The Bodily Self from Psychosis to Psychedelics
This experimental study (n=75) investigates the impact of psychedelic and psychiatric experiences on the sense of self. It finds that patients with psychosis exhibited reduced Body Ownership and Sense of Agency, while participants with substantial psychedelic experiences reported enduring subjective changes to the sense of self, although no differences were found at the level of the bodily self when compared to control participants.
Authors
- Harduf, A.
- Harel, E. V.
- Panishev, G.
Published
Abstract
The sense of self is a foundational element of neurotypical human consciousness. We normally experience the world as embodied agents, with the unified sensation of our selfhood being nested in our body. Critically, the sense of self can be altered in psychiatric conditions such as psychosis and altered states of consciousness induced by psychedelic compounds. The similarity of phenomenological effects across psychosis and psychedelic experiences has given rise to the “psychotomimetic” theory suggesting that psychedelics simulate psychosis-like states. Moreover, psychedelic-induced changes in the sense of self have been related to reported improvements in mental health. Here we investigated the bodily self in psychedelic, psychiatric, and control populations. Using the Moving Rubber Hand Illusion, we tested (N=75) patients with psychosis, participants with a history of substantial psychedelic experiences, and control participants to see how psychedelic and psychiatric experience impacts the bodily self. Results revealed that psychosis patients had reduced Body Ownership and Sense of Agency during volitional action. The psychedelic group reported subjective long-lasting changes to the sense of self, but no differences between control and psychedelic participants were found. Our results suggest that while psychedelics induce both acute and enduring subjective changes in the sense of self, these are not manifested at the level of the bodily self. Furthermore, our data show that the bodily self-processing, related to volitional action, is disrupted in psychosis patients. We discuss these findings in relation to anomalous self-processing across psychedelic and psychotic experiences.
Research Summary of 'The Bodily Self from Psychosis to Psychedelics'
Introduction
Harduf and colleagues frame the sense of self as a multi-level construct that is normally experienced as a unified, embodied phenomenon but can be perturbed in psychiatric and drug-induced altered states. Previous literature has identified substantial phenomenological overlap between psychosis and psychedelic experiences, including hallucinations and alterations of both the narrative self and the more basic 'minimal' or bodily self. Two core components of the bodily self receive particular attention: Body Ownership (BO), the feeling that one’s body or parts belong to oneself, and the Sense of Agency (SoA), the feeling of controlling one’s actions. BO is thought to arise from multisensory integration, whereas SoA depends on predictive sensorimotor processes that compare motor predictions with sensory feedback. This study set out to compare bodily self-processing across three groups: people with psychosis, people with a history of substantial psychedelic use, and neurotypical controls. Using the Moving Rubber Hand Illusion (MRHI), which combines visuotactile and visuomotor manipulations to probe BO and SoA within participants, the investigators aimed to test whether long-term or clinical alterations of selfhood are reflected in altered multisensory bodily processing and to assess whether psychedelic experiences produce effects comparable to psychosis at the level of the bodily self.
Methods
Seventy-five right-handed adults participated (N = 75; 27 females; mean age 29.9, SD = 8.85), divided into three equal groups: Control (N = 25; 16 females; mean age 24.84), Psychedelic (N = 25; 11 females; mean age 28.8) and Psychosis (N = 25; all male; mean age 36.08). Control and Psychedelic participants self-reported no history of neurological, psychiatric, or tactile/motor disorders; Psychedelic participants were recruited via psychedelic social media groups and all reported prior psychedelic use, whereas Controls reported none. The Psychosis group consisted of inpatients at a mental health centre. Ethical approvals and written informed consent were obtained. The extracted text notes that the Psychosis group was older than the other groups. To quantify enduring subjective changes in self-experience, the Psychedelic and Control groups completed a bespoke questionnaire drawing items from established instruments (Ego-Dissolution Inventory, Cambridge Depersonalization Scale, Mystical Experience Questionnaire). The Psychosis group’s clinical symptoms were assessed with the Positive and Negative Syndrome Scale (PANSS). The MRHI experimental apparatus placed participants’ real right hand out of sight and a life-size gender-matched rubber hand in view; a cloth and glove preserved visual continuity. The experimental design included two stimulation modalities: Visuotactile (VT), in which the experimenter brushed the participant’s hidden right hand and the rubber hand, and Visuomotor (VM), in which participants produced index-finger movements at ~1 Hz that mechanically moved the rubber hand via a concealed rod. For both VT and VM there were three conditions: synchronous anatomically congruent (Sync), synchronous anatomically incongruent (Incong—participant index finger coupled to rubber middle finger), and asynchronous anatomically congruent with a 500 ms delay (Async). Each of the six combinations (2 modalities × 3 conditions) was administered in a separate one-minute run presented in pseudorandom order. After each run participants completed a 12-item MRHI questionnaire (3 BO items, 3 SoA items, and 6 control items) rating statements on a −3 to +3 Likert scale; Controls and Psychedelic participants used a num-pad, Psychosis participants responded verbally. Statistical analyses used mixed repeated-measures ANOVA with Condition (Sync/Incong/Async) as a within-subject factor and Group as between-subjects factor; post-hoc tests were Bonferroni corrected. For nonsignificant effects the authors performed Bayesian repeated-measures ANOVA, using Bayes factors of exclusion (BF01) to quantify evidence favouring models without the factor. Nonparametric Wilcoxon signed-rank and Mann-Whitney tests were used where distributions deviated from normality. Illusion strength was operationalised as the difference in subjective rating between Sync and Async conditions. Spearman correlations examined the BO–SoA relationship.
Results
Participant characteristics and questionnaire data: All three groups contained 25 participants. The Psychosis group was significantly older than the other groups (Psychosis vs Control p < 0.001; Psychosis vs Psychedelic p < 0.01). The Psychedelic group reported substantially more enduring altered-self experiences than Controls across the three questionnaire subscales: Ego-Dissolution Inventory (EDI) mean 7.28 (SD 2) versus Control mean 3.85 (SD 1.92), W = 68, p < 0.0001; Mystical Experience Questionnaire (MEQ) mean 8.27 (SD 1.65) versus Control 3.02 (SD 2.45), W = 38, p < 0.0001; Cambridge Depersonalization Scale (CDS) mean 5.33 (SD 1.99) versus Control 3.16 (SD 1.77), W = 127, p < 0.001. The Psychosis group did not complete this questionnaire. Visuomotor (VM) stimulation — Sense of Agency (SoA): VM synchronous (Sync) and anatomically coupled conditions produced positive SoA ratings overall (Sync mean 1.81, SEM 0.15, V = 2527, p < 0.001; Incong mean 0.85, SEM 0.21, p < 0.001), whereas the Async VM condition did not (mean −0.18, SEM 0.21, p > 0.05). There was a strong main effect of Condition (F(2,144) = 39.92, p < 0.001, partial η2 = 0.35). A main effect of Group was also observed (F(2,72) = 4.19, p < 0.05, partial η2 = 0.10), with post-hoc tests showing the Psychosis group rated SoA lower than the Control group (p < 0.05); the Psychosis vs Psychedelic comparison did not reach significance (p = 0.09). Bayesian analysis provided moderate evidence supporting no difference between Control and Psychedelic groups (BF01 = 3.69). There was no significant Group × Condition interaction (F(4,144) = 0.714, p = 0.58; BF01 = 4.58), indicating that the pattern of condition effects was similar across groups despite lower overall SoA in the Psychosis group. Visuomotor (VM) stimulation — Body Ownership (BO): VM Sync produced modest BO overall (mean 0.4, SEM 0.21, V = 1665, p < 0.05), whereas Incong and Async were negative and non-significant. Condition had a strong effect (F(2,144) = 30.15, p < 0.001, partial η2 = 0.29). The main effect of Group was not significant (F(2,72) = 0.202, p = 0.81; BF01 = 3.63), but there was a significant Group × Condition interaction (F(4,144) = 2.96, p < 0.05, partial η2 = 0.07). This interaction was driven by the Psychosis group’s lack of a BO experience in the Sync VM condition (group mean below 0). Comparing illusion strength (Sync − Async), the combined Healthy groups (Control + Psychedelic) showed greater BO illusion strength (mean 2.00, SEM 0.25) than the Psychosis group (mean 0.52, SEM 0.33; W = 916, p < 0.001). Pairwise Sync–Async differences were significant in Control (W = 252, p < 0.001) and Psychedelic (W = 290, p < 0.001) groups but not in the Psychosis group (W = 107, p = 0.07). Visuotactile (VT) stimulation — Body Ownership and SoA: VT Sync induced BO across the sample (mean 0.97, SEM 0.22, V = 2107.5, p < 0.001) while Incong and Async did not. Condition had a strong effect (F(2,144) = 31.52, p < 0.001, partial η2 = 0.30). There were no significant group differences in VT-induced BO (F(2,72) = 0.43, p = 0.64; Bayesian BF01 = 8), and no significant Group × Condition interaction (F(4,144) = 1.45, p = 0.22; BF01 = 8.57), indicating similar VT BO across Control, Psychedelic and Psychosis groups. VT did not induce SoA in any group (Sync mean −0.87, SEM 0.21; other VT conditions even more negative), though Condition differences were present (F(2,144) = 12.04, p < 0.001). No Group effects emerged for VT SoA (F(2,72) = 0.16, p = 0.85; BF01 = 6.83). Additional observations: Control items were consistently rated negatively, supporting the specificity of illusion-related ratings. Finally, BO and SoA illusion strengths were positively correlated: in VM r = 0.34 (p < 0.01) and in VT r = 0.51 (p < 0.001), indicating that stronger ownership tended to co-occur with stronger agency reports.
Discussion
Harduf and colleagues interpret the findings as evidence that psychosis and psychedelic experiences differentially affect bodily self-processing. The principal observations were that all groups showed typical VT-induced BO, the Psychosis group exhibited reduced SoA across VM conditions, and only the Control and Psychedelic groups displayed VM-induced BO; Psychosis participants did not. From these results the investigators infer that aberrant bodily experiences in psychosis are specifically linked to disrupted processing of volitional actions and predictive sensorimotor mechanisms rather than a global multisensory ownership deficit. In other words, impaired SoA during active movement may in turn diminish ownership under VM conditions, while VT-induced multisensory integration for BO remains relatively intact in psychosis. Regarding psychedelics, the authors emphasise a dissociation between enduring subjective reports of altered self-experience and preserved basic multisensory bodily processing. Although the Psychedelic group reported significantly higher lifetime ego-dissolution, depersonalisation, and mystical-type experiences than Controls, their MRHI responses closely matched Controls in both VT and VM conditions. The investigators therefore argue that long-term psychedelic use does not produce persistent changes in the sensorimotor processes that underpin BO and SoA, suggesting limits to the psychotomimetic account when applied to bodily self-processing. The authors acknowledge several limitations that temper the conclusions. The Psychosis group was older and exclusively male due to practical recruitment constraints, and they did not complete the self-experience questionnaire. Acute psychedelic effects were not examined, so no claim is made about transient alterations during intoxication. The authors call for future work to examine different compounds, acute versus lasting effects, and to address demographic imbalances.
Conclusion
The authors conclude that psychosis is associated with diminished agency and reduced body ownership during volitional actions, implicating disrupted predictive sensorimotor processing in altered bodily selfhood. By contrast, substantial prior psychedelic use was linked to enduring self-reports of altered phenomenology but did not correspond to lasting changes in multisensory bodily processing as measured by the MRHI. Consequently, the findings suggest that considerable psychedelic exposure does not alter the low-level multisensory mechanisms underlying the bodily self, though it may affect more explicit or narrative aspects of selfhood.
View full paper sections
INTRODUCTION
The self is a central organizing principle of our cognition and experience. The sense of self is composed of different levels or modelssubserved by different neural systems, which in neurotypical states are manifested as a robust and unified phenomenological experience. Altered states of consciousness, such as those found in neurological, psychiatric, and psychedelic states, are known to alter this experience e.g., 14-16 . Notably, psychosis and psychedelic states both produce hallucinations and substantial alterations in the sense of self e.g.,, causing some researchers to consider the psychedelic experiences as psychotomimetic. For example, psychedelic experiences may induce changes across many levels of self, from the narrative selfto robust changes in the 'minimal self'. Changes in minimal self may include abnormal feelings of ownership over one's body, aberrant bodily sensations, loss of spatial self-location, and even a loss of the sense of self "ego dissolution". Indeed, "ego dissolution" is an extreme condition associated with high doses of psychedelics leading to a loss of segregation between oneself and surroundings and typically accompanied by a feeling of unity with the universe. Similarly, the disrupted self is a prominent feature in psychosis. Schizophrenia (SCZ) patients show deficits across different levels of self, including narrative, social, and bodilymodels of the self. Positive symptoms in SCZ include diminished control over one's thoughts and actions, which lead to "passivity" experiences like auditory hallucinations and thought insertion. It has been proposed that abnormal predictive sensorimotor processes blur the demarcation of self-generated sensations and drive self-deficits in psychosis. This, in turn, may cause one's inner voice, thoughts, and actions to be felt as if they are generated by some external agent leading to symptoms found in psychosis. Thus, both psychedelic and psychotic states are unique examples in which the fundamental aspects of the bodily self are altered. The bodily self consists of two fundamental aspects, Body Ownership (BO), the experience of identifying with the body e.g.,, and the Sense of Agency (SoA), the experience of control over one's actions e.g.,. BO relies on multisensory integration in which exteroceptive and interoceptive sensory signals are bound to form a coherent model of the self e.g.,. For example, when the brain receives converging visual, tactile, and proprioceptive signals congruent with prior implicit models of the self, a feeling of ownership arises for this limb. The SoA is grounded in predictive motor and inference processes linking volitional actions to the predicted sensory outcomes e.g.,. Previous work has shown that voluntary actions are accompanied by an efference copy, which is used to compute forward models predicting the expected sensory consequences e.g., 63,64,but see 65 . The comparison of the predicted and actual sensory feedback enables the agent to distinguish between internally and externally generated sensations. If they match, a feeling of a selfgenerated movement arises (that the action is self-generated), and this is accompanied by both behavioraland neuralsuppression of the sensory consequences. If there is a large discrepancy between an action's predicted and actual sensory outcomes, the action is not attributed to oneself. For example, if the visual outcome of an action is modified temporally e.g.,, anatomically e.g.,, or spatially e.g., 76-78 , participants do not experience authorship of the action. Thus, both BO and SoA rely on integration of sensory signals, however, SoA utilizes information from volitional action which can be used to improve inferences on the bodily self. The experimental study of the bodily self has relied heavily on the induction of illusory states of BO and SoA. In their seminal work, Botvinick and Cohen revealed that visuotactile (VT) stimulation on a visible rubber hand, synchronous to tactile stimulation on one's unseen real hand, induces illusory ownership over the rubber hand. The rubber hand illusion (RHI), as well as full body and virtual versions of this paradigm, have become a mainstay of experimental research on the bodily self e.g., 1,13,80-82 . Indeed, numerous subsequent studies have shown that induction of illusory BO is accompanied by replicable behavioral, physiological, and neural states. SoA is typically tested by introducing discrepancies between participants' movements and their visual consequences. For example, several studies have used temporal or spatial visuomotor (VM) discrepancies to alter and decrease the experience of SoA e.g.,. Most experiments have focused on either BO or SoA such that their interactions were neglected. The Moving Rubber Hand illusion (MRHI), a paradigm introduced a decade ago by Kalcaket and Ehrsson allowed examining both BO and SoA by introducing both VT and VM illusions in a within-subject design. Here we used the MRHI to examine alterations in the bodily self across participants with substantial psychedelic experience, participants with psychosis, and neurotypical controls. In psychosis, there is substantial evidence of impairments in SoA. Across different paradigms, participants on the psychosis spectrum show a reduced ability to discriminate self-initiated actions. Recent evidence has shown that this deficit is even present in non-symptomatic 22Q11DS participants with a genetic propensity for psychosis, suggesting SoA deficits may be a precursor of the psychotic state. This aberrant SoA has been suggested to stem from abnormal predictive sensorimotor processes, which have been causally shown to induce psychosis-like states. While there is converging evidence for SoA deficits in psychosis, the state of BO is less clear. While early reports reported abnormal processing of the RHI in SCZ, recent empirical and meta-analytical work has shown no differences between control and SCZ participants for BO. Classic psychedelics, which have been suggested to mimic psychosis symptoms, also induce striking alterations of self, including strange sensations of the bodily self e.g.,. These have been suggested to drive by changes in neural connectivity primarily due to the activation of 5HT2A receptors 98-100, for review see 24 . However, while subjective reports of acute and long-term changes in the sense of self after psychedelic use exist 101,102 , experimental evidence is conspicuously lacking. In this experiment, we compared BO & SoA using the MRHI in participants with psychosis, participants with a history of substantial psychedelic experiences, and neurotypical controls.
MATERIALS AND METHODS
Participants. Seventy-five, right-handed participants took part in the experiment (27 females; Age M = 29.9 SD = 8.85). There were three groups of participants: Control (N = 25; 16 females; Age M =24.84, SD = 6.28), Psychedelic (N = 25; 11 females; Age M = 28.8 SD = 6.01), and Psychosis (N = 25; all male; Age M = 36.08 SD = 10.39). The Control and Psychedelic groups included participants who self-reported no history of neurological, psychiatric, or tactile and motor disorders and any use of associated medications. Participants in the Psychedelic group were collected by approaching psychedelic social media groups and all participants reported psychedelic substance use, whereas in the Control group, they reported no use (see Supplementary Fig.). The Psychosis group included patients with psychosis hospitalized at the Beer Yaakov-Ness Ziona Mental Health Center (see Supplementary Table). The experiment was performed following the ethical standards of the Declaration of Helsinki and was approved by the Gonda Multidisciplinary Brain Research Center ethics committee (for the Control and Psychedelic groups) and by the Beer Yaakov-Ness Ziona Mental Health Center ethics committee (for the Psychosis group). All participants gave written informed consent before the experiment. The Psychosis group had a higher mean age than other groups, Wilcoxon rank-sum test (Psychosis vs. Control: W = 505.5, p < 0.001; Psychosis vs. Psychedelic: W = 447.5, p < 0.01).
ASSESSMENT OF PSYCHEDELIC USE AND CHANGES IN SELF.
The Psychedelic and Control groups self-reported their frequency of use of psychedelic substances as well as other substances (e.g., alcohol and other compounds see Supplementary Fig.). To assess changes in the sense of self following psychedelic use, we constructed a questionnaire examining altered self-experiences using 18 statements gathered from several well-established questionnaires (Ego-Dissolution Inventory (EDI), Cambridge Depersonalization Scale (CDS), and the Pahnke-Richards Mystical Experience Questionnaire (MEQ) 104 that examine the frequency and duration of experiences such as depersonalization, ego-dissolution, internal and external unity, and transcendence of time and space (see Supplementary Tablefor full questionnaire details). The questionnaire was completed by both the Psychedelic and Control groups. To assess whether psychedelic use is associated with increased altered self-experiences, we compared the questionnaire scores for each section (EDI, CDS, and MEQ) between the Psychedelic and Control groups using a Mann-Whitney test. Psychosis patients' clinical symptoms were assessed by a clinician via the Positive and Negative Syndrome Scale (PANSS) (see the clinical characteristics of Psychosis patients in Supplemental Table).
EXPERIMENTAL SETUP.
We used an adaptation of the MRHI setup to assess both BO and SoA. Participants sat at a table with their right hand hidden from their view and placed inside a box (35 cm × 25 cm × 12 cm). A realistic life-size and gendermatched rubber hand model was placed on top of the box and covered with a latex glove identical to the participants' glove. The participant's forearm was covered with a soft black cloth to ensure visual continuity of the rubber hand with the participant's arm (Fig.).
EXPERIMENTAL DESIGN.
Participants underwent two types of sensorimotor stimulation: Visuotactile (VT) and Visuomotor (VM). In the VT stimulation, as used in classical RHI experiments to induce illusory BO, participants were instructed to relax their hand while the experimenter brushed both the real and rubber hand with small brushes. In the VM stimulation, used in MRHI experiments to elicit an illusion of BO and SoA over the rubber hand, participants raised their index finger at a semi-regular rhythm of 1 Hz, which was practiced before the experiment. This movement resulted in a movement of the rubber hand by thin metal rod hidden inside the box. In both VT and VM stimulations, three different conditions were used. The brush strokes (i.e., VT) or active movement (i.e., VM) of the participant's index finger was accompanied by either anatomically congruent and temporally synchronous stimulation of the rubber hand's index finger (Sync), or anatomically incongruent and temporally synchronous stimulation of the rubber hand's middle finger (Incong), or anatomically congruent and asynchronous stimulation with a 500ms delay inserted between stimulation of the real and rubber hand (Async) (Fig.). In the VM Async condition, the metal rod was connected to the experimenter's finger, allowing the experimenter to manipulate the rubber hand's movement in a matter unknown to the participant. Each combination of stimulation type and condition was induced in a separate run lasting one minute, resulting in a total of six runs presented in a pseudorandomized order. After each run, participants completed the MRHI questionnairewhich assesses their subjective experiences of BO and SoA. The questionnaire comprises 12 statements, including 3 statements referring to BO, 3 to SoA, 3 BO control statements, and 3 SoA control statements (see Supplementary Tablefor the full list of statements). The control statements in the questionnaire do not capture the specific phenomenological experiences of BO or SoA, but rather have several similarities to the illusion-specific statements in general. Each statement was rated on a 7-point Likert scale ranging from -3 "totally disagree" to +3 "totally agree", with 0 indicating "neutral". Control and Psychedelic participants were presented with the statements on a computer screen and responded with a num-pad, while the Psychosis participants responded verbally.
FIGURE 1 -EXPERIMENTAL PARADIGM. (A)
Experimental design illustration of VM and VT stimulations, each divided for the Sync (synchronous coupling/stroking on both index fingers), Incong (synchronous coupling/stroking on the subject's index finger and the rubber hand's middle finger), and Async (coupling/stroking on both index fingers with a temporal delay) conditions. Red connecting lines in VM represent the mechanical connection between the subject's index finger and the rubber hands index finger in Sync, the middle finger in Incong, or the experimenter's finger for a delayed movement (Async). (B) Experimental procedure of a single run. In each run a single stimulation type (VT/VM) and condition (Sync /Incong / Async) was administered, after which the subject completed the MRHI questionnaire that consists of 12 statements probing SoA and BO. The order of runs was randomized between subjects.
DATA ANALYSIS.
Statistical analyses were performed using JASP 0.16.3 105 and R 106 . To estimate the main effects of Condition (Sync / Incong / Async), Group (Control / Psychedelic / Psychosis), and the interaction between them, we used mixed repeated measures ANOVA with Condition as a within-subject factor and Group as a betweensubjects factor. All post-hoc comparisons were Bonferroni corrected. For nonsignificant results, we performed equivalent Bayesian repeated-measures ANOVA to assess the evidence's strength of the null hypothesis. Specifically, we used a Bayes factor of exclusion. Briefly, BF exclusion was obtained by comparing the evidence of models that included the factor of interest to the evidence of models that did not include the factor. The models that were compared were of similar complexity including the same number of variables 107 . Thus, BF exclusion represents how likely it is to observe the data under models that exclude the factor compared to models that include this factor. BF values between 1 to 3, 3 and 10, and 10 and 30, were interpreted as anecdotal, moderate, and strong evidence for the hypothesis respectfully. The Shapiro-Wilk test was used to assess normality (p > 0.05), and the appropriate nonparametric tests were applied when one or more of the corresponding data sets failed to meet the criteria for normal distribution. To examine if a condition induced an illusion of BO or SoA over the rubber hand, we conducted a Wilcoxon signed-rank test by comparing the participants' ratings to 0 (i.e., neutral experience). Conversely, the control ratings were compared to 0 in the opposite direction using a Wilcoxon signedrank test, to ensure all were rated negatively below zero. To estimate the difference in BO illusion strength between the Healthy group (Control and Psychedelic) and the Psychosis group, we perform Mann-Whitney test. The illusion strength for each group was computed as the difference in BO subjective rating between Sync and Async and compared using a paired Wilcoxon signed rank. We used Spearman's correlations to estimate the association between the strength of SoA and BO illusion in each simulation.
PSYCHEDELIC EXPERIENCE IS ASSOCIATED WITH ALTERED SELF-EXPERIENCES
To examine reported differences in self-related experiences between the Control and Psychedelic groups, we compared the mean scores of their statements across sections of the self-experience questionnaire. The Psychedelic group's mean in the EDI statements (M = 7.28, SD = 2) was higher than the Control group's mean (M = 3.85, SD = 1.92) (W = 68, p < 0.0001). Also, in the MEQ statements, the Psychedelic group showed a higher mean (M = 8.27, SD = 1.65) than the Control group (M = 3.02, SD = 2.45) (W = 38, p < 0.0001; For the analysis of each question, see Supplementary Fig.and Table). Additionally, in the CDS statements the Psychedelic group's mean score (M = 5.33, SD = 1.99) was higher than the Control (M = 3.16, SD = 1.77) (W = 127, p < 0.001). These results support previous reports of the association between psychedelic experiences and changes in the sense of self.
REDUCED SENSE OF AGENCY DURING VISUOMOTOR STIMULATION IN PSYCHOSIS
As expected, VM stimulation induced an experience of SoA in the Sync condition (M = 1.81, SEM = 0.15, V = 2527, p < 0.001), the Incong condition (M = 0.85, SEM = 0.21, V = 2042, p < 0.001), where temporal coupling occurs, and was non-significant in the Async condition (M = -0.18, SEM = 0.21, p > 0.05; see Fig.). A main effect of Condition was found, (F(2,144) = 39.92, p < 0.001, ηp 2 = 0.35), and post-hoc comparisons showed significant differences between all the combinations of conditions (p < 0.001 for all pairs). Importantly, a main effect of Group was found (F(2,72) = 4.19, p < 0.05, ηp 2 = 0.1), and a post-hoc test revealed that this was driven by a significant difference between the Psychosis group and the Control group (p < 0.05), but was not significantly different between the Psychosis and Psychedelic groups (p = 0.09). A Bayesian analysis revealed moderate evidence supporting the lack of an effect between the Control and the Psychedelic groups (BF01 = 3.69). Despite lower SoA ratings of the Psychosis group, they were similarly affected by the different conditions in the VM stimulation, as can be seen by the lack of significant interaction between Group and Condition factors (F(4,144) = 0.714, p = 0.58). Bayesian analysis showed moderate evidence supporting the lack of an effect of this interaction (BF01 = 4.58). Notably, all the control statements ratings were negatively rated (p < 0.0001; Bonferroni Corrected). Thus, the Psychosis group showed reduced SoA compared to the Control group in the VM stimulation yet were comparably impacted by the modulation of anatomical congruence and temporal synchrony (Fig.).
REDUCED BODY OWNERSHIP DURING VISUOMOTOR STIMULATION IN PSYCHOSIS
As expected, VM stimulation induced an experience of BO in the Sync condition (M = 0.4, SEM = 0.21, V = 1665, p < 0.05), and a lower and non-significant in the Incong (M = -0.89, SEM = 0.19, V = 606, p = 1), and Async (M = -1.09, SEM = 0.2, V = 452, p = 1) conditions (Fig.). A main effect of Condition was found (F(2,144) = 30.15, p < 0.001, ηp 2 = 0.29). The main effect of the Group was not significant (F(2,72) = 0.202, p = 0.81). Bayesian ANOVA revealed a BF01 = 3.63, showing moderate evidence supporting the exclusion of Group factor. Notably, conditions affected the groups differently, as evidenced by a significant interaction between Condition and Group factors (F(4,144) = 2.96, p < 0.05, ηp 2 = 0.07). This interaction was driven by the lack of BO experience for the Psychosis group in the Sync condition (i.e., the group's mean is below 0). Importantly, the illusion strength of the Healthy group (Psychedelic and Control) was higher than the Psychosis group (Healthy: M = 2, SEM = 0.25; Psychosis: M = 0.52, SEM = 0.33; W = 916, p < 0.001). Additionally, the pairwise comparison between the typical difference of Sync and Async conditions was only significant in the Control (W = 252; p < 0.001) and Psychedelic (W = 290; p < 0.001) groups, but not significant in the Psychosis group (W = 107; p = 0.07). Notably, all the control statements ratings were negatively rated (p < 0.0001; Bonferroni Corrected). This lack of difference in the Sync and Async conditions, and diminished experience of BO illusion compared to the Healthy group, further supports our finding that the Psychosis group have impaired experience of BO under VM stimulation.
VISUOTACTILE STIMULATION INDUCED SIMILAR BODY OWNERSHIP ACROSS GROUPS
The VT stimulation is similar to the classic RHI, which is characterized by inducing BO over the rubber hand. As expected, the participants experienced BO over the rubber hand in the Sync condition (M = 0.97, SEM = 0.22, V = 2107.5, p < 0.001). The BO ratings were non-significant and gradually declined in the Incong (M = -0.03, SEM = 0.22, V = 1282, p = 0.49), and Async (M = -0.91, SEM = 0.19, V = 579, p = 1) conditions (Fig.). This was supported by a significant main effect of Condition (F(2,144) = 31.52, p < 0.001, ηp 2 = 0.3), and post-hoc comparisons which showed significant differences between all possible combinations of conditions (p < 0.001 for all pairs). Comparing BO between groups, we did not find significant differences between Control, Psychedelic, and Psychosis groups across the different conditions. The main effect of Group was not significant (F(2,72) = 0.43, p = 0.64). Bayesian ANOVA revealed moderate (BF01 = 8) evidence for excluding Group factor. Finally, the interaction of Group and Condition was not significant (F(4,144) = 1.45, p = 0.22, η2 = 0.04), and Bayesian analysis revealed moderate evidence for exclusion of the interaction factor (BF01 = 8.57). Notably, all the control statements ratings were negatively rated (p < 0.0001; Bonferroni Corrected). Thus, no significant differences between the Psychosis, Psychedelic and Control groups were found in the VT induction of illusory BO (Fig.) and moderate evidence against such differences was present.
VISUOTACTILE STIMULATION DID NOT INDUCE SENSE OF AGENCY ACROSS GROUPS
In general, the VT stimulation did not induce SoA experience over the rubber hand (Sync: M = -0.87, SEM = 0.21, V = 743, p = 1; Incong: M = -1.33, SEM = 0.18, V = 345, p = 1; Async: M = -1.85, SEM = 0.16, V = 154, p = 1; see Fig.). The repeated measures ANOVA revealed a main effect of Condition (F(2,144) = 12.04, p < 0.001, ηp 2 = 0.14). A post-hoc Bonferroni test showed significant differences between all comparisons (p < 0.001). The main effect of the Group was not significant (F(2,72) = 0.16, p = 0.85). Bayesian ANOVA revealed moderate (BF01 = 6.83) evidence for excluding Group factor. We did not find a significant interaction between Group and Condition (F(4,144) = 2.13, p = 0.08, η 2 = 0.05). Bayesian analysis revealed moderate evidence for excluding the interaction factor (BF01 = 3.49). Notably, all the control statements ratings were negatively rated (p < 0.0001; Bonferroni Corrected). Thus, in the absence of active movement, no experience of SoA over the rubber hand was formed, and this was similar across groups.
POSITIVE CORRELATION BETWEEN BO AND SOA
We wanted to determine whether BO experience was correlated with SoA experience in both simulations (VM and VT). To examine it we calculated the subtraction between Sync and Async for BO and SoA for each stimulation. The differences from the subtraction which represent the BO and SoA illusion strength. BO and SoA illusions in VM were significantly correlated (r = 0.34, p < 0.01; see Fig.). Likewise, BO and SoA illusions in VT were also significantly correlated (r = 0.51, p < 0.001; see Fig.). These results stand with the idea that when one experiences SoA over the rubber hand, one also tends to experience BO over the rubber hand.
DISCUSSION
The current study aimed at experimentally comparing how the bodily self is impacted in populations with modulations of the sense of self driven by psychotic or psychedelic experiences. Using the Moving Rubber hand Illusion, we tested both BO and the SoA in these populations. Our study revealed several interesting findings. First, as expected all populations showed modulations of BO using the classical VT stimulation. Second, the Psychosis group showed a reduced SoA during VM stimulation. Third, in the Control and Psychedelic groups, synchronous VM stimulation induced BO, which was absent in the Psychosis group. Finally, while the Psychedelic group showed strong effects of a history of altered self-experiences, we found evidence that there are no lasting modulations of the bodily self in this group. As expected, synchronous VT stimulation induced illusory BO over the rubber hand. No significant differences were found between the Control, Psychedelic, and Psychosis groups and Bayesian analysis revealed moderate evidence (BF = 8) against differences between groups. This is in line with previous experimental and metaanalytical data showing no differences between Control and Psychosis groups in multisensory BO processing. Similarly, this suggests that psychedelic experiences which often include striking modification of the phenomenological experiences of BO, do not cause long-lasting alterations of multisensory bodily processing. Previous theoretical and experimental work has highlighted abnormal SoA and predictive sensorimotor mechanisms as a central impairment across the SCZ spectrum. For example, irregularities in SoA attribution have been shown in neurotypical individuals with high psychotypical scoresand even in healthy individuals with a genetic propensity for SCZ. In acute psychosis abnormalities in SoA processing are pronounced with both low accuracy in judgments and aberrant meta-cognition e.g.,. Indeed, in a recent study embodied agency judgment allowed to classify psychotic participants with a high degree of accuracy. The current results show low sensations of agency in the Psychosis group across all conditions indicating that they were aware of the conditions but felt a reduced sensation of control over the movements. This finding is in line with the passivity experiences in schizophrenia, which involve a reduction in SoA. Furthermore, previous studies using body illusions, revealed that individuals on the SCZ spectrum experience a reduction in SoA similar to healthy individuals when exposed to VT/VM asynchronous stimulations, as evidenced by decreased SoA compared to synchronous stimulations. Synchronous VM stimulation also induced sensations of BO in the Control and Psychedelic groups but not the Psychosis group. Several studies have shown that SoA and BO are often correlated, suggesting that they impact each other. Reduced BO during synchronous VM stimulation suggests that aberrant sensations of bodily self in psychosis may stem not from deficits in BO processing per se see 94 but rather from reduced SoA which in turn impacts the feeling of ownership over body parts. Thus, while the Psychosis group showed typical BO and SoA in the VT conditions, these experiences were reduced specifically in VM conditions. This result joins previous evidence that the experience of the bodily self is disrupted in psychosis patients. Nonetheless, our findings suggest that these altered bodily experiences in psychosis patients are not a global impairment, but rather related to aberrant processing of volitional actions. Importantly, our findings highlight that while psychedelic experiences are associated with dramatic changes in the sense of self, including the bodily self, which were not associated with experimentally induced changes in BO or SoA. In fact, across all experimental conditions the Psychedelic group's results were nearly identical to those of the Control group. This finding is especially salient in light of the clear differences in subjective experiences of modifications of the self between these groups (See Fig.and Supplementary Table). This suggests that while some changes to the sense of self induced by psychedelic experiences are retained at explicit levels of the selfmodel, fundamental sensorimotor processing of the bodily self is not altered in an enduring manner. Thus, our data provide evidence against psychotomimetic approaches to psychedelic experiences at least for bodily self-processing, in such that the Psychosis and Psychedelic groups showed no similarities in their results.
LIMITATIONS
Our study has several limitations which should be addressed. First, due to the age limits of the population at the mental health center, the psychotic group was on average older than the other groups. Nonetheless, age differences were previously not found to affect the illusion strength 122 . Second, because the experiment was conducted in a male only ward of the Beer Yaacov-Ness Ziona mental health center, the Psychosis group consisted only of men. Third, due to the time limit for an experiment on the population with psychosis, the psychosis group didn't complete the questionnaire examining altered self-experiences. Finally, due to ethical and experimental limitations, we could not collect psychedelic participants in the acute stage of the psychedelic experience, limiting any inferences regarding their acute effects. Further studies are needed to explore how different psychedelic compounds impact the sense of self, during and after the psychedelic experience.
CONCLUSIONS
Our study shows that psychosis impacts the bodily self with diminished agency and body ownership during volitional actions. Contrarily, while substantial psychedelic use was associated with reported enduring changes in the experience of self, it showed no impact on processing of the bodily self. Our results suggest that even considerable psychedelic use doesn't alter multisensory bodily processing underlying the sense of self, but may manifest changes in more explicit, narrative models of the self.
Study Details
- Study Typeindividual
- Populationhumans
- Characteristicsobservational
- Journal