Optimised Multi-Channel Transcranial Direct Current Stimulation (MtDCS) Reveals Differential Involvement of the Right-Ventrolateral Prefrontal Cortex (rVLPFC) and Insular Complex in those Predisposed to Aberrant Experiences

Research has shown a prominent role for cortical hyperexcitability underlying aberrant perceptions, hallucinations, and distortions in human conscious experience – even in neurotypical groups. The rVLPFC has been identified as an important structure in mediating cognitive affective states / feeling conscious states. The current study examined the involvement of the rVLPFC in mediating cognitive affective states in those predisposed to aberrant experiences in the neuro-typical population. Participants completed two trait-based measures: (i) the Cortical Hyperex-citability Index_II (CHi_II, a proxy measure of cortical hyperexcitability) and (ii) two factors from the Cambridge Depersonalisation Scale (CDS). An optimised 7-channel MtDCS montage for stimulation conditions (Anodal, Cathodal and Sham) was created targeting the rVLPFC in a single-blind study. At the end of each stimulation session, participants completed a body-threat task (BTAB) while skin conductance responses (SCRs) and psychological responses were recorded. Participants with signs of increasing cortical hyperexcitability showed significant suppression of SCRs in the Cathodal stimulation relative to the Anodal and sSham conditions. Those high on the trait-based measures of depersonalisation-like experiences failed to show reliable effects. Collectively, the findings suggest that baseline brain states can mediate the effects of neuro-stimulation which would be missed via sample level averaging and without appropriate measures for stratifying individual differences.

What is particularly striking is that elevated levels of cortical hyperexcitability have also been revealed in neurotypical groups in the complete absence of a tractable underlying pathology (Ohayon, 2000;Johns & Van Os, 2001;Barkus, et al., 2007;Braithwaite et al., 2011;Diederen et al., 2012;Braithwaite, Broglia, et al., 2013;Preti et al., 2014;Braithwaite, Marchant, et al., 2015;Braithwaite, Mevorach, et al., 2015;Kråkvik et al., 2015;McGrath et al., 2015;Van Os & Reninghaus, 2016;Baumeister et al., 2017).The emerging view is one of a continuum of cortical hyperexcitability / predisposition to aberrant perceptions along which individuals can be placed (Tien, 1991;Van Os et al., 2009).Crucially, the presence of such experiences in neurotypical groups provides insight not only into the characteristics of aberrant experience that require explanation, but also more fundamental aspects of human consciousness.
According to the Diagnostic and Statistical Manual 5th Edition / DSM 5, depersonalisation disorder (DPD) is a clinical disorder characterised by dissociative experiences where the patient typically describes a feeling of unreality for the bodily self (depersonalisation), unreality of their surroundings (derealisation) or both (Sierra & David, 2011;American Psychiatric Association, 2013; see also Sierra, 2009 for a review).Patients report feelings of estrangement from themselves; remoteness from their bodies, thoughts, and actions, coupled to a radically altered sense of 'presence' and a dampening of emotional affect (numbness).Collectively, DPD reflects what is, in essence, a profound shift and change in self-consciousnessa sense of 'feeling unreal' (Sierra & Berrios, 1998, 2000;Sierra et al., 2005;Sierra, 2009;Sierra & David, 2011;Medford, 2012;Seth et al., 2012;Seth, 2013;Clark, 2013).DPD can occur comorbidly with other conditions, but also exists in its own right as a specific disorder.
Importantly, unless occurring co-morbidly with other conditions or disorders, DPD is not typically associated with sensory hallucinations or delusions (reality-testing is left intact; Sierra, 2009;Sierra & David, 2011).Instead, the aberrant perceptions are more accurately defined as 'distortions' in human experience as opposed to perceiving something which has no external reference (e.g., hallucination).
As with the concept of cortical hyperexcitability, "depersonalisation-like experiences" (DLEs) are also reported in the neurotypical/ nonclinical population (Sierra, 2009;Dewe et al., 2016Dewe et al., , 2018;;Sierra & David, 2011;Braithwaite et al., 2020).The prevalence and occurrence of DLE's in the neurotypical population occurs with estimated lifetime rates of between 26 and 74% (Hunter et al., 2004;Michal et al., 2009;Sierra & David;2011).Accordingly, these experiences reflect many of the thematic components of DPD, albeit in attenuated form (but are no less striking to the observer).Again, aberrant body experiences are a core component of DLEs (Craig, 2009;Sierra & David, 2011;Seth, 2012Seth, , 2013;;Clark, 2013;Jay et al., 2014Jay et al., , 2016)).Conceptually, DLEs themselves can be seen as a continuum, similar to that of cortical hyperexcitability and the concept of Schizotypy (Claridge, 1997), where the frequency and intensity of such experiences can vary on an individual basis.The relationship between DLEs and DPD and whether the presence of increasing DLEs represents a vulnerability to transition to disorder awaits clarification.
As a net consequence of these interactions, affective 'feeling states' are divorced from colouring the typical integration between perception and cognition resulting in an attenuated emotional experience, a reduced sense of presence, subjective feelings of 'unreality' and profound alterations in self-consciousness.The above suggestions are also supported by examinations revealing major neuronal projections from the rVLPFC into the AIC and functional imaging, showing that activity in these areas display a functional interdependencewhere increased activity in rVLPFC occurred in concert with significantly decreased activation (or absence of activation) in the AIC (see Craig, 2009;Uddin, 2015;Trambaiolli et al., 2022).
Behavioural, brain-imaging and neurostimulation research also support the contention that the rVLPFC has an aberrant overinhibitory role over the AIC in patients with DPD.Jay et al. (2014) used low-frequency (1 Hz) repetitive transcranial magnetic stimulation (rTMS) directed at the rVLPFC with DPD patients.The underlying rationale here was to inhibit the brain networks that were thought to be over-inhibiting the AIC, consequently liberating other brain regions receiving major projections from it (the AIC) from such suppressive modulation.Jay et al. (2014) found that the symptoms of DPD were alleviated by low frequency rTMS, as the brain stimulation had an inhibitory effect on inhibitory neural processes.In essence, inhibiting an inhibitory system that is functionally (and anatomically) connected to another neural system, releases these other networks, to function at a more typical level.
Surprisingly, despite aberrant body experiences being identified as a core component of DPD, until recently, body-based aversive imagery or threats have not been used to examine how such stimulation interacts with autonomic responding in such groups.Accordingly, the role of these networks in neurotypical levels of DLEs is unclear.Dewe et al. (2016) (see also, Dewe et al., 2018) were the first to examine DLEs in neurotypical samples where rather than viewing generic images, the observer's own physical body received a threat (via a movie prop syringe).This study provided evidence of suppressed skin conductance responses (SCRs) and autonomic responding to such threats and the magnitude of the suppression was associated with the strength of the DLEs reported.In a parallel development, Braithwaite et al. (2020) introduced a novel Body Threat Assessment Battery (BTAB) that contains dynamic stimuli specifically showing body and non-body (baseline) aversive threats and correspondingly elicit increased autonomic activity (measured by skin conductance responses) to further research body-based processing (and aberrations therein).
Although numerous demonstrations of the efficacy of tDCS have been reported, the underlying biophysics of tDCS remains somewhat unclarified.It is generally thought that tDCS exerts its effects by modifying spontaneous neuronal activity via shifting the resting membrane potential in a polarity-dependent manner.By this account, anodal (excitatory) stimulation induces an increase in the background spontaneous firing rate by moving cell membranes more towards a depolarized state thereby making it more likely that they fire.In contrast, cathodal (inhibitory) stimulation reduces cortical excitability by moving cells more towards a hyperpolarized state making them less likely to fire (Lauro et al., 2014).
These observations highlight that it is becoming increasingly important to measure and take into account potential individual differences in baseline excitability (Benwell et al., 2015;Hsu et al., 2016;Fertonani & Miniussi, 2017;Juan et al., 2017;Dubreuil-Vall et al., 2019) because the differential effects of tDCS may be due to differences in the excitatory/inhibitory (E/I) balance across cortical regions and layers and within their network dynamics (Boroojerdi et al., 2000;Jacobson et al., 2012;Alekseichuk et al., 2016;D'Souza et al., 2016;Romei et al., 2016;De Graaf et al., 2017;Fertonani & Miniussi, 2017;Silvanto et al., 2018;Yang & Sun, 2018).Differences in latent cortical excitability across individuals could thus create heterogeneity in both individual predisposition to aberrant experiences and their responses to tDCS.However, many previous tDCS studies have focused their analysis at the whole sample level without accounting for background trait or state factors.This may explain, at least in part, failures to replicate findings and why some meta-analyses have failed to find significant tDCS effects (Horvath et al., 2015;Medina & Cason, 2017).Not accounting for baseline excitability may mask or indeed miss subtle, though important, interactions between baseline brain states and stimulation.Current evidence suggests that these network-level interactions play a critical role in mediating the response to low-amplitude brain stimulation (Dmochowski et al., 2011;Miranda et al., 2013;Fox et al, 2014;Ruffini et al., 2014;Kunze et al., 2016;daSilva et al., 2018).
Recent advances and development in technology has resulted in optimised multi-electrode arrays consisting of several anode and cathode configurations, which can be used to simultaneously modulate regions of a distributed functional network model based on electroencephalograph (EEG) systems (Miranda et al., 2013;Ruffini et al., 2014;Fischer et al., 2017;Ruffini et al., 2018).The advantages from this approach are of both methodological and theoretical importance.For example, compared to standard bipolar methods that utilise large electrodes, a model-driven stimulation design using small electrodes and realistic head models for estimation of current flow, will lead to superior focality and spatial resolution, helping to ensure that the stimulation occurs maximally within the targeted networks and minimally affecting other brain areas (Ruffini et al., 2014(Ruffini et al., , 2018)).The increased quantitative properties of multi-channel tDCS (MtDCS) lend themselves to a more precise theory-building process with regards to the targeted brain networks and their ascribed functions.

Overview:The present study
The present study examined the role of the rVLPFC and its inferred functional relationship with the insula complex in mediating cognitive-affective responses to aversive body-threat stimuli in a neurotypical sample.Participants were measured for: (i) predisposition to aberrant experiences thought to reflect increased degrees of visual cortical hyperexcitability, and (ii) proneness to DLEs.The role of the rVLPFC was assessed via a targeted and optimised MtDCS montage.The montage consisted of multiple anode and cathode electrodes to stimulate this brain region which participates within a distributed functional network.Using electric field modelling techniques, optimal parameters for MtDCS montages were determined (stimulation current and location of all electrodes) to directly identify the optimal targets and parameters for accurate brain stimulation (Miranda et al., 2013;Ruffini et al., 2014Ruffini et al., , 2018)), using a realistic template head model.
In line with previous brain-stimulation findings, brain-imaging and known neuroanatomical pathways, it was assumed that stimulating the rVLPFC would impact the operation of the insula cortex (a region identified as important in saliency networks, default networks, the generation of conscious feeling/states, interoceptive awareness, predictive-coding, and the mediation of autonomic SCRs (Critchley, 2005;Medford et al., 2006;Lemche et al., 2007Lemche et al., , 2008;;Jay et al.. 2014;Xia et al., 2017;Vinberg et al., 2021) and this would primarily manifest in autonomic SCR responses to body-threat stimuli.
In addition, consistent with findings on transcranial electrical stimulation (tES) (Horvath et al., 2015;Hsu et al., 2016;Romei et al., 2016;Medina & Cason, 2017;Silvanto et al., 2018), it was expected that the efficacy of MtDCS stimulation would not be all-or-none and that stimulation could be mediated further by individual differences in trait-based signs of cortical hyperexcitability.Furthermore, the study was extended to explore if these brain regions may also be implicated in mediating DLEs in neurotypical samples in line with what has been reported for clinically diagnosed depersonalised patients (Jay et al., 2014(Jay et al., , 2016)).Collectively, the present study sought to quantify trait-based factors pertaining to aberrant experience (potentially reflecting cortical hyperexcitability) by stratifying participants on a recently devised measure (the Cortical Hyperexcitability Index II -CHi_II - Fong et al, 2019;Braithwaite, Marchant, et al., 2015) and on particular factors pertaining to DLEs (the aberrant body experience and alienation from surroundings factors of the Cambridge Depersonalization Scale: Sierra & Berrios, 2000;Sierra et al., 2005).
Previous work on the CHi_II measure revealed a 3-factor solution; "Heightened Visual Sensitivity and Discomfort" (HVSD), Auralike Hallucinatory Experiences" (AHE) and "Distorted Visual Perception" (DVP) (Fong et al., 2019).Neurophysiological investigations (electroencephalogram EEG / visual evoked potentials VEP) and behavioural work (Pattern-glare tasks) have validated the underlying assumptions of the measure with migraine patients and neurotypical samples showing different performance patterns as a function of the different factors (Fong et al., 2019(Fong et al., , 2020)).This work also demonstrated that the factor "AHE" best predicted underlying aberrant degrees of cortical hyperexcitability.Therefore, it was taken a-priori as the primary indicator of aberrant hallucinatory experiences most likely mediated by central cortical processes.This factor was used to examine trait-based estimates of baseline neural states mediating the efficacy of the MtDCS procedure.
The current study is also the first to utilise optimised MtDCS to influence the activity in the rVLPFC and indirectly the insula complex to mediate cognitive-affective processes.For this purpose, a specific 7-channel optimised montage was created to target the rVLPFC.
In summary, this study examined the presence of trait-based predisposition to aberrant experience (reflecting increased cortical hyperexcitability) in mediating cognitive-affective responses as a result of optimised MtDCS directed at the rVLPFC.In addition, the sample was also screened for predisposition to DLEsas these have been thought to reflect aberrant processing in rVLPFC and insula regions.Psychophysiological (skin conductance responses) and psychological responses (ratings) were quantified where participants viewed a novel computer-based task depicting body-threat scenarios (the BTAB: Braithwaite et al, 2020) under different MtDCS conditions.

Participants
Thirty-five participants were recruited from Lancaster University, Department of Psychology, UK.Participants' ages were between 18 and 25 years (M = 19 years, SD = 1.48) of which, 22 were Female and 13 were Male.Safety and exclusion criteria prevented participants with debilitating fear of blood/gore/needles, any fitted electrical/medical device, a history of psychiatric/dissociative diagnoses as well as epilepsy/fainting/seizures of unknown origin from taking part.The study was approved by the Lancaster University Ethics Committee (FST19024).All volunteers were compensated for their time with course credits.
A-priori sample size estimation could not be straightforwardly based on a power analysis of previous work because of the novelty of the methods used in the current study and their novel combined use.Nonetheless, a-priori estimations of 30 -35 participants were based on i) exceeding the sample sizes typically used in previous tDCS research (e.g., 13 to 16 participants, e.g., Antal et al., 2003Antal et al., , 2011;;Peña-Gómez et al., 2011), and ii) a consideration that testing half the sample size of our previous brain stimulation research (Braithwaite, Mevorach, et al., 2015) was appropriate given the fact that MtDCS is a far superior method of brain stimulation (focally and in terms of intensity).In addition, the use of Bayes Facor analyses provided further information regarding the strength of evidence for or against the effect of our independent variables (see Results section).
Out of the 35 participants that took part, two participants could not continue due to initial failed impedance (safety) checks on the MtDCS electrodes.A further 5 participants were excluded from analysis as they did not complete all required sessions and thus failed to produce fully balanced data.An additional participant was also removed from analysis as SCR non-responder based on established guidelines (Dawson et al., 2007;Braithwaite, Watson et al., 2013;Boucsein et al., 2012;Braithwaite et al., 2020).The final sample used for analysis consisted of 27 participants, 10 male and 17 female (M Age = 19 years, SD = 1.53).

Screening measures
Two validated screening questionnaires were used to measure individual predisposition to aberrant experiences associated with cortical hyperexcitability and depersonalisation-like experiences.
Cortical Hyperexcitability Index II (CHi_II).This measure examines a collection of aberrant visual experiences thought to reflect the presence of cortical hyperexcitability (Fong et al., 2019).It has high internal consistency and reliability (Cronbach Alpha 0.90).An exploratory factor analysis revealed a three-factor solution as being the most parsimonious: these were "Heightened Visual Sensitivity and Discomfort" (HVSD), "Aura-like Hallucinatory Experiences" (AHE) and "Distorted Visual Perception" (DVP).All factors had loadings of over 0.40 and there were no cross-loadings of the items.For the present study, we focussed on the "AHE" factor as a measure of cortical hyperexcitability as this arguably represents more centrally mediated hallucinatory experiences and has shown greater predictive power with neurophysiological EEG measures (Fong et al., 2019(Fong et al., , 2020)).Therefore, the AHE factor was used to stratify participants in terms of their baseline cortical hyperexcitability levels and hence to examine the efficacy of MtDCS brain stimulation.
Cambridge Depersonalisation Scale (CDS) -Anomalous Bodily Experiences and Alienation from Surroundings.The CDS examines susceptibility to aberrant and dissociative experiences related to depersonalisation disorder (Sierra & Berrios 2000;Sierra et al. (2005).Previous work has identified four factors (Sierra et al., 2005) two of which Anomalous Bodily Experiences: (ABE) and Alienation from Surroundings: (AFS) have been shown to work well in neurotypical populations and these factors represent the more core components of DPD, so only these factors were measured and then pooled into one indicator of "depersonalisation-like-experiences" (DLEs: see Braithwaite, Broglia et al., 2013;Dewe et al., 2016Dewe et al., , 2018;;Braithwaite et al., 2020 for a similar approach).The combined range of scores of the DLE factor (which contains 13 items in total) is 0-130.
The screening measures (CHi_II and DLE) were standardised by dividing their sum by the number of items on that factor (to control for different number of items on each factor).

Multi-channel transcranial direct current stimulation (MtDCS)
An optimised 7-channel montage was created with a weighted target map specifically designed to target the rVLPFC while leaving the rest of the brain maximally unaffected.This was done initially by highlighting the rVLPFC region via the Stimtargeter tool (Neuroelectrics, Barcelona, ESP) to create a weighted target map for montage optimisation performed by the Stimweaver multi-focal algorithm (Neuroelectrics, Barcelona, ESP) which revealed an optimised solution for the target region (Ruffini et al., 2012(Ruffini et al., , 2014(Ruffini et al., , 2018;;Ho et al., 2016).The solutions were modelled on a standard generic reference brain, which was also used to determine the optimised montage.
The design of this focal montage was based on a weighted target map reflecting with a strong weight (W = 10) the desired normal electric field on the rVLPFC cortical surface (E = 0.25 V/m) with the additional requirement of zero electric field on the rest of the brain (W = 10).The final optimization chosen was a montage with the best goodness of fit metric ERNI (error with respect to no intervention) for focusing on the target required to be stimulated [resulting in WCC (Weighted cross correlation of target map and delivered field) = 0.53 and ERNI = − 4245 mV 2 /m 2 -Fig.1].The high definition headcap provides 64 points based on a subset of the 10-10 EEG system (Seeck et al., 2017) for electrode positioning.The electrodes identified for optimal ERNI were AF8, F6, F8, AF4, FC4, FP1, T8 for stimulation (see Fig. 2 and Table 1).The polarity of the electrode combination was determined by the condition (Anodal or Cathodal), i.e., the cathodal condition montage was obtained by reversing the currents in the anodal condition.For safety, during optimization, the total output was constrained to not exceed 2 mA and each stimulation electrode was capped at a maximum current of 1 mA.The stimulator used was Starstim 8 controlled by the NIC2 Software (Neuroelectrics, Barcelona, ESP).
Stimulation occurred for 11 mins in total per session which included a ramp-up (30secs) and ramp-down period (30secs) of electrical current (10mins of full stimulation).In the Sham condition, the stimulator only used the ramp-up and ramp-down protocol to mimic the experience of stimulation.Each participant completed three neuromodulation sessions (Anodal, Cathodal and Sham) in a randomised order and were blind1 to the nature of the conditions.Sessions were a minimum of one week apart to wash out any potential carry-over stimulation effects across sessions.

Table 1
The optimised weighted channels for the Anodal condition from the optimised montage used for the rVLPFC site.

Stimulation Site
Current per electrode (µA) Note: Polarities of the applied current per electrode were reversed for the Cathodal condition.

The Body Threat Assessment Battery (BTAB)
The BTAB consists of a selection of high-definition dynamic clips depicting various threats directed to a human body and non-body baselines (Braithwaite et al. 2020).The original task contained 12 body-threat clips (6 from an Egocentric perspective and 6 from Exocentric perspective) and 3 non-body baseline stimuli.In the original study, each clip was shown individually, and normative psychological ratings (arousal, valence, sense of pain, and realism of threat) and psychophysiological responses (Specific threat-related SCRs and non-specific SCRs) were determined (Braithwaite et al., 2020).In addition, at the beginning of the threat clips a 'set-up shot' (an upper body/torso) was included to avoid startle responses and artefacts in the SCRs at the start of stimuli presentation.
The present study used a modified a blocked design (programmed in E-Prime 3.0) where the clips (i.e., a series of threats) were grouped together with the psychological ratings coming at the end of this short series of clips.In each session, participants viewed a total of 10 stimuli separated into three blocks (2 baseline, 4 Ego Threat and 4 Exo Threat).The clips chosen were pseudo-randomised according to the X SCR (µS) values (where X = mean of difference between SCR onset and its maximum peak and µS = microseimens) described in Braithwaite et al. (2020) so that they were matched in terms of autonomic potency.The 3 blocks of stimuli were randomised within E-Prime 3.0.
Stimuli were presented on a 22 in., 16:9 aspect ratio monitor at 1920 × 1080 resolution, in a darkened room with a viewing distance of ~ 80-90 cm.Additionally, to avoid unfamiliarity to the procedure and stimuli, a practice trial based on a neutral stimulus (body-based setup with a non-threatening stimulus i.e., brush stroking and arm) was shown along with an example on how to complete the rating scale (arousal, valence, pain and realism of threat) questions before stimulation began (see Fig. 3).The total time taken for participants to complete this task was ~ 7 min (5 min to view all video clip blocks and ~ 2 min to answer the psychology self-report ratings).

Psychological self-report ratings
During the BTAB task, participants were asked to report their experiences after each block had finished.These subjective ratings were based on a Likert-type scale across 4 dimensions, namely, emotional valence (− 5 to +5), emotional arousal (0 to 9), experience of illusionary/sensory pain (0 to 9) and realism of threat (0 to 9).Fig. 3.A schematic illustration of the present procedure completed in Session 1. Note: The screening measures were not repeated in Sessions 2 and 3 as these were trait-based measures and were not dependent on the stimulation.

Skin conductance responses (SCRs)
SCRs were obtained using a BIOPAC MP36R data-acquisition unit (BIOPAC Systems Inc., Goleta, CA).This was connected to a PC running 64-bit Windows 10 Home.All signals were recorded with a 0.05 Hz high-pass filter and sampled at 2000 Hz.Data were collected by applying a small continuous low voltage (0.5 V) current through two disposable pre-gelled EL507 electrodes attached to SS57L sensor leads.The electrodes and leads were attached to the distal phalanges of the index and middle finger on the left hand of the participant.These were attached 10 mins before data was acquired to obtain the clearest/high quality signal.
All SCR responses were gathered and processed in BIOPAC AcqKnowledge v5.0 (BIOPAC Systems Inc., Goleta, CA).An SCR was defined as a magnitude delta function (µS), between the peak value and SCR onset (for more detail, see Braithwaite, Watson et al., 2013;Braithwaite et al., 2020).Where there were no SCRs detected (for a given block), a zero value was recorded.As per Braithwaite et al, (2013), the SCR threshold was set at 0.01 µS.The SCR of interest was defined as the largest/strongest response that occurred during the presentation of a stimulus.All other SCRs during the presentation of the dynamic clips were classed as non-specific SCRs (NS-SCRs) which were analysed for their frequency (F-SCR) and strength as additional measures of autonomic arousal (Nikula, 1991;Boucsein, 2012;Braithwaite et al., 2013b).
All signals were visually inspected for artefacts and when encountered, the signal was down sampled by 200 samples/sec (iteratively) to remove them in that section of the signal.In line with recommended practice, SCR data were normalised by using [SCR (Log + 1)] transformations and standardized by converting to Z-Scores (Dawson et al., 2007;Boucsein et al., 2012;Braithwaite, Watson, et al. 2013;Braithwaite & Watson, 2015;Braithwaite et al., 2020).Previous research (Braithwaite et al., 2020) showed that the two video perspectives (Ego and Exo) were equally efficient at eliciting strong responses and so responses to both viewpoints were combined (average of the largest SCR in each block).Additionally, the Baseline block SCRs, NS-SCRs and F-SCRs were merged into a single Baseline value for analysis.

Procedure
Each participant began the study by completing the trait-based screening measures (CHi_II and DLE) with order randomised across participants.Following this, participants were shown the practice trial for the BTAB, during which they could ask questions regarding the task.Next, the participants were setup for the MtDCS protocol, which was done in two steps.First, the cranial perimeter / head circumference was measured to find a suitably sized headcap.Second, the headcap was positioned by ensuring the Nasion electrode (FPZ), Central electrode (CZ), Inion electrode (IZ) and the preauricular points (T7 and T8) were correctly aligned on the participant's head (the CZ corresponded to the vertex of the participants' head).Then, the water-soluble electrode gel was injected into each of the 7 stimulation electrodes and subsequently connected to the stimulator (StarStim Necbox 8 -Neuroelectrics, Barcelona, ESP).Safety checks were conducted by confirming the earthing clip (on the earlobe) was secure and the electrodes were guaranteed to be within normal impedance levels (0 -10 kΩ).In addition, during this stage, the electrodes for the SCRs were attached to the index and middle distal phalanges.Once completed, the stimulator was turned on and stimulation was delivered for 11 mins (including 30 s ramp-up and ramp-down) where safe impedance levels were continuously monitored.During stimulation, participants were asked not to make extraneous movements and to immediately report any adverse effects.Directly after stimulation had ended, the BTAB task began, during which SCRs were recorded (~7 min).This was followed by an exit-questionnaire to ensure that participants did not experience any severely adverse effects2 during the stimulation.This was repeated for each participant across all sessions (see Fig. 3).Each session took approximately 1.5 h per participant.

Overall statistics
Analyses were conducted using SPSS v27 and JASP v0.14.1.Non-parametric tests (Mann-Whitney U independent samples t-tests, Freidman's χ 2 tests, Kendall's Tau correlations) were conducted for non-normal data.Independent sample t-tests, paired samples ttests, two-way mixed ANOVAs, within-subject ANOVAs and Pearson's two-tailed r correlations were used for the self-report measures to distinguish between the Anodal, Cathodal and Sham data.Where sphericity was violated with an ε of <0.75, a Greenhouse-Geiser correction was used and >0.75, a Huynh-Feldt correction was used.Additionally, all multiple testing was corrected using the False Discovery Rate (FDR) method, FDR correct p values are indicated as B&H in the analysis, if the uncorrected p value is less than the B&H corrected p value, the comparison is considered significant (Benjamini & Hochberg, 1995, see also -Braithwaite et al., 2020).
Comparisons between trait-based predisposition to CHi_II and DLE and responses based on stimulus presentation (psychophysiological and psychological) were achieved through correlational analysis.Follow up analysis was conducted using median splits to identify high and low scorers on each trait-based predisposition screening measure to clarify the pattern of effects for the different stimulation conditions.
Where appropriate, Bayesian statistics (using default Cauchy priors) are reported alongside Frequentist statistics (p values).Bayes Factor analysis (BF) not only informs about the alternative hypothesis (BF 10 > 1) but also indicates strength of the null hypothesis (BF 10 < 1).In accordance with Jarosz and Wiley (2014), a BF 10 between 1-3 would mean the data are inconclusive, 3-10 shows good evidence for the alternative, 10-100 as strong and >100 as decisive evidence.For clarity, both BF 10 (in support of the alternative) and corresponding BF 01 (support for the null) are provided.
When analysed at the whole sample level, no reliable differences were observed in the SCRs across the three conditions (Anodal Threat, Cathodal Threat, Sham Threat) by a within-subject ANOVA; F (2,52) = 0.663, p > 0.05, BF 10 = 0.204, BF 01 = 4.906.To summarise, in line with previous research, the present sample replicated and showed significantly increased SCR responses to the threat stimuli, relative to baseline stimuli, evidencing that the stimuli were effective at inducing increased autonomic responses.In addition, when viewed across the whole sample, MtDCS did not appear to significantly impact autonomic processes.

BTAB magnitudes of NS-SCRs
Similarly, an analysis of the magnitudes of all non-specific SCRs (NS-SCRs: which were all other SCRs, except the maximum one, that occurred during the viewing of the clip), which provide an index of background autonomic arousal, was conducted using a withinsubject ANOVA at 4 levels (Baseline, Anodal Threat, Cathodal Threat and Sham Threat) (Fig. 5).There was a significant main effect of
Taken together, these findings show evidence for the efficacy of BTAB in distinguishing between non-body stimuli and body-threat stimuli.

Threat SCRs
Pearson's two-tailed r correlations were conducted between the AHE factor from the CHi_II and Threat SCRs from the BTAB task in all three stimulation conditions (Anodal, Cathodal and Sham) (Table 2).A significant negative correlation was observed between AHE and Cathodal condition Threat SCRs suggesting that those scoring higher on this measure showed lower autonomic responses to the aversive body stimuli when subject to the Cathodal stimulation to the rVLPFC (Fig. 7).No significant findings were observed with in the Anodal and Sham conditions.The Bayes values here can be used to infer that the correlation coefficient for the cathodal condition is indeed reliably/significantly distinct from the other two conditions.
To further explore that the different brain-stimulation conditions were indeed inducing significant differences, the AHE measure was used as the independent variable to create a high predisposition group (high AHE, n = 13, M = 1.48,SD = 1.04) and low predisposition group (low AHE, n = 13, M = 0.12, SD = 0.16) by a median-split approach to identify the pattern of effects on brain stimulation (Fig. 8).

Table 2
Pearson's coefficients (FDR Corrected) and Bayes values between AHE and Threat SCRs (Z-Scored) under all three stimulation conditions (Anodal, Cathodal, Sham).Overall, the findings show that those scoring high on measures of cortical hyperexcitability displayed significantly more suppression of autonomic responses in the cathodal stimulation condition relative to those scoring low on this factor.It is particularly noteworthy there was some suggestion of mirror-reversed effects for the two groups (Fig. 8).

Magnitudes of NS-SCRs
As above, Pearson's two-tailed r correlations were conducted between the AHE factor and the strength of Threat NS-SCRs under all three stimulation conditions (Anodal, Cathodal and Sham).The frequentist analysis revealed there was a suggestion that the magnitudes of NS-SCRs in the Cathodal condition were correlating negatively with scores on the AHE measure (meaning as AHE scores increased, NS-SCR magnitudes decreased) however, the Bayes Factor analysis modifies this interpretation and suggests it is inconclusive (Table 3).
As the interaction was significant, further analysis in terms of within-subject ANOVA on the high AHE and low AHE groups was conducted.In the high AHE group, a significant main effect F (2,24) = 4.182, p = 0.028, η 2 ρ = 0.258, BF 10 = 6.205,BF 01 = 0.155) was revealed.Pairwise comparisons using FDR corrections in the high AHE group showed that under Cathodal stimulation NS-SCRs were

Frequency of NS-SCRs
Similarly, Pearson's two-tailed r correlations were conducted between the AHE factor and the frequency of Threat NS-SCRs in all three stimulation conditions (Anodal, Cathodal and Sham) though, no significant findings were observed (Table 4).
Independent sample t-tests between the two groups in the Frequentist analysis (FDR corrected) did not show any differences in any

BTAB and self-report ratings
The baseline block responses and threat block responses were compared across all four dimensions to examine the efficacy of the BTAB task in eliciting responses to non-body aversive and body-aversive imagery.

CHi-II and self-report ratings
In line with the psychophysiological data, the CHi-II AHE factor was subjected to Pearson two-tailed r correlations and a mediansplit (to quantify the sample into a high AHE group and low AHE group) analysis.The psychological ratings reported during the BTAB in each session were then compared based on the two groups and the findings are reported below for each dimension (that is, emotional arousal, emotional valence, sense of pain and realism of threat).
Emotional Arousal, Sense of Pain and Realism of Threat.The results from the correlational analysis and two-way ANOVA showed no significant effects (all p's > 0.05, BF 10 < 1, BF 01 > 3) between the two groups in the three stimulation conditions (see Fig. 12  a, c and d).Emotional Valence.As this data followed normality, a Pearson's two-tailed r correlation was conducted to compare responses on the AHE factor and the psychological ratings for valence however, no significant correlations were observed.

Depersonalisation-like experiences (DLEs)
The analysis utilising DLE responses largely produced null findings (all p's > 0.05, BF 10 < 1, BF 01 > 3) or in some cases inconclusive with BF 01 range between 1-3.Table 5 provides a summary overview of the main findings.In the case of "Sense of Pain" ratings in the threat condition, correlational analysis between DLE responses and this component showed a significant positive correlation (see Table 5) in the Anodal condition, suggesting that as the predisposition to DLE increased so did participants sense of illusory pain to viewing the Threat videos in this condition.Further analysis using mixed 2 × 3 ANOVA's and independent t-tests showed that that sense of pain ratings were not affected by stimulation condition (i.e., Anodal, Cathodal and Sham) but instead by predisposition to DLE (High DLE vs. Low DLE).In the interests of clarity, conciseness and transparency, the detailed analyses of these findings are reported in Appendix A.

General discussion
The current study examined the presence of trait-based predisposition to aberrant experience in mediating cognitive-affective responses as a result of optimised MtDCS directed at the rVLPFC / insula complex.Psychophysiological (skin conductance responses) and psychological responses (ratings) were obtained from participants whilst they viewed a novel aversive body-threat task (the BTAB: Braithwaite et al., 2020) under different MtDCS brain stimulation conditions.
The magnitudes of the maximum SCRs were significantly higher for threat stimuli relative to baseline stimuli.This was also the case for the magnitudes of the non-specific SCRs (NS-SCRs) -where the strength of NS-SCRs was reliably increased for threat stimuli relative to non-threat baseline stimuli.In addition, the frequency of NS-SCRs (a measure of tonic autonomic arousal) was significantly increased for threat stimuli relative to baseline stimuli for all but the cathodal condition, which appeared to show some evidence of a suppression in the frequency of NS-SCRs generated (although this was not conclusive).
At the whole sample level, there were no reliable effects of MtDCS brain stimulation on the strength of maximum threat SCRs relative to the sham condition.There were also no reliable effects of MtDCS brain stimulation on the frequency and magnitudes of NS-SCRs.This implies that MtDCS had no effects on the current sample.However, what is striking and noteworthy is that there were clear and significant effects of MtDCS when the sample was examined in relation to scores on a proxy measure of cortical hyperexcitability.This opposes the view that MtDCS had no discernible effects on neural processes, lending support to the notion that using such stratifying measures can speak to the trait-based baseline effects which can interact with and mediate the effects of brain stimulation (Basten et al., 2011;Krause & Cohen Kadosh, 2014;Horvath et al., 2014;Bell et al., 2022).
Cathodal stimulation of the rVLPFC via an optimised 7-channel MtDCS montage produced significantly suppressed SCR magnitudes, relative to the Sham and Anodal condition.As scores on the cortical hyperexcitability measure increased, Threat SCR magnitudes significantly decreased under the Cathodal condition.The Bayes Factor analysis, complementing the correlational analysis and the median-split analysis, support that that this effect was only observed for those scoring high on the CHi_II proxy measure of cortical hyperexcitability.This was not the case for those scoring low on the CHi_II measurewhere there were no reliable differences relative to the sham condition.Clearly the inferred degree of trait-based cortical hyperexcitability further mediated the effects of MtDCS and produced a highly selective effect -one that would be missed if averaging only at the whole sample level.
Anodal stimulation produced no reliable effects on SCR responses (maximum Threat SCRs or NS-SCRs) in terms of the strength or the frequency of responses relative to the sham condition.Furthermore, although there appeared to be a difference in the strength of threat SCRs for the sham condition for the high and low AHE groupsthis was not reliable and thus both groups displayed similar SCR profiles under no-stimulation (sham) conditions.
Collectively, the psychophysiological findings demonstrated that the MtDCS procedure was successful in influencing autonomic processing and responding but only for those participants showing elevated and increasing signs of cortical hyperexcitability and only for the cathodal condition.The suggestion here is that this is likely due to the stimulatory influence directed at the rVLPFC which in turn impacted on the operation of the anterior insula regiona network with a known functional involvement in saliency networks, the generation of conscious feeling states, interoceptive awareness, predictive-coding, and the mediation of autonomic skin conductance responses (Critchley, 2005;Medford et al., 2006;Lemche et al., 2007Lemche et al., , 2008;;Jay et al., 2014;Xia et al., 2017;Vinberg et al., 2021).
There is the possibility that the lack of influence from anodal stimulation (for both groups) may reflect something of a 'ceiling effect' in autonomic responding, where no additional response could occur from the application of brain stimulation.This is perhaps supported by the fact under anodal conditions, Threat SCRs were not reliably stronger than sham, and both were matched for strength of response.This could be due, at least in part, to the potency of the videos to elicit an autonomic response.
Whilst some novel effects in relation to proxy signs of cortical hyperexcitability were observed, this was not the case when the sample was examined in terms of DLEsone component of which was focused on aberrant body experiences.The DLE measure did not reveal any significant findings for either the magnitudes of maximum SCRs or NS-SCRs.However, it is interesting that the frequency of NS-SCRs did imply signs (p = 0.044, BF 10 = 1.650) of a reduction (increased inhibition) in autonomic responding as a function of predisposition to DLEs under cathodal stimulation.Whilst this did not survive FDR corrections, the corresponding Bayes value shows this effect to be inconclusive (and therefore not significant), the trend does go in the opposite direction to that expected from the idea of 'liberating' the AIC through inhibition of the rVLPFC.
Although interpreting null effects should be approached cautiously, the Bayes Factor analysis does allow for some extrapolation here and a tentative explanation can be explored.One reason for the non-significant DLE effects might be due to the questions on the DLE measure being more multifacetedcovering more mid and higher-level multisensory experiences relative to the very specific and focused visual items on the CHi_II measure (which centre on aberrant visual experiences only).In addition, the random-sampling approach led to the high DLE group with a mean score of 3.32 out of a maximum of 13 and so the present sample might not be sufficiently predisposed to aberrant body experiences in order for effects to strongly emerge (Sierra et al., 2005).Irrespective of these possibilities, the general pattern observed is consistent with the findings from the CHi_II and both contrast with the notion of inhibiting an inhibitory region results in increased autonomic responding (at least for neurotypical groups).
On the whole, the psychological ratings followed the pattern seen for the psychophysiological SCR data relating to the BTAB task.Threat stimuli were rated as more arousing, having more realism and being viewed more negatively (valence) relative to non-threat baseline stimuli.However, the results indicated that brain stimulation did not have a reliable impact on the overall psychological ratings.This was also observed even when the sample was split on the basis of the screening measures (AHE and DLE).There was a general trend for the high groups (on both measures) to rate the threat videos as more negatively valanced, emotionally arousing, generating a higher sense of illusory pain, and as more threatening, but the stimulation conditions did not reliably further influence these psychological ratings relative to sham.
A possible reason for this could be that psychological findings do not necessarily always follow SCRs and can at times reflect a dissociation between autonomic processing and reported conscious experience or subjective evaluation (Silvert et al., 2004;Christopoulos et al., 2019).In particular, autonomic/somatic responses to aversive stimuli have been shown to be the initial and immediate perceptual stage (as a survival strategy) following which interpretation, responses and learning are processed ( Öhman & Soares, 1993;Silvert et al., 2004;Christopoulos et al., 2019).This implies that the latter stages may be influenced (by various confounding factors, e. g., prior experiences, phobias etc.) and change conscious responses, i.e., degree of threat experienced may be differentially interpreted.

Theoretical implications
The current study investigated the role of the rVLPFC as an inhibitory network propagating into the AIC which has been implicated in the mediation of a variety of neurobiological processes such as negative cognitive-affective states, autonomic activity, saliency networks and empathy (Craig, 2002(Craig, , 2003(Craig, , 2009;;Critchley et al., 2004;Critchley, 2005;Ochsner & Gross, 2004, 2005;Lemche et al., 2007;Eippert et al., 2007;Klumpers et al., 2010;Gu et al., 2013;Seth, 2013;Clark, 2013).In addition, the current work examined the role of trait-based predispositions for cortical hyperexcitability and its potential to mediate the efficacy of MtDCS brain stimulation.
Findings from proxy signs of cortical hyperexcitability are particularly interesting.The present findings show, for the first time, that, as aberrant experiences thought to reflect underlying cortical hyperexcitability increase so does the efficacy of MtDCSbut only for the cathodal condition resulting in significantly suppressed autonomic SCRs.In the opposite to a hypothesised result, far from 'releasing' the AIC from aberrant suppression, the protocol adopted here increased suppression and this suppression increased in sympathy with proxy indicators of cortical hyperexcitability.
One possible solution might be that this reflects electrical stimulation primarily interacting with the most active neurons presentand thus effects are manifest here for the high scoring group on proxy measures of hyperexcitability as arguably background neural processes may well be operating at an elevated level of activation (Nitsche & Paulus, 2000;Liebetanz et al., 2002, Nitsche et al., 2005;Li et al., 2015).By this account, it is possible that cathodal stimulation increased activation in networks dedicated to the inhibitory regulation of the AIC, leading to a significant suppression of autonomic responses.
Consequently, rather than inhibiting the action of rVLPFC and increasing the frequency and/or strength of SCRs, cathodal stimulation simply increased the degree of inhibition mediating SCRs.Conceptually this is akin to a kind of functional steering or targeting of specific brain networks (Bikson & Rahman, 2013;Bestmann et al., 2015;Jackson et al., 2016;Kronberg et al., 2017;Vergallito et al., 2022).
Put more simply, the apparent friction with regards to how do proxy signs of increased hyperexcitability (the CHi_II responses) relate to signs of increased inhibition from brain stimulation, can be seen as a three-stage process where; at Stage 1 -there is a latent degree of background trait-based hyperexcitability occurring within the AIC (most likely via projections from hyperexcitable sensory cortices).At Stage 2 -this elicits a stronger inhibitory response from the rVLPFCresulting in these inhibitory systems working harder to regulate neurophysiological activity in the AIC.Therefore, in Stage 3 -MtDCS is applied to the rVLPFC and it interacts more strongly with the inhibitory processes as those networks are already working harder (e.g., functionally targeted).
Both the rVLPFC and AIC receive projections from multisensory areas, including vision (Crick & Koch, 1995;Miller, 1999;Barcelo et al., 2000;Benchenane et al., 2011).The rVLPFC has been shown to have an increased inhibitory role for negative emotions / feeling states and behavioural responses (Aron et al., 2004;Phan et al., 2005;Banks et al., 2007;Wager et al., 2008;Berkman & Liebermann, 2009;Szczepanski & Knight, 2014;Vergallito et al., 2018;Chick et al., 2020;Gallucci et al., 2020) and the insula complex is also centrally involved in attentional, somatosensory and autonomic processing (Critchley, 2004;Pollatos et al., 2007;Craig, 2009;Uddin, 2015;Wang et al., 2019;Chen et al., 2021).Consequently, during the processing of potent and aversive body-threatening imagery, activity within the insula would be increased (even more so with the presence of incoming hyperexcitability from visual systems) and associated by an increased involvement of suppressive processes from the rVLPFC.Accordingly, the increased involvement of inhibitory processes from rVLPFC make it more malleable to cathodal brain stimulation and such a mechanism could underlie the observation here of greater autonomic suppression from Cathodal MtDCS over the rVLPFC.
These findings also provide evidence that trait-based indicators of baseline levels of neural activity (via a proxy measure of aberrant experiences which reflect cortical hyperexcitability) can be used to reveal important differences in the efficacy of MtDCS to influence brain processes.This is consistent with a broader literature showing that baseline excitability can have different effects when brain stimulation is applied (Peña-Gómez et al., 2011;Jacobson et al., 2012;Sarkar et al., 2014;Benwell et al., 2015;Romei et al., 2016;De Graaf et al., 2017;Bell et al., 2022).
Based on the current findings, trait-based baselines are important to consider when examining brain stimulation methods.Specifically, stratifying individuals on the basis of trait-based predisposition to cortical hyperexcitability or even the inclusion of a statebased task could provide more thorough insights into the functional interaction between brain-stimulation montages and the background latent activity of neural systemsthat may otherwise be missed by general averaging approaches.Along with previous work on exploring EEG profiles (Fong et al., 2019(Fong et al., , 2020)), the present findings also considerably extend the utility of the CHi_II measure itself in relation to examining the efficacy of MtDCS brain-stimulation.
A prominent theory for the profound phenomenological experiences reported by patients with depersonalisation posits that the rVLPFC over-inhibits the AIC and thus prevents cognitive-affective processes from colouring consciousness (Hunter et al., 2003;Jay et al., 2014Jay et al., , 2016;;Critchley, 2005;Craig, 2009;Seth, 2013;Clark, 2013).Jay et al. (2014) provided evidence that by suppressing the rVLPFC (using rTMS), autonomic SCR activity was significantly increased in clinically depersonalised patients.In essence, the experimental protocol was being used to inhibit a brain network that was responsible for aberrant degrees of inhibition and in so doing, liberating the AIC (which receives prominent projections from rVLPFC) from such suppression.As far as trait based DLEs in the current neurotypical sample go, the current findings showed that predisposition to DLEs did not further mediate SCRs in relation to brain stimulation, at least for our neurotypical sample.
It should be acknowledged that the present study differs from that of Jay and colleagues (2014) in several important ways.First, the current study utilised a neurotypical sample that was measured for predisposition to DLEs and not patients with clinically diagnosed levels of these experiences.Second, the sample was also measured for proxy signs of cortical hyperexcitability to examine the efficacy of brain stimulation as a function of differences in background trait-based indicators.Previous research has not addressed this.Third, the current study utilised a form of transcranial electrical stimulation (optimised MtDCS) and not low frequency rTMS.Crucially, cathodal MtDCS applied to the rVLPFC did not 'release' the AIC and produce increased autonomic responding (SCRs) -quite the opposite.
From this, a highly speculative suggestion could be that the present study and that of Jay et al. (2014) might imply an important redescription in the functional relationships of these networks between neurotypical groups and those who have transitioned to disorder.Such possibilities require further investigation.

Limitations & Further study
While the use of trait-based measures is a noteworthy advancement in the field, future studies would benefit from an examination of additional state-based measures (i.e., the level of excitability at the time of testing).This is also interesting as some studies have revealed a complex interplay between these factorssuggesting that both are mediated by contributions from interdependent yet also distinct neural systems (see Kühn & Gallinat, 2012;Zmigrod et al., 2016; see also Smith et al., 2013; for evidence from an examination of P50 potentials).Trait-and state-based processes are not one and the same and one does not necessarily always predict the otherthough exactly how both interact with brain stimulation protocols remains an area of further study.
Work in this area with aberrant experiences in neurotypicals could also be improved by having a larger range of scores on the DLE measures used here.The absence of effects in this regard may reflect the observation that the present sample was not overly predisposed to DLEs and therefore such effects could not manifest.The present study adopted random sampling and so could not control for this.Future work might want to specifically target and recruit individuals that are highly predisposed to DLEs to examine this matter further and determine how the presence of such experiences may interact with MtDCS (or similar) protocols.
Finally, it should be acknowledged that the a-priori sample size estimations used here were speculative.This was partly due to the fact that a power analysis was deemed inappropriate due to the fact that many of the conditions and methods used in the present study are novel and have not been combined in this way before.While this makes our approach exploratory in this regard, established alternatives do exist and these approaches are based on what the literature has shown previously.In addition, our use of Bayes Factor analyses provides clear evidence of the central effects being either 'decisive' or 'strong' suggesting our sample is more than sufficiently powered to support the main claims.In addition, the computer-modelled optimised multi-channel approach used in the current work results in much more intense stimulation being steered more directly to the specific intended brain regions and thus is more focused and reliable than previous research that has used bipolar approaches with tDCS.Irrespective of these observations, future work which extends the sample to target and include high-scoring DLE participants as well as the sample size per-se could provide be helpful in providing additional insights.

Conclusion
The current study provides evidence that the rVLPFC (a region with strong functional and anatomical connections to the AIC) shows S.D. Joshi et al. differential involvement in mediating cognitive-affective responses to aversive body-threat stimuli in neurotypical individuals predisposed to aberrant experiences that reflect cortical hyperexcitability.This study has also shown that baseline brain states are important and trait-based stratification of the sample shows differences in the neuromodulatory paradigm that would be otherwise missed.At the whole sample level, no differences in autonomic responses to salient aversive stimuli were observed between Anodal, Cathodal and Sham conditions.However, examination of different stimulation conditions when the sample was divided based on predisposition to aberrant experiences showed suppression of autonomic responses for the high-scoring group only.This raises interesting and exciting prospects for the field, such as, whether benchmark findings may be weakened or modified and if previous studies with null findings are a result of group averaging.The effects from the current study demonstrated that cortical hyperexcitability can be extended beyond the visual and extra striate cortices and plays a role in the inhibitory functioning of the rVLPFC.Whether this is a result of visual projections into the region (already hyperexcitable), or a more domain-general degree of hyperexcitability that exists outside of visual networks but present in the brain region remains to be seen.Taken together, this provides additional evidence to the underlying neural mechanisms mediating stable consciousness and that cortical hyperexcitability may be a critical variable involved in these experiences.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.(FfWG) during the period in which this study was conducted and as such would like to thank the foundation for their generous endowment

Depersonalisation-like experiences and optimised MtDCS brain stimulation
Responses from the ABS and AFS factors of the depersonalisation scale were combined and normalised (responses divided by number of items) to make a composite independent variable labelled "DLE".

Threat SCRs
Pearson's two-tailed r correlations were conducted between the responses on the DLE factor and magnitude of Threat SCRs under all three stimulation conditions (Anodal, Cathodal and Sham) but no significant correlations were observed (Table A.1).  (Jarosz & Wiley, 2014).AH = Alternative Hypothesis.
The DLE factor was then split using a median to identify a high DLE group (n = 13, M = 3.32, SD = 0.82) and a low DLE group (n = 13, M = 0.85, SD = 0.56) to observe the pattern of effects on brain stimulation (Fig.

Magnitudes of NS-SCRs
As above, Pearson's two-tailed r correlations were used to compare the responses on the DLE measure and strength of Threat NS-SCRs under all three stimulation conditions (Anodal, Cathodal and Sham) however, in line with above, no significant findings were noted (Table A.2).

Frequency of NS-SCRs
Similarly, Pearson's two-tailed r correlations were used to compare the responses on the DLE measure and frequency of Threat NS-SCRs under all three stimulation conditions (Anodal, Cathodal and Sham) but no significant correlations were noted (Table A.3).

Depersonalisation-like experiences and self-report ratings
Similar to the above analysis, Pearson's two-tailed r correlations (for normal data) and Kendall's Tau (τ) correlations (for nonnormal data) followed by two-way ANOVA (for normal data and Friedman's tests for non-normal data) were conducted to compare between the DLE factor and psychological ratings (on all dimensions).

Emotional arousal, emotional valence and realism of threat
This data followed normality however we failed to note any significant effects for the Pearson's two-tailed r correlations or between the groups in all three conditions (all p's > 0.05, BF 10 < 1, BF 01 > 3) (

Sense of pain
As this data did not follow normality, non-parametric analysis was conducted.Kendall's tau correlations between the DLE factor and psychological ratings for sense of pain in all three stimulation conditions (Table A .4).This indicated that higher the scores on the DLE factor in the Anodal stimulation condition higher the sense of pain ratings.No significant correlations were observed in the Cathodal and Sham conditions.

Table A.4
Kendall's τ correlation coefficients (FDR Corrected) and Bayes values between DLE and psychological ratings (sense of pain) for all stimulation conditions (Anodal, Cathodal, Sham).* Significant correlations after using FDR corrections, the correlation pairs are ranked in ascending order.** Bayes values and interpretation according to (Jarosz & Wiley, 2014).AH = Alternative Hypothesis.
No significant differences were noted between the stimulation conditions (Anodal, Cathodal, Sham) in the high DLE and low DLE groups (Fig. A.4 c).However, Mann-Whitney U independent sample t-tests revealed that participants in the high group perceived the stimuli as more painful in every stimulation condition; Anodal (U = 141, p = 0.003), Cathodal (U = 123.5,p = 0.044) and Sham (U = 123.5,p = 0.044).

Fig. 2 .
Fig. 2.Representation of the placement of electrodes over the rVLPFC.Note: AF8, F6 and F8 as stimulation channels and AF4, FC4, FP1 and T8 as return channels and magnitude of the electric field.
S.D.Joshi et al.

Fig. 9 .
Fig. 9. Magnitudes of Threat NS-SCR's (Z-Scored) under all brain stimulation conditions (Anodal, Cathodal and Sham) for the high AHE group and low AHE group.

Fig. 10 .
Fig. 10.Frequency of Threat NS-SCRs (CPM) under all brain stimulation conditions (Anodal, Cathodal and Sham) for the high AHE group and low AHE group.

Fig. 11 .
Fig. 11.Differences between Baseline and Threat blocks in each stimulation condition (Anodal, Cathodal and Sham) for the four psychological rating dimensions.Note: In the above the four psychological ratings are illustrated for (a) emotional arousal, (b) emotional valence, (c) sense of illusory pain and (d) realism of threat.

Fig. 12 .
Fig. 12. Differences between high AHE vs. low AHE groups for all stimulation conditions (Anodal, Cathodal, Sham) in the four psychological rating dimensions.Note: In the above figure, the four psychological ratings are illustrated for (a) emotional arousal, (b) emotional valence, (c) sense of illusory pain and (d) realism of threat.
Fig. A.4 a, b and d).

Fig. A. 4 .
Fig. A.4. Differences between high DLE vs. low DLE groups for all stimulation conditions (Anodal, Cathodal, Sham) in the four psychological rating dimensions.Note: In the above figure, the four psychological ratings are illustrated for (a) emotional arousal, (b) emotional valence, (c) sense of illusory pain and (d) realism of threat.

Table 5
Summary of the main findings from the DLE responses.