (Cheyne and Girard, 2009).If visual content is present during the experience, the scenery depicted may either mirror the individual's current physical setting (e.g., one's bedroom) or portray entirely unfamiliar surroundings (Tart, 1967(Tart, , 1968;;Twemlow et al., 1982;Blackmore, 1984;Blanke et al., 2004;Rabeyron and Caussie, 2016).The context and state of mind in which OBEs arise is also diverse.For example, OBEs can be triggered during life-threatening events, near-death experiences and episodes of impaired consciousness, but also in fully awake and conscious individuals (for review see Bünning and Blanke, 2005).The populations targeted to investigate OBEs also differ, including neurological patients with brain lesions or seizures (Blanke et al., 2002(Blanke et al., , 2004)), healthy individuals who self-reported the ability to induce OBEs at will (Tart, 1967(Tart, , 1968;;Smith and Messier, 2014), and healthy individuals who are susceptible to hypnotic suggestion (Zeev-Wolf et al., 2017;Facco et al., 2019).Likewise, a wide variety of OBE definitions have been used to describe OBEs to experimental subjects in different studies.For instance, Twemlow et al. (1982) defined OBEs as "an experience where you felt that your mind or awareness was separated from your physical body", whereas Milne et al. (2019) conceptualized OBEs as "an experience where you have perceived/experienced the world from a vantage point ouside of the physical body".Finally, methods employed by researchers to gather and analyse phenomenological reports of OBEs also present significative differences, with some researchers characterizing the OBE phenomenology through either semi-structured interviews or questionnaires that can be collected right after the experience (Blanke et al., 2004;Facco et al., 2019) or years after the experience (Twemlow et al., 1982;Blackmore, 1984;Rabeyron and Caussie, 2016).
Overall, the variability of OBE reports and the methods used to collect them can lead to a misrepresentation of past results.For example, different studies may be dealing with diverse but interrelated experiences, or with different levels or subtypes of the same experience.Take as an example the OBE-like experience described by Smith and Messier (2014), which was named "extra-corporeal experience" to distinguish it from other OBEs reported in the literature.Similarly, Zeev-Wolf et al. (2017) distinguished between "high-OBEs" and "low-OBEs" to refer to the level of self-reported OBE success achieved by hypnotizing their study participants.Interestingly, these two groups (high versus low OBEs) showed differences in brain activity as measured by magnetoencephalography, including graded (or dose-response) relationships (Zeev-Wolf et al., 2017).In this context, we can formulate the following questions: 1) Do we need a more specific classification of OBEs to fully understand this phenomenon?2) If we decide to categorize OBEs more specifically, should these categories be kept completely separated from each other-"OBE" versus "extra-corporeal experience" in Smith and Messier (2014); or rather be considered as part of a continuum of hallucinatory experiences-"high-OBE" versus "low-OBE" in Zeev-Wolf et al. ( 2017)?
In any case, if more work is to be done on either of these classifications, the necessary and sufficient condition(s) that any experience must present to be classified as an "OBE" need to be defined first.In line with Carruthers (2018), we will use the term "core OBE" to refer to this concept, which includes only those phenomenological features that are common to all OBEs.
We propose to define such a core OBE as the subjective experience of being located outside one's physical body (extracorporeal self-location 1 ).The definition proposed here excludes states that involve certain OBE-like sensations (e.g.: floating sensations), but where the subject still reports feeling located inside their physical body (for example, see pg. e59 in Rabeyron and Caussie, 2016).Thus, this definition only takes into account individuals who feel they are situated in a particular or abstract location that is described as occurring outside of their physical body, independent of whether they experience any visual content-such as seeing the world from an elevated perspective (Bünning and Blanke, 2005)-or not.An example of the latter are "out-of-body feelings", where individuals feel located outside of their physical body while describing no visual content (De Ridder et al., 2007;Cheyne and Girard, 2009;Terhune, 2009;Lopez et al., 2010;Braithwaite et al., 2011).
Depersonalization episodes would be further excluded from this definition.These are defined as feelings of being detached from one-self according to the DSM-V (American-Psychiatric-Association, 2013), and might be mistaken for OBEs if not analysed in more detail.However, OBEs and depersonalization episodes differ in three key aspects: 1. Individuals during depersonalization episodes experience themselves-or certain parts of themselves, such as any emotions or body parts-as unfamiliar or detached from the subjective self.Thus, identification of the self with one's physical body or emotions is disturbed during depersonalization (American--Psychiatric-Association, 2013).In addition to this, people during OBEs experience their entire self as if located outside of their physical body-i.e., both self-location and self-identification with the physical body are disturbed during OBEs (Irwin, 1988;Bünning and Blanke, 2005).2. Depersonalization episodes are commonly reported along with feelings of unreality (derealization), whereas OBEs are most likely described as hyper-realistic experiences (Tart, 1967;Twemlow et al., 1982;Levitan et al., 1999;De Ridder et al., 2007;American-Psychiatric-Association, 2013;Rabeyron and Caussie, 2016).3. OBEs can appear in the context of depersonalization/derealization disorder along with other symptoms, but experiencing extracorporeal self-location is not a necessary condition for the diagnosis of this disorder (American-Psychiatric-Association, 2013).
The reader should note that, when not using the appropriate questions, some depersonalization episodes may be classified as either "OBEs" or "out-of-body feelings", and vice versa.This must not be overlooked when providing a definition of OBEs or depersonalization to study participants that might not be aware of this distinction. 2Finally, another phenomenon that can cause confusion in this regard are episodes of "ego-dissolution" (for review see Millière, 2017) or experiences that partially or completely lack a sense of body boundaries or embodiment (Ataria et al., 2015;Dor-Ziderman et al., 2016;Alcaraz-Sánchez et al., 2022;Campillo-Ferrer et al., 2023), such as the experience of "merging" with one's surroundings.While these experiences can be compared to OBEs where individuals report feeling as if they were "evaporated" (see Footnote 1), these features do not necessarily appear together with the subjective experience of being located outside one's physical body.As an example, see Campillo-Ferrer et al. (2023), where Subject 3 reported feeling as if "not having a body", yet the feeling of self-location during this experience remained to be inside of the physical body.
We argue that the examples provided above should all be considered to develop a more comprehensive classification and definition of OBEs.Establishing extracorporeal self-location as the necessary and sufficient condition of OBEs, and additionally using variable and non-necessary components of the experience to either describe different types of OBEs, or a scale with different levels, might constitute a promising path to follow.This approach can be used to distinguish feelings of extracorporeal self-location that occur in combination with seeing one's own body (OBE with autoscopy) or not (OBE without autoscopy).In line with this, different phenomena like experiencing floating or flying sensations might be considered associated experiences that can either be placed or not inside of a continuum of hallucinatory experiences (Out-of-body spectrum).This idea will be implemented in Section 5 to characterize OBEs occurring during sleep, as well as other related phenomena.To this aim, we will next review the literature linking OBEs to sleep and dreaming.

Sleep-onset REM periods may set the stage for out-of-body experiences during sleep and dreaming
A majority of OBEs occurring within the healthy population are reported to originate from low-vigilance states such as resting, dozing, or transitional states such as falling asleep or waking up (Twemlow et al., 1982;Blackmore, 1984;Braithwaite et al., 2011).During these transitional periods, sleep-wake hallucinations may occur, including hypnagogic hallucinations (while falling asleep) and hypnopompic hallucinations (while waking up) (Waters et al., 2016;Ghibellini and Meier, 2023).As such, some authors have speculated that certain types of OBEs may be part of this set of sleep-wake hallucinations, building the hypnagogic model of OBEs (for review see Terhune, 2009).OBEs have also been reported during vivid and emotional dreaming states, which are, in turn, more likely to appear during the rapid eye movements (REM) sleep phase (Aserinsky and Kleitman, 1953;Carr and Solomonova, 2019).For instance, 35 % of participants in one study reported having been dreaming while experiencing an OBE (Twemlow et al., 1982).In a separate study, OBEs were also classified by 49 % of participants under 1 This phenomenon is typically referred to as disembodiment in the literature (Blanke and Arzy, 2005;Blanke et al., 2005;Bünning and Blanke, 2005).In this paper, we have decided to employ the term 'extracorporeal self-location' to mitigate potential misconceptions that can arise from the term 'disembodiment'.Thus, we define disembodiment as the subjective sensation of not being embodied.It is important to note that this sensation is not always reported in the case of OBEs; some individuals describe what we interpret as a state of 'distorted embodiment' where they perceive themselves as a "sphere of light", a "circular thing", a "point in space", or as if being "evaporated", "gaseous", "fluid-like" or devoid of limbs (McCreery and Claridge, 1996a;Tressoldi et al., 2015;Rabeyron and Caussie, 2016).Other subjects have reported sensations of "melting" or feeling as if the body were "empty" prior to the OBE onset (LaBerge, 1986;Devinsky et al., 1989).
2 Here we do not argue against the possibility of a relationship between OBEs and depersonalization.Our primary intention is to emphasize the necessity for more precise definitions and classifications of OBEs.In fact, we encourage future research to explore the potential connection that may exist between OBEs and depersonalization.The studies conducted by Devinsky et al. (1989) and Lopez and Elzière (2018) can serve as valuable inspirations for achieving this goal.The former documented cases where depersonalization and detachment preceded OBEs and autoscopy in the context of seizures (Devinsky et al., 1989).The latter study identified depersonalization-derealization as key predictors of OBEs in both patients experiencing dizziness symptoms and healthy subjects (Lopez and Elzière, 2018).
T. Campillo-Ferrer et al. the category "like dreaming" and were associated with dream vividness, flying dreams and lucid dreams (Blackmore, 1984).
Alternatively, sleep-related OBEs may be linked to atypical, transitional or mixed states of sleep, such as the so-called "dissociated REM states".This latter term arises from the Theory of Dissociated States of Mind proposed by Mahowald and Schenck (2005), which suggests that REM and non-REM sleep may differ but are not strictly mutually exclusive.Thus, there are instances where REM and non-REM sleep can overlap, forming dissociated REM states (Mahowald and Schenck, 2005).The idea that different sleep phases may overlap, potentially leading to dissociations, can in turn be linked with that of "local sleep": the notion that sleep oscillations characterizing each sleep phase do not necessarily affect broad cortical networks, but only a set of specific brain regions (Siclari and Tononi, 2017;Krueger et al., 2019).Thus, an individual might seem physiologically awake while some neuronal populations are nonetheless exhibiting distinctive traits of being locally asleep (Nobili et al., 2012;Andrillon et al., 2019).
Dissociated REM states may potentially occur more frequently in individuals who have also experienced OBEs in the past.In this context, an interview-based study showed that individuals who have had neardeath experiences (N=55) are more likely to experience REM sleep intrusions when they have also experienced OBEs, either when these OBEs were reported during sleep or in the context of near-death experiences (Nelson et al., 2007).In the latter study, REM sleep intrusions were defined as sleep features that are characteristic of REM sleep-such as hallucinations or paralysis of the major muscles-but which are present during an apparent state of wakefulness (Nelson et al., 2006(Nelson et al., , 2007)).
Following this reasoning, sleep-related OBEs might just be another type of dissociated REM state presenting EEG features of both wakefulness and REM sleep.This idea has been suggested by several authors (Mahowald et al., 2011;Raduga et al., 2020), but with current literature we are not yet able to confirm this assumption.In particular, EEG and polysomnographic studies on sleep-related OBEs show contradictory results.This could be explained by the fact that they were not performed with state-of-the-art methodology, and the phenomenology of OBEs was not clearly defined in these studies (for review see Bünning and Blanke, 2005).Consequently, it is challenging to attribute a specific EEG pattern or sleep stage to this experience.
Despite the limited empirical evidence, some studies still suggest an indirect connection between OBEs and dissociated REM states.Specifically, OBEs have been linked to several sleep phenomena (such as sleep paralysis and lucid dreaming), which have been in turn associated with sleep-onset REM periods-i.e., episodes of REM sleep that exhibit an uncommonly short latency (Singh et al., 2006).Thus, while regular REM sleep emerges after a relatively long episode of non-REM sleep, individuals during sleep-onset REM periods can either enter the REM state directly from wakefulness or after a very brief period of non-REM sleep (see Fig. 1).Sleep-onset REM periods may lead to ambiguous polysomnographic recordings combining elements from both wakefulness and REM sleep (Takeuchi et al., 1992), a state that could be considered as one of "dissociated REM sleep".
Sleep-onset REM periods are present in the healthy population (Bishop et al., 1996), but these are most frequently reported by individuals who suffer from narcolepsy-a sleep-wake disorder characterized by involuntary sleep attacks and sudden loss of muscle tone or cataplexy (Akintomide and Rickards, 2011;Kornum et al., 2017)and/or by individuals who experience sleep paralysis or lucid dreams (Vogel, 1960;Hishikawa and Kaneko, 1965;Takeuchi et al., 1992Takeuchi et al., , 2002;;Levitan et al., 1999;Gott et al., 2021;Mainieri et al., 2021).In line with this, hybrid states with features of both wakefulness and REM sleep have been observed in EEG recordings obtained from narcoleptic patients and in the context of sleep paralysis and lucid dreaming (Hishikawa and Kaneko, 1965;Takeuchi et al., 1992Takeuchi et al., , 2002;;Voss et al., 2009;Terzaghi et al., 2012;Y. Liu et al., 2015;Mainieri et al., 2021).
In the specific case of narcolepsy, the presence of sleep-onset REM periods is nowadays used as a criterion for the diagnosis of this disorder (Vogel, 1960;Singh et al., 2006).Interestingly, patients with narcolepsy also tend to experience lucid dreams more frequently (Dodet et al., 2015;Rak et al., 2015), and can furthermore present sleep paralysis and sleep-wake hallucinations as a symptom (Akintomide and Rickards, 2011;Kornum et al., 2017).On the other hand, sleep paralysis can additionally appear out of the context of narcolepsy (named "isolated sleep paralysis", see Takeuchi et al., 2002) and has been linked to higher susceptibility to experience sleep-related OBEs and lucid dreams (Cheyne and Girard, 2009;Denis and Poerio, 2017;Raduga et al., 2020).These periods of "conscious loss of postural tone" (or sleep paralysis) can be accompanied by auditory, visual and vestibular hallucinations (Denis et al., 2018) and are often reported during sleep-onset REM periods in both narcoleptic patients (1 out of 1 episodes in Terzaghi et al., 2012; and 4 out of 5 episodes in Mainieri et al., 2021) and healthy participants (5 out of 6 episodes in Takeuchi et al., 1992; and 6 out of 8 episodes in Takeuchi et al., 2002).
Research on lucid dreaming also supports the hypothesis that OBEs are associated with sleep-onset REM periods.In this regard, a study found that OBE content appears more frequently during wake-initiated lucid dreams compared to other types of lucid dreams (Levitan et al., 1999).Wake-initiated lucid dreams are defined as lucid dreaming episodes in which subjects remain conscious during direct transitions from wakefulness to REM sleep (LaBerge et al., 1986;Levitan et al., 1999).Thus, individuals experiencing wake-initiated lucid dreams retain consciousness throughout the falling asleep process and become aware of the subsequent dream state (necessarily experiencing a sleep-onset REM period, with a wake-REM transition lasting less than 2 minutes).In the aforementioned study, wake-initiated lucid dreams were contrasted with dream-initiated lucid dreams in particular, where subjects become conscious and aware of dreaming once they have reached REM sleep (not necessarily experiencing a sleep-onset REM period, with a wake-REM transition lasting more than 2 minutes) (Levitan et al., 1999).
In this context, it should be emphasized that the sentence "I felt myself leaving my body" was only reported at the onset of lucid dreams in this study (Levitan et al., 1999), suggesting that this specific sensation may arise from transitions to REM sleep.
The hypothesis that OBEs may be linked to sleep-onset REM periods is further supported by another study describing the polysomnographic features of an individual who self-reported the ability to experience OBEs during sleep (Tart, 1968).During the first experimental session, the participant did not experience any OBEs, but entered REM sleep directly from wakefulness.Thus, it could be speculated that her ability to experience spontaneous sleep-onset REM periods could have also led her to experience OBEs more frequently.
Considering the reviewed literature, we hypothesize that the presence of sleep-onset REM periods in the polysomnographic recordings can set the stage for some episodes of sleep paralysis, wake-initiated lucid dreams and sleep-related OBEs.In this regard, Levitan et al. (1999) proposed that sleep-related OBEs may appear under the same physiological conditions as wake-initiated lucid dreams: rapid transitions from wakefulness into REM sleep, coupled with increased levels of cerebral activation and meta-cognition, a state that can be considered as one of "conscious falling asleep".Thus, a direct transition from waking into REM sleep without losing consciousness in the process could lead the subject to experience a state of conscious loss of postural tone (muscle atonia).This might explain why sleep paralysis has been associated with OBEs and lucid dreams for so long (Fox, 1962;Monroe, 1971;Muldoon and Carrington, 1974;LaBerge, 1986;Blackmore, 1988;Irwin, 1988;Levitan et al., 1999;Cheyne and Girard, 2009;Raduga et al., 2020).Moreover, a conscious wake-REM transition may also contribute to the emergence of common hallucinations in all of these experiences, a phenomenon that has previously been remarked by several authors (LaBerge, 1986;Blackmore, 1988;Irwin, 1988;Levitan et al., 1999).Examples of these hallucinations are: • Auditive hallucinations, such as hearing rushing sounds.
• Unusual bodily experiences, including: -Feeling immobility, body distortion, numbness or electrical sensations.-Having specific vestibular-motor hallucinations, such as experiencing motion, lack of gravity or strong bodily vibrations.
Regarding the latter, floating, flying, falling, lifting and spinning hallucinations have all been reported during sleep paralysis episodes together with the sensation of a violent extraction of the self from the body (Cheyne et al., 1999;Cheyne, 2003;Cheyne and Girard, 2009).These are similar to floating, flying or falling hallucinations and sensations of being "pulled" or "pushed" out from the physical body reported at the onset of sleep-related OBEs (Fox, 1962;Monroe, 1971;Muldoon and Carrington, 1974;McCreery and Claridge, 1996a) and OBEs or unusual bodily experiences triggered by brain stimulation (Blanke et al., 2000(Blanke et al., , 2002(Blanke et al., , 2004)).Floating, flying and falling sensations are also typical dream themes (Schönhammer, 2000), although these have traditionally been associated with lucid dreams and sleep-related OBEs (Blackmore, 1984;Stumbrys et al., 2014).Thus, imagining similar actions such as levitating, falling down, rolling over or head rotation have been used to induce lucid dreams and sleep-related OBEs (Raduga, 2021).An example of the latter is the "roll-out" method, where OBE practitioners must imagine themselves as if turning over in bed in order to "go out of the body" (Tart, 1967).
In conclusion, maintaining consciousness during transitions from wakefulness to REM sleep may initiate the physiological conditions leading to certain types of OBEs, sleep paralysis and lucid dreams, as well as to shared hallucinations experienced in these different states.A key question that may emerge in this context is: if sleep-related OBEs, sleep paralysis episodes and lucid dreams all occur under similar circumstances, how can we distinguish between these different phenomena?In response to this question, some authors in the 20th century argued that the similarities between OBEs and lucid dreams were sufficient to classify sleep-related OBEs as merely a specific type of lucid dream (LaBerge, 1986;Levitan et al., 1999).However, this idea is inconsistent with more recent literature.While a majority of early studies showed a statistically significant association between OBEs and lucid dreams, a meta-analysis carried out in 1988 indicated a relatively weak association between these two phenomena (combined phi coefficient = 0.20, see Irwin, 1988).Nevertheless, this association only refers to OBEs in general, and not to sleep-related OBEs specifically.
Recently, a new study found a significant relationship between lucid dreams and OBEs appearing exclusively during sleep (Raduga et al., 2020).This study showed that regular lucid dreamers are more likely to often experience sleep-related OBEs (24 %) when compared to non-lucid dreamers (10 %).Moreover, 41 % of regular lucid dreamers in this study had never experienced sleep-related OBEs, in contrast to 64 % of non-lucid dreamers (Raduga et al., 2020).These results suggest that lucid dreamers tend to report more sleep-related OBEs than non-lucid dreamers, but they do not support the classification of sleep-related OBEs as a specific type of lucid dream.In addition, it should be noted that these percentages were derived from a limited sample population consisting exclusively of individuals who regularly experience lucid dreams vs. individuals who never experience lucid dreams (Raduga et al., 2020).Therefore, the likelihood of experiencing sleep-related OBEs among individuals who occasionally report lucid dreams (non-regular lucid dreamers) remains unquantified.What we can infer from these data is that some lucid dreamers never experience sleep-related OBEs, just as some sleep-related OBE experiencers never experience lucid dreams.
In this discussion, we argue that sleep-related OBEs, sleep paralysis episodes and lucid dreams are distinct yet interconnected states of consciousness that may arise under similar physiological conditions (e. g., sleep-onset REM periods and potentially other dissociated sleep states, see Section 5).Accordingly, we will suggest in the following section different experiential dimensions that can be used to categorize sleep-related OBEs and other related phenomena into separate groups.

Phenomenological differences between out-of-body experiences and other related sleep phenomena
To further shed light on the differences that exist between sleeprelated OBEs and episodes of sleep paralysis or lucid dreaming, we will next offer a descriptive analysis of the main phenomenological features characterizing each phenomenon.This section of the paper will begin by comparing the conscious states reported during both lucid and non-lucid dreaming, with the aim of establishing a foundation for comparison (Section 4.1).Next, we will conduct an in-depth phenomenological examination of the various states of consciousness that can be encountered during sleep-related OBEs, lucid dreams, and episodes of sleep paralysis, with a specific emphasis on highlighting the differences existing between sleep-related OBEs and lucid dreams.Since there are different phenomenological factors that can define consciousness in dreams, we will limit our comparison to specifically describe differences between these phenomena in the following categories: lucid insight (Section 4.2), realism (Section 4.3), control, memory and thought (Section 4.4), positive and negative emotions (Section 4.5), and dissociation (Section 4.6).These categories (see Figs. 2 and 3 for visual representation) were derived from the Lucidity and Consciousness in Dreams (LuCiD) scale described by Voss et al. (2013), who identified (1) insight, (2) realism, (3) control, (4) memory, (5) thought, (6) positive emotion, (7) negative emotion and (8) dissociation as factors which best describe dream consciousness.

Consciousness in lucid versus non-lucid dreams
Non-lucid dreams are generally characterized by the lack of both logical thinking and the ability to plan things ahead, remember or reflect about the past or future ('thought' and 'memory' factors).Non-lucid dreamers seem "forced" to fully engage in the immediate present of the experience, which can be highly emotional ('positive emotion' and 'negative emotion' factors).They are also unable to control what is occurring in the dream ('control' factor) and unable to become aware of being in a dream state ('insight' factor).Sometimes, this behaviour is continued even when the dream content becomes extremely bizarre ('realism' factor).
In lucid dreams, there is a re-emergence of especially the 'insight' and 'control' factors (Voss et al., 2013).Despite 'thought' being present in non-lucid dreams, during lucid dreams it can become more creative and/or logical, and not only restricted to the immediate present (Stumbrys and Daniels, 2010;Konkoly et al., 2021).In combination with regaining memories of their waking life (Stumbrys et al., 2014;Carr et al., 2020;Mallett, 2020), lucid dreamers can conclude that they are dreaming (Windt and Metzinger, 2007;Noreika et al., 2010) and may be able to control their behaviour in the dream, or even change its plot, characters or setting (Stumbrys et al., 2013(Stumbrys et al., , 2014;;Stumbrys and Erlacher, 2017).Another two factors that are representative of lucid dreams are 'positive emotions' and 'dissociation' (e.g., third-person perspective), which have both been found significatively higher in states of lucid versus non-lucid dreaming (Voss et al., 2013;Schredl et al., 2022).With regards to 'negative emotions', only a small proportion of lucid dreams seem to be negatively toned (10 % in Stumbrys, 2021).

Lucid insight
The level and type of lucid insight reported within the experience may be used to distinguish between sleep-related OBEs and episodes of sleep paralysis or lucid dreams.Lucid insight is defined here as the ability to be aware of one's current state of consciousness (e.g., dreaming) and is derived from the 'insight' factor isolated by Voss et al.  (2013). 3This concept is linked to lucid dreams, but can also be applied to both altered and non-altered states of wakefulness (Noreika et al., 2010).Since lucid insight is not an all-or-nothing phenomenon, several authors have defined different levels to quantify this feature.For instance, Windt and Metzinger (2007) and Noreika et al. (2010) distinguished four types of lucid states that can appear while dreaming. 4hree of them (A/awareness, B/behavioural and E/emotional lucidity) are considered pre-lucid forms when occurring in isolation, and only the presence of C/conceptual lucidity-or having the thought that one is dreaming-is necessary to classify a dream as lucid.
According to these authors, an example of A-lucidity is the awareness that the environment or characters in a dream have a dreamlike or virtual appearance while the experience is taking place.Nevertheless, without being able to conceptualize the experience as a dream (Clucidity), the subject would not be able to realize or conclude that they are dreaming.Other pre-lucid forms involve the ability to behave (Blucidity) or process emotions (E-lucidity) as if the individual knew that the ongoing experience is only a dream (Windt and Metzinger, 2007;Noreika et al., 2010).Important to note, C-lucidity as defined by these authors is not necessarily present in states of wakefulness.Thus, one way to activate C-lucidity while being awake would consist on conceptualizing that our current conscious experience is only a representation of the actual physical world created by the human brain (Noreika et al., 2010).
The level of lucid insight reported by lucid dreamers has repeatedly been compared to that reported in sleep-related OBEs.However, these two must be clearly differentiated: lucid dreamers and sleep-related OBE   experiencers both share the ability to realize that they are not awake at the moment of the experience (Blackmore, 1988;Irwin, 1988;Levitan et al., 1999).In other words, they both recognize a state of consciousness that is different from the standard waking state.In contrast, while lucid dreamers conclude that they are dreaming (Windt and Metzinger, 2007), a substantial number of sleep-related OBE experiencers conclude that they are actually located outside of their physical body (for discussion see Levitan et al., 1999).Most remarkably, it is common during sleep-related OBEs to reflect upon the fact that the physical body is asleep during the experience.Even in this case, some subjects still claim to have been outside of their physical body upon waking up (Fox, 1962;Monroe, 1971;Muldoon and Carrington, 1974).
Sleep paralysis episodes can also exhibit partial lucid insight (Terzaghi et al., 2012;Jalal and Hinton, 2013;Jalal et al., 2014Jalal et al., , 2015;;Mainieri et al., 2021).By definition, individuals are aware during sleep paralysis that their physical body is paralyzed while the experience is still taking place (Denis et al., 2018).Additionally, some sleep paralysis sufferers have demonstrated the ability to signal the onset of this experience in the same way lucid dreamers do, by performing a left-right-left-right eye signal that can be visible in the polysomnographic recordings (Mainieri et al., 2021).Nevertheless, similar to sleep-related OBEs, sleep paralysis episodes may also occur together with a sense of confusion between the hallucinogenic-driven world and the real physical world, a characteristic that may persist even after waking up from this experience (Jalal and Hinton, 2013;Jalal et al., 2014Jalal et al., , 2015)).Terzaghi et al. (2012) provided a descriptive example of the latter phenomenon, presenting the case of a subject who engaged in an actual conversation with a neurologist while experiencing sleep paralysis.Once the subject was fully awake, this person was unable to determine if the conversation had actually taken place or was just a dream.This phenomenon is similar to dream-reality confusion experienced in the context of narcolepsy, which can lead narcoleptic patients to occasionally remember dream occurrences as real events (Wamsley et al., 2014).
Therefore, if we only compare the level of lucid insight described during the experience, sleep-related OBEs seem to be more akin to sleep paralysis episodes than to lucid dreams.However, this is not always the case.Next, we will present other phenomenological factors where sleep related OBEs appear to exhibit opposite characteristics to sleep paralysis episodes.An example of this is the likelihood of experiencing negative or positive emotions (see Section 4.4), where lucid dreams constitute an intermediate state between sleep paralysis (with more negative emotions) and sleep-related OBEs (with more positive emotions).

Realism
Different levels of realism within the experience may also be used to distinguish sleep-related OBEs from sleep paralysis and lucid dreaming episodes.For example, in contrast to lucid dreaming, scenes portrayed during OBEs are more likely to depict realistic representations of the environment in which subjects are physically located (58-75 % of OBE reports in Twemlow et al., 1982;Blackmore, 1984), such as the laboratory environment or their own bedroom (Tart, 1967(Tart, , 1968;;Jalal et al., 2014;Rabeyron and Caussie, 2016).In addition, the scenery of sleep-related OBEs may also seem realistic without necessarily being an exact copy of the sleeper's physical location.For example, lack or addition of pieces of furniture has been noticed in some OBE reports (for discussion see Irwin, 1988;Levitan et al., 1999), or changes in the normal visual field, such as experiencing panoramic or tinted vision (Twemlow et al., 1982;Rabeyron and Caussie, 2016).In up to 36 % of the cases, OBEs appear to be embedded in sceneries that are familiar or unfamiliar to the sleepers (Twemlow et al., 1982;Blackmore, 1984), but which do not represent the immediate individual's physical environment (for reports see Tart, 1968;LaBerge, 1986;Rabeyron and Caussie, 2016).
In this regard, several authors have suggested that increased levels of realism within the OBE environment may limit lucid insight during sleep-related OBEs (LaBerge et al., 1986;Levitan et al., 1999).According to this hypothesis, minor inconsistences within the OBE scenery or storyline might be detected by experienced lucid dreamers, who would thereby become fully aware of their dreaming state, achieving higher levels of lucid insight and disrupting the sense of reality previously experienced (LaBerge et al., 1986;Levitan et al., 1999).In this context, sleep-related OBEs might be deemed as only a type of pre-lucid dream that emerges from highly realistic dream settings.The use of dream incongruences to trigger lucidity is, in fact, a common practice.A well-known technique employed to achieve lucidity is the "reality check", "reality testing" or "reflection technique" (Tholey, 1983;Stumbrys et al., 2012;Tan and Fan, 2022), where individuals make a conscious effort to question whether they are dreaming or not multiple times throughout the day (i.e., lucid insight is cultivated during wakefulness).By developing this habit, the question of whether they are awake or asleep can eventually permeate into the dream state and individuals can actively seek for inconsistences within the dream-such as the presence of elements that defy the laws of our waking reality-to determine whether they are dreaming or not (Tholey, 1983).
Thus, the ability to recognize incongruences in the dream world could indeed become more challenging for OBE experiencers compared to lucid dreamers since the environment of sleep-related OBEs tends to present fewer unusual elements (Levitan et al., 1999).We will next discuss the literature supporting and opposing this claim, and then proceed to describe the differences between the two following terms: 1) Environmental bizarreness (i.e., the presence of unusual vs. realistic elements in the dream setting); 2) Sense of reality within the experience (i.e., to subjectively experience a dream as a real vs.dream-like event).These two concepts can act independently from each other and will help us to stablish a distinction between sleep-related OBEs and lucid dreaming that goes beyond a mere shift in the dream content or level of lucid insight.
The hypothesis that increased realistic elements within the dream environment may limit lucid insight in sleep-related OBEs is supported by two sleep laboratory studies.These studies (N=10 each) found that pre-lucid dreams are the most bizarre when compared to both lucid and non-lucid dreams.Additionally, these studies found no significant differences between lucid and non-lucid dreams (Tyson et al., 1984;Hunt and Ogilvie, 1988).These results suggest that bizarreness, conceptualized as the combination of cognitive control (e.g."confused thinking") and visual hallucinosis (e.g."out of place" visual intrusions), may play a role in triggering lucidity but not in maintaining this lucidity (Hunt, 1982;Tyson et al., 1984;Hunt and Ogilvie, 1988).
Other studies have investigated the link between dream bizarreness and lucid dreaming, leading to mixed results (for review see Gackenbach, 1988).More recently, multiple questionnaire-based studies found no significant differences in bizarreness when comparing lucid and non-lucid dreams (N=100 in Schredl et al., 2022), no significant differences in bizarreness density when analysing dream reports with "high" versus "low" lucidity scores (N=16 in Stocks et al., 2020), and significatively less bizarreness in lucid versus non-lucid dreams (N=67 in C. Yu and Shen, 2019).Therefore, it remains unclear whether low dream bizarreness (or higher realism in the dream), may lead to decreased levels of lucid insight in sleep-related OBEs, as hypothesized by several authors (LaBerge et al., 1986;Levitan et al., 1999).
Important to note, dream bizarreness in these more recent studies was tested in the context of 'lucid dreams' or 'lucidity scores', two concepts that may also involve other consciousness factors different from 'lucid insight', such as 'control' or 'thought'.Only two additional studies assessed the correlation between dream bizarreness and different levels of lucid insight specifically (Voss et al., 2013;Mallett et al., 2022).One found a significant positive correlation between lucid insight and bizarreness (N=13 in Mallett et al., 2022), while the other found no correlation between these two factors (n=151 dream reports in Voss et al., 2013).Nevertheless, neither of the latter studies compared lucid T. Campillo-Ferrer et al. versus pre-lucid states, and neither of them distinguished between environmental bizarreness and other types of dream bizarreness (for a classification see Revonsuo and Salmivalli, 1995). 5Moreover, only one of them collected objective sleep data in a controlled laboratory setting (n=117 dream reports in Voss et al., 2013).Thus, it is hard to draw solid conclusions from these results about whether increased realism of the dream environment may result in sleep experiences with decreased lucid insight (such as sleep-related OBEs) or not.
Research on sleep paralysis also challenges the notion that reduced environmental dream bizarreness could decrease lucid insight in sleeprelated OBEs.For instance, wake-sleep hallucinations that arise during sleep paralysis are often experienced in the same environment where individuals are currently sleeping-e.g., feeling an intruder presence that is located in one's bedroom or bed (Denis et al., 2018).However, instead of hindering the emergence of lucid insight, individuals who report sleep paralysis also tend to experience lucid dreams more frequently (for review, see Ableidinger and Holzinger, 2023).Additionally, sleep paralysis has been mentioned as a potential tool to induce lucid dreams by several practitioners (Fox, 1962;Monroe, 1971;Conesa, 2002;Emslie, 2014;Mallett et al., 2022).
Another sleep phenomenon that also simulates the immediate sleeper's physical environment are false awakenings, where individuals dream about waking up and engaging in daily-life activities (Buzzi, 2019).Individuals during false awakenings may eventually come to the realization that the entire event is only a dream, consequently waking up in reality or transitioning into a lucid dream (Barrett, 1991;LaBerge and DeGracia, 2000;Buzzi, 2011Buzzi, , 2019)).Interestingly, several studies have found false awakenings to be associated with sleep paralysis (Buzzi, 2019;Raduga et al., 2020), sleep-related OBEs (Raduga et al., 2020), and to occur during sleep-onset REM periods and sleep states exhibiting mixed waking-REM polysomnographic components (Mainieri et al., 2021).
Therefore, the absence of unusual or bizarre elements in the dream environment (see 'environmental bizarreness' as a sub-factor of 'realism' in Fig. 2) does not necessarily imply less lucidity, as both sleep paralysis and false awakenings may facilitate lucid dreams.Moreover, it should be noted that merely having a dream setting that accurately represents the sleeper's surroundings-such as during false awakenings-is insufficient to achieve the sense of reality described during sleep-related OBEs and sleep paralysis (see 'sense of reality' as a sub-factor of 'realism' in Fig. 2).
The distinction between 'environmental bizarreness' and 'sense of reality' becomes apparent when we compare reports obtained upon waking up from each of these experiences.On one hand, individuals during false awakenings believe that they are genuinely getting up from bed and starting their day.However, they realize upon waking up (or while still in the dream) that the experience was only a dream (Buzzi, 2011(Buzzi, , 2019)).On the other hand, some individuals during sleep-related OBEs and sleep paralysis episodes continue to maintain the belief that they have been out of their physical body (Levitan et al., 1999) or that an intruder presence was observing them in the case of sleep paralysis (Jalal and Hinton, 2013;Jalal et al., 2014Jalal et al., , 2015) ) long after waking up.
Thus, it is important not to overlook the observation that sleeprelated OBEs and sleep paralysis episodes are perceived as more genuine when compared to false awakenings or wake-initiated lucid dreams.This may indeed suggest that distinct neural mechanisms underlie these different phenomena.Later, we will speculate about the possibility that a disruption in sensory integration during sleep might lead individuals to experience illusory bodily sensations as if they were real events during sleep paralysis and sleep-related OBEs (see Section 5.2).

Control, memory, and thought
Control over the experience, memories of waking life and thinking abilities appear to be similar across sleep-related OBEs, sleep paralysis episodes and lucid dreams.Similar to lucid dreamers, individuals during sleep-related OBEs can also remember certain aspects of their waking life and exhibit increased logical thinking when compared to non-lucid dreamers.This combination enables them to exert some kind of control over the experience, such as intentionally passing through walls and other physical barriers, initiating the action of flying, or travelling to certain locations by merely using their thoughts or mental intentions (Fox, 1962;Tart, 1967Tart, , 1968;;Monroe, 1971;Muldoon and Carrington, 1974;Twemlow et al., 1982;Blackmore, 1984;LaBerge, 1986;Rabeyron and Caussie, 2016).Memories of waking life can also be present during sleep paralysis episodes, making it possible to signal the onset of this experience in a sleep laboratory setting (Mainieri et al., 2021), or to control the content of the reported hallucinations (Kliková et al., 2021).
In addition, it is worth noting that the ability to control dreams-including the ability to stop, alter or choose the dream content-is closely associated with the ability to control OBEs.In this regard, a questionnaire-based study with N=36 participants revealed that individuals who can deliberately induce OBEs, also experience high dream control in their daily lives compared to those who only experience spontaneous OBEs (Blackmore, 1986).In the latter study, all participants who reported to have experienced "deliberate OBEs" (N=14), had also experienced lucid dreams previously (Blackmore, 1986).These findings reinforce the notion that, although sleep-related OBEs and lucid dreams seem to represent different experiences with distinct phenomenological characteristics, they may also share some common neural mechanisms.

Positive and negative emotions
Another phenomenological feature that can be used to distinguish between sleep-related OBEs and sleep paralysis or lucid dreaming are positive and negative emotions reported within the experience.The majority of research studies have either linked lucid dreams to more positive emotions when compared to non-lucid dreams, or have found no differences in the emotional value of these two experiences (for review see Schredl et al., 2022).In contrast, sleep paralysis episodes are typically accompanied by intense fear and distress (Cheyne et al., 1999;Cheyne and Girard, 2009;Jalal et al., 2014;Kliková et al., 2021), especially for individuals who categorize these experiences as supernatural (Jalal and Hinton, 2013;Jalal et al., 2015).For example, a recent survey-based study with a total sample size of N=172 revealed that only 23 % of individuals are able to enjoy sleep paralysis episodes (Kliková et al., 2021).On the other hand, most individuals report positive emotions during and immediately after OBEs, with only a few describing fear (up to 33 %), sadness (12 %) or feeling like going crazy (4 %) (Twemlow et al., 1982;Blackmore, 1984).
While sleep-related OBEs, lucid dreams and sleep paralysis episodes are distinguishable by the likelihood of experiencing negative or positive emotions, they also share one common characteristic: the emotional value of these phenomena appears to be influenced by (1) the type of hallucinations reported and (2) the degree of control exerted over the experience.For instance, 87 % of individuals reporting pleasant sleep paralysis in the latter study (Kliková et al., 2021) were also able to shape the content of their hallucinations, with most of them choosing to experience flying, floating and OBEs as opposed to intruder experiences (sensing, seeing and/or hearing a threatening presence) and incubus experiences (choking sensations, breathing difficulties or pressure and pain on the chest and other body parts).Additional studies have linked unusual bodily experiences-including floating sensations and sleep-related OBEs-to positive emotions, and specifically to feelings of bliss (Cheyne et al., 1999;Cheyne, 2003).Likewise, control over experience has also been associated with positive emotions during lucid dreams, and not only in the context of sleep paralysis.In this regard, a recent study conducted a content analysis of 400 lucid dream reports extracted from the social media platform Reddit, revealing a significant association between high dream control and positive emotions in dreams (Mallett et al., 2022).Overall, these findings are in line with the idea that sleep-related OBEs, sleep paralysis episodes and lucid dreams may represent different yet closely related states of consciousness.

Dissociation
The level of dissociation within the experience may also help us to distinguish between sleep-related OBEs, sleep paralysis and lucid dreams.In this regard, numerous studies have investigated the presence of dissociative factors in lucid dreams, considering specific aspects such as6 : • Autoscopy ("while dreaming, I was seeing myself from the outside") • Disrupted identification with the dream self ("while dreaming, I was not myself but a completely different person") • Third-person visual perspective ("I watched the dream from the outside, as if on a screen").
In a survey-based study involving n=151 recent dream reports (n=117 of them collected in a sleep laboratory setting), it was found that the combination of these three factors was more prevalent in lucid dreaming compared to non-lucid dreaming reports (Voss et al., 2013).A separate study focused on lucid dreaming induction (N=27) discovered a positive correlation between these three factors and the level of lucid insight reported by the study participants (Voss et al., 2014).In contrast, a third study involving N=40 participants found no significant differences in the combined factors of 'autoscopy' and 'third-person visual perspective' between lucid and non-lucid dreams induced in a controlled laboratory environment (Carr et al., 2020).In the latter study, those participants who reported lucid dreams but did not perform the eye-signal verification (N=4) scored higher on the combined 'autoscopy' and 'third-person perspective' dissociative factor, implying that full immersion within the dream body might be crucial for controlling eye movements during the dream (Carr et al., 2020).
Together with an elevated visual perspective, 'third-person visual perspective' and 'autoscopy' are also considered key elements of OBEs with visual content (Blanke and Arzy, 2005;Blanke and Mohr, 2005;Bünning and Blanke, 2005).For instance, individuals may describe seeing their physical body as if they were looking at it from the ceiling and from a third-person perspective (Blanke et al., 2004;Jalal et al., 2014).In this context, we would like to emphasize the need for caution in assuming that higher dissociation ratings in lucid dreams necessarily account for episodes of sleep-related OBEs.This is not necessarily the case: as mentioned above, one of the main factors used to measure dissociation in dreams is 'disrupted identification with the virtual (dream) self' (Voss et al., 2013(Voss et al., , 2014)).Conversely, individuals during OBEs frequently identify their own self with the virtual (parasomatic) body rather than the physical body (see Footnote 1).Moreover, neither of the aforementioned studies on lucid dreaming considered 'self-location' as a factor of 'dissociation', nor did they control for this factor at all.Even if participants saw their bodies from an elevated visual perspective during these lucid dreams ("out-of-body autoscopy" in Cheyne and Girard, 2009), we cannot be certain if they also experienced their own self as if located outside of their physical body during that time.
Based on subjective self-reports from multiple studies, episodes of sleep paralysis seem to be typically experienced from a first-person perspective and to involve self-identification of the sleeper with the paralyzed physical body (see reports in Jalal and Hinton, 2013;Jalal et al., 2014;Jalal et al., 2015).In some instances, 'autoscopy' and 'third-person visual perspective' may also intrude this experience (Cheyne et al., 1999;Cheyne, 2003;Cheyne and Girard, 2009;Kliková et al., 2021).In this case, these two factors seem to be associated with vestibular-motor hallucinations and sleep-related OBEs in specific.For instance, when observed in the context of sleep paralysis, 'autoscopy' and 'third-person visual perspective' were classified by principal components analysis in the same category as floating sensations and OBEs-under the 'motor-vestibular hallucinations' factor-in contrast to the 'intruder' and 'incubus' factors (Cheyne et al., 1999;Cheyne, 2003).Moreover, these two factors have been observed to occur during episodes that also present an elevated visual perspective (Cheyne et al., 1999;Cheyne, 2003) or during transitions from sleep paralysis to OBEs (see Case Example I in Jalal et al., 2014).
Thus, 'autoscopy' and 'third-person visual perspective' seem to appear during sleep paralysis specifically in association with other OBElike features, such as having an elevated visual perspective or experiencing vestibular-motor hallucinations.This supports the necessity for a classification of this phenomena that follows a continuum-like pattern (see Section 2).Such a continuum holds the potential to predict transitions from one experience to another, where sleep paralysis and OBElike phenomenology may coexist.In the subsequent section, we will introduce two continua that may shed light on this aspect.

Potential neurophysiological correlates of sleep-related outof-body experiences
The body of research discussed previously suggests that sleep-related OBEs, wake-initiated lucid dreams and sleep paralysis episodes may manifest under the same physiological conditions (e.g., sleep-onset REM periods).However, they may not be identical states but rather potentially represent distinct experiences within a broader experiential spectrum.This divergence could predict the observed phenomenological differences characterizing each phenomenon.For example, minor variations in the neural and physiological correlates revealed by polysomnography may correlate with the heightened sense of reality described during OBEs and sleep paralysis episodes when compared to wake-initiated lucid dreams.
Building on this idea, we will next discuss how OBEs, sleep paralysis and lucid dreams might constitute a continuum of experiences appearing sequentially along the wake-REM transition (Sleep-onset REM mental events continuum, Section 5.1).By applying the predictive coding framework to the dream state (Bucci and Grasso, 2017), we will further suggest that sleep paralysis, OBEs and other unusual bodily experiences reported during sleep might all represent different manifestations of a single continuous phenomenon (Sleep-related OBE spectrum, Section 5.2).This phenomenon would be characterized by the capacity to retain awareness of the fading of bodily sensations and muscular tone, experienced in an intermediate state between wakefulness and sleep.It is important to note that the following discussion is speculative and requires validation in future research.Nonetheless, it offers a predictive framework for guiding future empirical studies in this area (for discussion see Section 5.3).

The Sleep-onset REM mental events continuum
In keeping with the literature, we identified at least three phases that may be present during direct wake-REM transitions (see Fig. 5a):   c) depicting how wake-REM transitions may be experienced either while falling asleep (from wake to REM) or while waking up (from REM to wake).Shaded cells in d) remark one of the many possible combinations for transitions from one phenomenon to another.The direction of this transition is pointed out by two black arrows.In this case, the transition goes from sleep paralysis and/or unusual bodily experiences to nonborderline OBEs and lucid dreams.Within this last step (from non-borderline OBEs to lucid dreams), bodily awareness is lost and the individual achieves lucid insight instead.
(1) Wakefulness (2) Altered sleep onset (3) REM sleep With 'wakefulness' here we refer to the first phase of sleep onset (SO1) as defined by Hori et al. (1994).In this stage, individuals exhibit an EEG pattern that is dominated by a high amount of alpha frequency activity and elevated muscular activity (see Box 1).Although the sleeper may seem physiologically awake during this period, several studies have demonstrated that hypnagogic imagery can already occur as early as during this phase (Foulkes and Vogel, 1965;Germain and Nielsen, 2001).On the other hand, the 'altered sleep onset' phase represents a state in which muscle atonia is already present but may remain unstable.During this period of sleep immobility, the EEG signal exhibits intermediate alpha frequency levels between REM sleep and wakefulness, resembling the second (SO2) and third (SO3) phases of sleep onset isolated by Hori et al. (1994) (see Box 1).This state has been observed during sleep paralysis episodes, hypnagogic hallucinations, false awakenings and pre-lucid dreams, in contrast to periods of eyes-closed wakefulness and/or lucid dreams (Tyson et al., 1984;Takeuchi et al., 1992Takeuchi et al., , 1994Takeuchi et al., , 2002;;Terzaghi et al., 2012;Mainieri et al., 2021).Individuals who have the ability to experience wake-initiated lucid dreams are able to navigate throughout the entire wake-REM transition without losing consciousness in the process (LaBerge et al., 1986;LaBerge and Rheingold, 1991).This provides them with the opportunity to experience sleep paralysis, numbness or distinctive vestibular-motor hallucinations during the 'altered sleep onset' phase, before entering the lucid dream environment (Levitan et al., 1999).Given that most lucid dream reports have been reported to occur during REM sleep (LaBerge, 1988;Baird et al., 2019), we assume that the (wake-initiated) lucid dream environment most likely appears after the 'altered sleep onset' phase, when the individual has already transitioned into REM sleep.
Experiences involving sensations of "entering" and/or "going out of" one's physical body-here referred to as "borderline OBEs"-are commonly reported during sleep paralysis episodes (Cheyne et al., 1999;Cheyne and Girard, 2009;Kliková et al., 2021) and at the onset of lucid dreams (Levitan et al., 1999).This suggests that borderline OBEs may also emerge from the 'altered sleep onset' phase previously described.Additionally, OBE episodes which do not involve any sensations of "entering" or "going out of" the physical body (non-borderline OBEs) may also appear in a state of 'altered sleep onset'.The latter assumption is consistent with the high proportion of "flying", "falling", "levitation" or "OBE" sensations described during sleep paralysis episodes (24.4 % in Kliková et al., 2021) and wake-initiated lucid dreams (at least 30 % in Levitan et al., 1999) compared to awakening reports from regular sleep onset (ranging from 0.25 % to 3.3 % in Foulkes and Vogel, 1965;Rowley et al., 1998) and dream-initiated lucid dreams (at least 12 % in Levitan et al., 1999).
When compared to borderline OBEs, non-borderline OBEs would most likely appear at latter stages within the 'altered sleep onset' phase or at the onset of REM sleep, when complex dreaming is more frequent and there is heightened narrative development compared to earlier sleep stages (see Fig. 5b) (Foulkes, 1962;Foulkes and Vogel, 1965;Vogel, 1991;Rowley et al., 1998;Carr and Solomonova, 2019;Martin et al., 2020).This conclusion is based on the distinct phenomenological features observed in non-borderline OBEs compared to borderline OBEs.For example, individuals reporting non-borderline OBEs typically describe being present in a particular setting (e.g., their bedroom) and, in rarer cases, being able to travel to landscapes (44 % in Blackmore, 1984) or engage in conversations with other people (for example, see Tart, 1968).In contrast, individuals experiencing borderline OBEs commonly report the subjective sensation of "entering" or "leaving" the physical body, accompanied by other hallucinations that often lack a coherent narrative structure, such as hearing rushing sounds or feeling strong bodily vibrations (Twemlow et al., 1982;Blackmore, 1984;Cheyne et al., 1999;Cheyne and Girard, 2009;Rabeyron and Caussie, 2016).Therefore, the phenomenology of these experiences suggests that borderline OBEs occur earlier in the falling asleep process, followed by non-borderline OBEs (see Fig. 5c).
Moreover, the amount of alpha power present in the EEG may predict the emergence of kinaesthetic imagery during the 'altered sleep onset' phase (compare Figs. 5a and b).In this regard, one sleep study involving N=26 participants found that kinaesthetic imagery reported at sleep onset (SO4 phase, see Box 1) was associated with higher alpha power in prefrontal and right temporal-parietal brain regions, when compared to visual imagery occurring in the same stage (Germain and Nielsen, 2001).Another sleep study with N=26 participants found that auditive and somatosensory imagery tends to be more prominent in the earlier stages of regular sleep onset, when alpha activity is at its highest (Hori et al., 1994).In addition, the proportion of visual imagery in this study progressively increased as alpha activity diminished, peaking at later stages of sleep onset (SO6 phase, see Box 1).
Together, these results further support the idea that borderline OBEs, which are more typically associated with auditive, somatosensory and kinaesthetic hallucinations (e.g., hearing rushing sounds or feeling strong bodily vibrations) may occur earlier while falling asleep when compared to non-borderline OBEs.Moreover, it also supports the idea that mental events during wake-sleep transitions unfold following a continuum (Sleep-onset REM mental events continuum, see Figs. 5b and c), making it possible to find intermediate states in which different phenomena may overlap-for instance, transitions from sleep paralysis to OBEs or lucid dreams mentioned by several individuals (Monroe, 1971;LaBerge, 1986;Conesa, 2002;Emslie, 2014;Jalal et al., 2014).Nevertheless, it is important to note that the mentioned studies both focused Box 1 Polysomnography of regular sleep onset.
Regular sleep onset resembles REM sleep in all polysomnographic markers.During sleep onset, we can observe a shift in the EEG signal transitioning from alpha frequencies, which are characteristic of wakefulness (McKinney et al., 2011;Darracq et al., 2018), to theta frequencies, which are characteristic of REM sleep (Simor et al., 2020).This shift is accompanied by a gradual decrease of muscular tone which is less pronounced than the muscle atonia typically observed during REM sleep (Kales and Rechtschaffen, 1968;Chokroverty, 2017;Berry et al., 2020).Sleep onset is also characterized by the presence of both rapid and slow eye movements, similar to the rapid eye movements observed during REM sleep (Foulkes and Vogel, 1965;Vogel et al., 1966).Based on these polysomnographic features, the sleep onset transition can be categorized into nine sub-phases (Hori et al., 1994;Tanaka et al., 1997;Germain and Nielsen, 2001).The initial three stages (SO1-SO3) illustrate the progressive drop of alpha activity associated with drowsiness (Tanaka et al., 1997;Prerau et al., 2014).The subsequent six phases correspond to the flattening of the EEG signal (SO4), followed by the emergence of low-voltage theta waves (SO5), vertex sharp waves (SO6-SO8), and eventually sleep spindles (SO9).According to the sleep scoring rules provided by the American Academy of Sleep Medicine (Berry et al., 2020), phases SO1 and SO2 correspond to a state of 'wakefulness', while phases SO3-SO8 and SO9 correspond to the first and second phases of non-REM sleep-'N1' and 'N2', respectively (see Fig. 4).
T. Campillo-Ferrer et al. on regular sleep onset (Hori et al., 1994;Germain and Nielsen, 2001), rather than sleep-onset REM periods.Thus, further research is needed to clarify whether this sequence of appearance (from kinaesthetic/auditive imagery to visual imagery) or any other sequence is also applicable to direct wake-REM transitions.
Finally, we hypothesize that, as individuals wake up from REM sleep (REM-wake transition), the events may disappear in the reverse order following the Sleep-onset REM mental events continuum (see the twodirectional arrows in Fig. 5c).For instance, most sleep paralysis episodes in the literature were reported in the context of sleep-onset REM periods (Takeuchi et al., 1992(Takeuchi et al., , 2002;;Terzaghi et al., 2012;Mainieri et al., 2021).However, not all of them were linked to the process of falling asleep, but they were also reported to occur upon waking up from these sleep-onset REM episodes (2 out of 3 patients in Mainieri et al., 2021), where individuals may describe hypnopompic hallucinations (Denis et al., 2018).In addition, some individuals have also described sensations of "re-entering" the physical body at the end of their sleep-related OBEs (Blackmore, 1984;Irwin, 1988).This sensation has been described in many different ways, including the experience of being "magnetically drawn back into the body" (LaBerge, 1986) or a change in the point of view that can be coupled with a "wave of vibrations" (Rabeyron and Caussie, 2016).Thus, a mixed stage with similar characteristics to the 'altered sleep onset' phase-combining elements from both wakefulness and REM sleep-could potentially also emerge while awakening from REM sleep.This may initiate conditions leading to sleep paralysis episodes and potentially sleep-related OBEs.

The Sleep-related OBE spectrum
So far, the Sleep-onset REM mental events continuum described before only provides a potential framework for common hallucinations experienced during certain episodes of sleep paralysis, wake-initiated lucid dreams and sleep-related OBEs.Yet, it also makes predictions about the temporal order in which the different phenomena are most likely to appear during sleep-onset REM periods-from sleep paralysis and borderline OBEs, to non-borderline OBEs, to finally lucid dreams.The notion that each phenomenon could manifest at a slightly different polysomnographic stage within the wake-REM sleep transition, may partially account for the distinct phenomenological differences observed between lucid dreams, sleep-related OBEs and sleep paralysis.Nevertheless, this continuum does not fully clarify why perceptual realism is higher in sleep-related OBEs and sleep paralysis when compared to lucid dreams, or why not all wake-initiated lucid dreamers-who remain conscious during the entire wake-REM transition-do not experience sleep paralysis or OBEs.Additionally, it remains an open question whether sleep-related OBEs can also occur out of the context of sleeponset REM periods and therefore in another or similar sleep stage.Some of this variability can be predicted when establishing a distinction between two different types of awareness: (1) lucid insight, and (2) environmental/bodily awareness.In this context, lucid insight refers to being aware of the current state of consciousness, such as being aware of dreaming during lucid dreams (see Section 4.2).On the other hand, we define environmental/bodily awareness as being aware, in this case, of the progressive loss of muscular tone and sensory input coming from the body and external environment while falling asleep.We propose that bodily awareness is not necessarily experienced by wakeinitiated lucid dreamers, but this type of awareness is necessarily present during sleep paralysis episodes-where individuals are by definition aware of their inability to move in an intermediate state between wakefulness and REM sleep (Takeuchi et al., 1992;Takeuchi et al., 2002;Terzaghi et al., 2012;Mainieri et al., 2021)-and potentially during OBEs and other unusual bodily experiences reported during sleep.In this aspect, sleep paralysis and OBEs could be regarded as a single type of phenomenon manifested at different phases or levels of intensity across the wake-REM transition.
The previous speculation aligns with the perspective that dreams can be understood within the predictive coding framework, as proposed by several authors (Bucci and Grasso, 2017;Windt, 2018;Simor et al., 2022).Predictive coding is a theoretical framework suggesting that the human brain continually generates predictions about the external physical world (top-down information) and subsequently compares these predictions with the input received from the senses (bottom-up information).The final goal of this process is to minimize prediction errors and create the most accurate approximation to "reality" as possible (Clark, 2013;Hohwy, 2013).Within this context, sensory input from the periphery is constrained during sleep, leading the brain to estimate this information as "unreliable".This may cause the brain to predominantly rely on initial predictions about the external world (based on prior experience) rather than on actual sensory input, thereby giving rise to the dream experience (Bucci and Grasso, 2017).However, in an intermediate state between wakefulness and REM sleep, the brain may integrate conflicting information concerning the body and external environment, including: (1) externally generated sensory information; and (2) internally generated predictions rooted in past memories and internal constructs.As the brain is partially disconnected from the external environment during this period, it may struggle to effectively reconcile these predictions with incoming sensory input, leading to a disruption in sensory integration.When this ambiguous information becomes accessible to conscious perception at the onset of sleep (i.e., when bodily/environmental awareness is preserved during this process), we speculate that it may result in experiences such as visual or auditory hallucinations (when environmental awareness is preserved) or hyper-realistic unusual bodily experiences (when bodily awareness is preserved).
As the transition from wakefulness to REM sleep progresses, the stream of information from the external senses may progressively become more attenuated (see Fig. 5c).This speculation is derived from studies of typical sleep onset, which indicate a gradual decline in stimulus perception during the falling asleep transition (for review see Goupil and Bekinschtein, 2012).We assume a similar process for wake-REM transitions, although this assumption requires confirmation in future empirical studies.If true, a significant decrease in external sensory input during wake-REM transitions may lead the brain to primarily rely on memories and internal constructs, subsequently giving rise to the creation of dream environments and narrative plots (Bucci and Grasso, 2017).Nevertheless, it is possible that this gradual decline in stimulus perception has not yet reached its peak at this stage (i.e., the threshold for external stimuli to influence the system has not yet reached its minimum).As a result, a significant number of elements from the external physical world may be incorporated into the dream setting when environmental and/or bodily awareness are retained at this point, causing individuals to experience a dream setting that mimics the current sleeper's surroundings (e.g., one's bed or bedroom).
Therefore, sleep paralysis, sleep-related OBEs and other unusual bodily experiences reported during sleep, may form part of a spectrum of experiences according to the predictive coding framework (Sleep-related OBE spectrum).Within this spectrum, internally generated (top-down) information might intrude into the waking state during the early stages of the wake-REM transition, while externally generated (bottom-up) information could intrude into the sleep state in subsequent stages.We hypothesize that the interplay between internal and externally sourced components in shaping one's body representation (i.e., retaining bodily awareness during wake-sleep intermediate states) is what results in sleep paralysis episodes and sleep-related OBEs as subjectively seeming more real compared to false awakenings and (lucid) dreams.

Other models of sleep and consciousness consistent with the Sleeprelated OBE spectrum
A similar hypothesis was initially proposed by Blackmore (1988), who additionally argued that when internally generated information becomes more dominant than externally generated information, the brain eventually labels internal constructs and memories as the new T. Campillo-Ferrer et al. "reality".As a result, the "model of reality" as understood by Blackmore (1988) suddenly shifts,7 leading individuals during OBEs or non-lucid dreams to experience these memories and imaginations as if they were actual representations of the external physical world (for more elaboration on this model, see Blackmore, 1988).Interestingly, the sudden shift of perception pointed out by Blackmore may account for descriptions of a "subjective break", "click sensation" or "momentary gap in the stream of consciousness" reported at the onset of sleep-related OBEs (Irwin, 1988;Rabeyron and Caussie, 2016).
The latter can be compared to the sudden shift of perception experienced during bistable perceptual illusions such as the Spinning Dancer Illusion, where the brain can only "choose" one configuration at a time (C.-H.Liu et al., 2012;Alipour and Kazemi, 2015).In the case of the Spinning Dancer Illusion, the observer can either see the dancer spinning clockwise or anti-clockwise, a direction that may spontaneously switch later.In line with the previous idea, an unbalanced brain activity shifted towards the Default Mode Network and slower frequencies of the EEG-at the expense of the Central Executive Network and faster frequencies of the EEG-has been proposed to underlie the switch from externally-oriented cognition to internally-oriented cognition during dreams (Northoff et al., 2023).
Another model that considers altered states of consciousness during sleep is the activation-input-modulation (AIM) model (Hobson et al., 2000;Hobson, 2009), where states of consciousness are conceptualized as variations in a three-dimensional space composed by three axes: (A) The activation axis, representing high or low levels of cortical activation.High levels of cortical activation here could also be interpreted as the activation of specific brain regions or the presence of faster frequencies in the EEG, such as during wakefulness and REM sleep (high activation) compared to non-REM sleep (low activation); (I) The input-output gating axis, reflecting the gradual inhibition of external inputs and motor activity regulated by the brainstem.This inhibition is minimal in wakefulness and maximal in REM sleep; (M) The modulation axis, depicting the ratio of neurotransmitters operating in the brain.During sleep, neuromodulation shifts from mostly aminergic and serotonergic activity in wakefulness, to mostly cholinergic activity in REM sleep (Hobson et al., 2000;Hobson, 2009).
Within the AIM model (see Fig. 6a), lucid dreaming and sleep paralysis have traditionally been depicted as intermediate states between wakefulness and REM sleep, considering both the modulation and inputoutput gating axes (Hobson et al., 2000;Hobson, 2009;Denis, 2018).In Fig. 6b, we present a reinterpretation of the AIM model that is also consistent with the Sleep-related OBE spectrum and more recent lucid dreaming literature.This reinterpretation depicts lucid dreaming as a state of consciousness occurring in REM sleep rather than in a mixed stage between wakefulness and REM sleep (Mainieri et al., 2021;Gott et al., 2024).Moreover, the emergence of lucid insight during lucid dreaming may be linked to the activation of specific brain regions and a possible cholinergic pressure (for reviews see Baird et al., 2019;Gott et al., 2024).This aspect has been depicted as a shaded area within the 'REM sleep' sphere.
On the other hand, our interpretation of the AIM model illustrates sleep paralysis, vestibular-motor hallucinations, and (borderline) OBEs as originating from the 'altered sleep onset' phase mentioned in Section 5.1.This 'altered sleep onset' phase has been placed in an intermediate position between wakefulness and REM sleep, as suggested by multiple research studies (Tyson et al., 1984;Takeuchi et al., 1992Takeuchi et al., , 1994Takeuchi et al., , 2002;;Terzaghi et al., 2012;Mainieri et al., 2021).Similar to lucid dreaming, the activation of specific brain regions associated with the experience of 'bodily awareness' (see Section 5.3 for potential candidates) may coincide with the occurrence of sleep paralysis, vestibular-motor hallucinations, and (borderline) OBEs.This aspect has been depicted as a shaded area within the 'altered sleep onset' sphere in Fig. 6b.
From Fig. 6b, it can be observed that non-REM sleep and the 'altered sleep onset' phase exhibit similar levels of neuromodulation (M) and input-output gating (I), differing only in the level of cortical activation (A).Thus, our interpretation of the AIM model allows for the possibility that OBEs could also emerge from non-REM sleep if this phase were coupled with heightened levels of cortical activation-for example, due to an increase in bodily awareness during non-REM sleep (see blue arrow in Fig. 6b).Alternatively, OBEs could also manifest in the context of other sleep transitions different from sleep-onset REM periods.While the sleep continuum is typically categorized into distinct stages, transitions from one phase to another are not necessarily abrupt and may manifest changes in arousability, muscle activity and the progressive modification of the EEG signal (Pilcher and Schulz, 1987;Hori et al., 1994;Simor et al., 2014;Carrera-Cañas et al., 2019).Therefore, these gradual alterations could eventually lead to ambiguous sleep phases triggering the OBE phenomenon.For example, Hobson et al. (2000) suggest that transitions from non-REM to REM sleep may occur in a dissociated manner within the AIM space, where changes in cortical activation may precede changes in neuromodulation and input-output gating (see Figure 19 in Hobson et al., 2000).Such a dissociation may give rise to a state resembling the 'altered sleep onset' phase-with high cortical activation but an intermediate position on the modulation and input-output gating axes-which may potentially facilitate the OBE phenomenon (see red arrow in Fig. 6b).

Other literature supporting the Sleep-related OBE spectrum
Several studies support the idea that OBEs form part of a continuum of experiences that emerge sequentially and possibly at different levels of intensity.For instance, research on a patient suffering from right temporal-lobe epilepsy (Blanke et al., 2002) suggests that the intensity of focal electrical stimulation of the right angular gyrus can modulate the perception of either (1) Somatosensory illusions and body distortion (at low intensities, 2-3 mA), or (2) OBEs involving the observation of one's body from the outside (at high intensities, ≥ 3.5 mA).Additionally, a study investigating OBEs in the context of sleep paralysis found a correlation between the variables of (1) illusory bodily-self movement experiences, (2) feelings of separation from one's body, and (3) experiences of observing one's own body from an above visuo-spatial perspective (Cheyne and Girard, 2009), which seem to most likely appear following a sequential order.These researchers conducted multiple regression analyses based on different surveys that were collected from two distinct samples: one comprising retrospective reports of sleep paralysis from 11,385 participants, and another one gathering prospective reports of recent individual sleep paralysis episodes from 314 participants (Cheyne and Girard, 2009).The best-fitting model found in this study supports the hypothesis that illusory bodily-self movement experiences (IMEs) precede feelings of separateness from one's body (OBFs), which in turn mediate the relationship between IMEs and episodes involving the observation of one's body from an elevated visuo-spatial perspective (OBAs).This results in the following model: IMEs → OBFs → OBAs (Cheyne and Girard, 2009).Our model is also in line with subjective reports from individuals who describe body distortion and other unusual bodily experiences prior to sleep-related OBEs and false awakenings (LaBerge, 1986;Rabeyron and Caussie, 2016;Mainieri et al., 2021;Campillo-Ferrer et al., 2023).
Additionally, multiple studies have demonstrated that direct stimulus incorporation is possible during dreams (for review see Salvesen et al., 2024).An example of this phenomenon can be found in Konkoly et al. (2021), where lucid dreamers perceived some light cues and sounds played by the researchers as if they were part of the dream plot rather than as if coming from the outside.Moreover, some memory elements can also be integrated into the dream scenery.This phenomenon is evident in the so-called "laboratory dreams", where subjects participating in sleep studies dream about different elements of the laboratory environment (Schredl, 2008).In this context, both sensory incorporation and meta-dreaming-the latter including lucid dreams, sleep paralysis, false awakenings and OBEs-were identified as common themes of laboratory dreams (Picard-Deland et al., 2021).
Thus, the direct incorporation of auditive, tactile and proprioceptive stimuli (bottom-up signals), in combination with the access to visual memories of the current physical space (top-down signals), may lead the individual to experience a scenery that closely represents their immediate surroundings.This would also predict cases where the OBE environment was described as familiar but did not represent the current sleeper's setting.For example, in one of his sleep-related OBEs, LaBerge (1986) initially believed he had left his body in the same environment where his physical self was sleeping, only to realize upon waking up that the scene depicted a house he had not lived in for over 20 years.
Finally, as modelled in the Sleep-related OBE spectrum, several authors have also highlighted the convergence of awareness and abnormal somatosensory/motor signalling (such as muscle atonia) as the common elements triggering OBE-like experiences (Blackmore, 1988;Levitan et al., 1999;Bünning and Blanke, 2005;Overney et al., 2009;Millière et al., 2018).For instance, the likelihood of experiencing OBEs during periods of anaesthesia has been suggested to be associated with surgical awareness and the administration of muscle relaxants (Bünning and Blanke, 2005).An additional study suggested that the combination of cannabis consumption and severe loss of motor and somatosensory function may underlie recurrent OBEs in a patient suffering from multiple sclerosis (Overney et al., 2009).
Therefore, maintaining 'bodily awareness' during wake-REM transitions may represent yet another means of becoming aware of a state characterized by (1) reduced bodily sensations and (2) abnormal motor function (muscle atonia), and thus just another way to increase the likelihood of experiencing OBEs.This combination of factors is not specific to conscious wake-REM transitions; rather, it may occur upon awakening from REM sleep, during non-REM sleep, or in many other conscious states such as during meditation and psychedelic experiences.

A potential tool to predict out-of-body experiences during sleep and dreaming
To summarize, the literature reviewed in this paper suggests that sleep paralysis, OBEs and other unusual bodily experiences form a single type of phenomenon that occurs at varying levels of intensity.This phenomenon would be characterized by heightened awareness of the physical body ('bodily awareness') in situations where there is limited external body-related sensory input available to the brain and an abnormal motor function, such as during wake-sleep transitions (Sleeprelated OBE spectrum).Figs.5c and d illustrate the interplay between the Sleep-related OBE spectrum and the Sleep-onset REM mental events continuum.The latter outlines the order in which sleep paralysis episodes, borderline OBEs, OBEs and lucid dreams would most likely appear when experiencing a sleep-onset REM period.
Thus, enhanced 'bodily awareness' during sleep may co-occur together with increased 'lucid insight' and 'environmental awareness'.In the latter case, increased awareness of the external environment in an intermediate state between wakefulness and sleep may potentially trigger OBEs with visual and auditive content.Additionally, 'lucid insight' and 'environmental awareness' can also manifest independently from 'bodily awareness' under this framework.For instance, this independence could be observed in lucid dreams (with only high 'lucid insight'), false awakenings (with only high 'environmental awareness') or in certain lucid dreaming episodes where the dream setting resembles the immediate physical surroundings of the sleeper (with both high 'lucid insight' and high 'environmental awareness', see "environmentally-realistic lucid dreams" in Fig. 5d).In fact, there might be many possible combinations leading to different states, reflecting the high phenomenological variability inherent to these experiences.
Investigating the neural correlates associated with 'bodily awareness' and 'environmental awareness' during non-wakefulness and altered states of consciousness-and particularly during sleep, meditation or psychedelic experiences-could elucidate the potential role of these forms of awareness in facilitating the OBE phenomenon.In this context, two potential candidates for investigation are the insular and temporo-parietal cortices.Neuroimaging studies show abnormal activity in the insular cortex and/or temporo-parietal junction during OBEs induced by electrical stimulation or epileptic seizures (Blanke et al., 2002(Blanke et al., , 2004;;De Ridder et al., 2007), as well as during unusual bodily experiences reported in psychedelic and meditative states (Dor-Ziderman et al., 2016;Tagliazucchi et al., 2016).Moreover, one study linked kinaesthetic imagery at sleep onset to higher alpha power in right temporal-parietal regions when compared to visual imagery (Germain and Nielsen, 2001).
The insular cortex and temporo-parietal junction are crucial for multisensory integration (Blanke and Mohr, 2005;Blanke, 2012;Simmons et al., 2013).Multiple studies suggest that these areas are involved in processing interoceptive and exteroceptive signals from the body, as well as mental concepts not directly linked to the body, such as recognizing one own's name (for meta-analysis see Qin et al., 2020).Consequently, the insular cortex and temporo-parietal junction have been proposed as potential neural correlates of consciousness (Craig, 2009) and bodily self-consciousness (Park and Blanke, 2019), possibly forming part of a broader neural network involved in different aspects of self-consciousness such as self-identification and self-location (Park and Blanke, 2019).Overall, this accumulating evidence aligns with the idea that the insular cortex and temporo-parietal junction (along with other possibly associated brain regions) may play a pivotal role in the accurate processing of sensory information related to the body, and thus in the potential preservation of bodily awareness during OBEs and sleep paralysis.

Conclusions and future perspectives
In conclusion, while studies over the last few decades have advanced our understanding of OBEs, robust empirical data and a comprehensive grasp of this phenomenon are still lacking.Firstly, the variability of the concept makes it challenging to establish a simple and clear-cut definition for OBEs.Secondly, there is a lack of research on OBEs occurring during sleep and dreaming, even though a significant portion of spontaneous OBEs reported by healthy individuals seems to take place in such states.To shed light on these aspects, we have conducted a literature review examining the link between OBEs and sleep states.We have paid particular attention to the phenomenological differences that appear to exist between sleep-related OBEs and episodes of lucid dreaming or sleep paralysis.Finally, we have developed a conceptual model that may prove useful in the future for predicting sleep-related OBEs and other related phenomena.
The conceptual model we present here is based on 4 primary hypotheses: • All out-of-body experiences are characterized by the subjective experience of being located outside of the physical body (extracorporeal self-location).Throughout this paper, we have presented scientific evidence consistent with these hypotheses or supporting them, we have briefly explored the potential involvement of other sleep stages in the occurrence of OBEs-including states of dissociated non-REM sleep-and finally discussed the possible neural correlates underlying both OBEs reported during sleep and OBEs reported in wakefulness.However, it should be emphasized that these assumptions are merely speculative, and that previous investigations have not yet revealed a clear polysomnographic pattern associated with sleep-related OBEs.Furthermore, only one study suggests a direct association between sleep-onset REM periods and OBE frequency (Levitan et al., 1999).
Future polysomnographic studies should incorporate larger sample sizes to investigate the sleep stages where OBEs are most likely to manifest.Additional research combining EEG and fMRI techniques would facilitate a deeper exploration of the neuroanatomical and neurofunctional basis of sleep-related OBEs, as well as their direct comparison with lucid dreaming states.For any such research, either individuals who experience OBEs frequently or the development of new induction methods would be necessary.For the latter, a combination of relaxation techniques (Palmer and Lieberman, 1975;McCreery andClaridge, 1996a, 1996b), meditation aimed at cultivating bodily sensory awareness (Campillo-Ferrer et al., 2023), and approaches based on inducing sleep-onset REM periods (Takeuchi et al., 1992(Takeuchi et al., , 2002) might be promising.Alternatively, studies assessing the prevalence of OBEs in populations experiencing frequent REM sleep intrusions-such as shift workers (Kotorii et al., 2001;Goldbart et al., 2014)-may also help to clarify the potential link between sleep-onset REM periods and OBEs.

Fig. 1 .
Fig. 1.Electrophysiology of sleep-onset REM periods.a) Polysomnographic recording showing a direct transition from wakefulness into REM.The activity of three EEG channels is shown (F3-M2, Cz-M1, O1-M2), including one frontal (F3), one central (Cz) and one occipital (O1) EEG channel, all referenced to either the left (M1) or right mastoid (M2).Two electrooculograms referenced to the left mastoid (Right or REOG-M1, Left or LEOG-M1) and one electrocardiogram (ECG) show the progressive slowing of ocular and cardiac activity during this period.One chin electromyogram (EMG) shows the progressive loss of postural tone (muscle atonia) characteristic of drowsiness and REM sleep.b) Hypnogram of a patient with narcolepsy type 1 (top figure) compared with the hypnogram of a healthy individual (bottom figure).Note the sleep-onset REM periods (red and grey arrows) appearing in the context of narcolepsy type 1. Reprinted from "Narcolepsy" by Birgitte R. Kornum et al., (2017), Nature reviews.Disease primers, 3, 16100.Copyright [2017] by Macmillan Publishers Limited (Springer Nature).Reprinted with permission.

Fig. 2 .
Fig. 2. Comparison of the factors 'lucid insight' and 'realism' in sleep-related OBEs and similar states of consciousness.Cells containing a star (*) under the '(sleep-related) OBEs' category indicate phenomenological features that were extracted from general OBE literature rather than from specific sleep-related OBE literature.Coloured boxes highlight similarities between different phenomena.

Fig. 3 .
Fig. 3. Comparison of the factors 'control', 'memory', 'thought', 'emotions' and 'dissociation' in sleep-related OBEs and similar states of consciousness.Cells containing a star (*) under the '(sleep-related) OBEs' category indicate phenomenological features that were extracted from general OBE literature rather than from specific sleep-related OBE literature.Colour gradients in all 'emotions' and 'dissociation' categories illustrate lucid dreams as a potential intermediate state between sleep paralysis and sleep-related OBEs within this context.

Fig. 5 .
Fig.5.The Sleep-onset REM mental events continuum and Sleep-related OBE spectrum.Colour gradients represent the fading of alpha EEG activity in a), the fading of muscle activity and kinaesthetic/auditive imagery in b), and the shift from externally-sourced sensory input to internally-sourced sensory input in c), all experienced during wake-sleep transitions.Reverse gradients in b) represent the increase of visual imagery, narrative development and complex imagery experienced over time during sleep.Note the two-directional arrows in a) and c) depicting how wake-REM transitions may be experienced either while falling asleep (from wake to REM) or while waking up (from REM to wake).Shaded cells in d) remark one of the many possible combinations for transitions from one phenomenon to another.The direction of this transition is pointed out by two black arrows.In this case, the transition goes from sleep paralysis and/or unusual bodily experiences to nonborderline OBEs and lucid dreams.Within this last step (from non-borderline OBEs to lucid dreams), bodily awareness is lost and the individual achieves lucid insight instead.

Fig. 6 .
Fig. 6.Activation-input-modulation (AIM) model of consciousness.a) AIM model of consciousness including lucid dreaming and sleep paralysis states as described by Hobson et al. (2000), Hobson (2009) and Denis (2018).b) Reinterpretation of the traditional AIM model taking into account literature reviewed in the current paper.A, I and M represent the three axes composing the AIM model: activation axis, input-output gating axis and modulation axis, respectively.Shaded areas within the 'altered sleep onset' and 'REM sleep' spheres depict movements in the three-dimensional AIM space potentially associated with sleep paralysis, vestibular-motor hallucinations, OBEs ('altered sleep onset' sphere), and lucid dreams ('REM sleep' sphere).The blue arrow indicates a potential cortical activation happening during non-REM sleep, which may give rise to a state similar to the 'altered sleep onset' phase.The red arrow indicates a potential dissociation of the non-REM to REM sleep transition, which may give rise to the 'altered sleep onset' phase.Adapted from "REM sleep and dreaming: towards a theory of protoconsciousness" by J. Allan Hobson, (2009), Nature reviews.Neuroscience, 10, 803-813.Copyright [2009] by Macmillan Publishers Limited (Springer Nature).Reprinted with permission.

Control, memory and thought Emotions Dissociation Autoscopy Visual perspective Self-identification (Sleep-related) OBEs* Lucid dreams Sleep paralysis
• Maintaining consciousness during transitions from wakefulness to REM sleep (or from REM sleep to wakefulness) provide the physiological background conducive to sleep paralysis episodes, wakeinitiated lucid dreams and sleep-related OBEs.• The emergence of the different phenomena-such as sleep paralysis, lucid dreams or sleep-related OBEs-follows a temporal order that may sometimes overlap (Sleep-onset REM mental events continuum).• Sleep paralysis episodes, unusual bodily experiences and sleeprelated OBEs emerge when stronger awareness of the external physical body is maintained in an intermediate state between wakefulness and sleep (Sleep-related OBE spectrum).This heightened bodily awareness may contribute to the perception of sleep-related OBEs and sleep paralysis episodes as more real compared to lucid dreams or false awakenings.