Abstract

Introduction:

Affective disorders (ADs) are characterized by profound emotional processing deficits involving disrupted neural network activity and connectivity, particularly within the default mode network and fronto-temporal circuits, with abnormalities in theta and alpha oscillatory patterns. While current treatments primarily target mood symptoms, emotional processing impairments often persist and predict relapses. Awe, a complex self-transcendent emotion, may counteract such deficits through its capacity to reduce rumination and enhance positive affect. However, the neural correlates of awe experiences in clinical populations remain unexplored.

Objective:

For the first time, this exploratory study investigated the electroencephalographic (EEG) correlates of awe induced by validated virtual reality (VR) scenarios in individuals with ADs compared to healthy controls (HCs).

Methods:

Participants were exposed to immersive VR scenarios designed to elicit different awe experiences (mountains, waterfall, Earth) and a reference (awe-neutral) scenario. EEG activity was recorded during VR exposure and at baseline, followed by emotional state questionnaires. Power spectral density and graph-theoretical connectivity indices – Nodal Positive Strength and Global Efficiency – were computed across theta, alpha, and beta bands.

Results:

Healthy controls showed high awe responses in awe-inducing scenarios with selective, scenario-specific modulations in alpha and theta band activity and connectivity, reflecting preserved cognitive flexibility. Conversely, ADs reported similar awe responses across all VR scenarios with reduced environmental differentiation. With respect to HCs, ADs showed elevated theta power in bilateral frontal and temporal regions, suggesting compensatory activity related to emotional processing alterations. Both groups exhibited VR-induced reductions in alpha-band global efficiency, more pronounced in ADs, suggesting compromised neural integration during complex emotional processing.

Discussion:

Taken together, the results suggest that the emotional processing deficits inherent to ADs may limit the capacity to engage differentially with emotionally complex stimuli such as awe, while nonetheless providing initial evidence that VR-based awe exposure combined with neurophysiological recording represents a valuable approach for discriminating differential cerebral emotional responses in clinical populations. This proof-of-concept work warrants further investigation in larger cohorts to evaluate the therapeutic potential of awe-based interventions for affective disorders.

1 Introduction

Affective disorders (ADs), also known as mood disorders, represent a group of psychological conditions characterized by alterations in emotional regulation, frequently presenting as depressive manifestations (Sekhon and Gupta, 2023). Among depressive conditions, Major Depressive Disorder (MDD) and Bipolar Disorder (BD) are leading causes of disability worldwide, significantly impacting quality of life and increasing the risk of medical comorbidities (Grande et al., 2016; Lam et al., 2024; Yan et al., 2024). Beyond their core diagnostic criteria, ADs are characterized by profound alterations in the cognitive-affective domain, while MDD is characterized by negative bias in implicit emotional processing predominates, with enhanced reactivity to negative emotional cues and reduced accessibility to positive emotional stimuli (Zhang et al., 2016), BD presents a more complex pattern of emotion dysregulation characterized by extreme affective reactivity, difficulties in emotion regulation across all mood states, and impairments in facial emotion recognition and reward processing (Townsend and Altshuler, 2012; De Prisco et al., 2023).

The neurobiological underpinnings of these emotional processing deficits involve functional alterations within multiple brain networks. Electroencephalographic (EEG) studies in ADs consistently report abnormalities in oscillatory activity associated with emotional and motivational processing. A well-replicated finding is frontal alpha-band asymmetry, which has been interpreted as a marker of impaired positive emotion processing and a bias toward negative affective states (Harmon-Jones et al., 2008; Fernández-Palleiro et al., 2020). In addition to alpha-band abnormalities, converging evidence highlights disruptions in theta-band activity across ADs. Increased frontal and anterior midline theta power, commonly associated with dysfunctional anterior cingulate cortex activity, has been robustly linked to impaired emotion regulation and heightened ruminative or internally focused processing on depression (Jaworska et al., 2012; Arns et al., 2015). Furthermore, Functional Magnetic Resonance Imaging (fMRI) studies on MDD show reduced deactivation in the pregenual anterior cingulate cortex, ventromedial prefrontal cortex, and posterior cingulate cortex, alongside altered activity in the dorsolateral prefrontal cortex when compared to control subjects (Grimm et al., 2009; Victor et al., 2013). Moreover, reduced functional connectivity has been observed between the medial temporal gyrus (MTG) and the ventral caudate nucleus, as well as other regions, alongside increased connectivity between the superior temporal gyrus and precuneus, and angular gyrus (Yang et al., 2017). The persistent hyperactivity of the Default Mode Network (DMN) during transitions from resting state to externally oriented tasks contribute to cognitive vulnerability by promoting rumination and depleting resources needed for adaptive emotional processing (Wang et al., 2016). Similarly, BD shows disrupted DMN connectivity, particularly in the precuneus, medial prefrontal cortex, and posterior cingulate cortex (Öngür et al., 2010; Townsend and Altshuler, 2012; Zhao et al., 2024). Importantly, BD is also associated with decreased anticorrelation between the DMN and the salience network, potentially reflecting difficulties in switching between internally and externally directed attention (Zhao et al., 2024).

Treatment options for AD primarily target mood symptoms, with current first-line treatments being a result of a combined use of pharmacotherapy with psychotherapy to achieve symptomatic relief (Sekhon and Gupta, 2023). In MDD, typical approaches include antidepressants and cognitive behavioral therapy, while BD management requires mood stabilizers and atypical antipsychotics, often combined with psychoeducation and family-focused therapy (Fountoulakis et al., 2005; Shah et al., 2017). However, emotional processing impairments persist beyond mood symptom resolution and predict relapses (Roiser et al., 2012; Ruhe et al., 2019). Beyond standard treatments, experiential interventions, particularly those targeting emotional processing deficits, may be necessary to fully address the emotional processing deficits of ADs (Pos et al., 2003; Yen et al., 2024). Recent research began to explore the therapeutic potential of complex emotions induced in controlled laboratory settings. Of particular interest is the emotional experience of awe, studied as a possible adjunctive pathway to enhance positive affect, shift the focus away from the Self, and broaden attentional scope, which could benefit patients with depression (Chirico and Gaggioli, 2021; Monroy and Keltner, 2022).

Awe is a complex emotion arising from vast, overwhelming, and/or threatening stimuli (Keltner and Haidt, 2003) that exceeds one’s current mental schemas and necessitates cognitive accommodation (Pérez et al., 2023). This emotion has been defined as a self-transcendent phenomenon (e.g., Yaden et al., 2017) featuring two core features, that is, an increased sense of connectedness to others and the environment, and a diminished sense of self, termed “the small self” (Piff et al., 2015; Yaden et al., 2017; Cavallaro and Rivera, 2025). At the same time, awe itself can trigger profound self-transcendent experiences characterized by alterations in the perception of space, time, and self-concept, placing it at the edge between an emotional state and an altered state of consciousness (Chirico and Yaden, 2018; Abatista and Cova, 2023; Jiang et al., 2024; Zhao et al., 2026).

The self-transcendent properties of awe – shifting attentional focus from self to others and fostering a sense of connection to something greater – present compelling mechanisms for addressing core features of ADs. According to the Matryoshka model of self-transcendence (Chirico and Gaggioli, 2021), these properties directly counteract fundamental aspects of depressive psychopathology, including rumination and maladaptive self-referential thinking (Hendricks, 2018; Lopes et al., 2020; Chirico and Gaggioli, 2021; Monroy et al., 2025). Beyond its self-transcendent characteristics, awe enhances creative thinking (Chirico et al., 2018b; Zhang et al., 2021; Pizzolante et al., 2024) and promotes prosocial behavior (Piff et al., 2015; Li et al., 2024), cognitive and social domains notably impaired in ADs (Chirico and Yaden, 2018; Chirico et al., 2018b; Jiang and Sedikides, 2022). Furthermore, awe’s capacity to trigger profound restructuring of mental schemas and self-perception suggests potential for neuroplastic modifications of the inflexible cognitive biases that are characteristic of these disorders (Beck and Bredemeier, 2016; Price and Duman, 2020).

At the neural level, preliminary evidence from fMRI suggests that awe engages neural circuits relevant to the pathophysiology of ADs, supporting its potential therapeutic mechanisms. Neuroimaging research in healthy individuals has identified activation of the frontoparietal network, including medial frontal gyrus, insula, and supramarginal gyrus, as well as reduced DMN activity (van Elk et al., 2019). Additionally, the left MTG plays a pivotal role in processing both positive and threatening variants of awe (Takano and Nomura, 2022), while awe intensity appears modulated by the inferior temporal cortex, hippocampus, inferior and middle frontal gyri, basal ganglia, and cerebellum (Ishizu and Zeki, 2014).

However, while fMRI provides valuable insights into the brain hemodynamic correlates of awe with high spatial resolution, it presents inherent limitations for investigating neuronal dynamics, including oscillatory and connectivity patterns that unfold on millisecond timescales. Additionally, the fMRI environment constrains experimental paradigms to predominantly passive audio and/or visual stimuli. In this context, EEG offers complementary advantages for studying fast neuronal dynamics and connectivity patterns (Coelli et al., 2017, 2019; Maggioni et al., 2021) underlying awe experiences. Particularly, EEG provides direct measurement of brain electrical activity with exceptional temporal resolution, capturing the rapid neural dynamics that characterize emotional processing through quantification of oscillatory activity via power spectral measures and characterization of dynamic interactions between brain regions via connectivity metrics across the main frequency bands. Beyond its temporal advantages, EEG’s compatibility with diverse experimental settings broadens the methodological repertoire for awe induction. Among diverse awe-inducing technologies, virtual reality (VR) has emerged as a promising approach to combine experimental control and high ecological validity in the laboratory settings (Gallagher et al., 2015; Chirico et al., 2017; Hofmann et al., 2021; van Limpt-Broers et al., 2024). Immersive VR environments enable multisensorial stimulation and the creation of realistic scenarios (Chirico et al., 2016), thereby affording a unique sense of presence and immersion that enhances the emotional intensity of awe experiences.

Despite these methodological advantages, relatively few studies have investigated the electrophysiological correlates of awe in healthy individuals, especially those using VR paradigms. In our previous EEG-VR study, Carbone et al. (2024) investigated the brain mechanisms that underlie VR-induced awe experiences using different frequency-resolved metrics, among which spectral power and entropy-based measures. Within this work, three awe-inducing scenarios were compared with respect to a reference VR one. Across the awe-inducing scenarios, consistent metrics increases were found in selective channels spanning from the frontal and temporal lobes (theta, alpha, beta, gamma bands) to the parietal (beta band) and occipital (gamma band) lobes. Despite highly overlapping results, entropy-based metrics emerged to be particularly sensitive in capturing subtler scenario-specific changes, underscoring the importance of employing complementary EEG metrics to fully characterize brain processes associated with awe. Similarly, van Limpt-Broers et al. (2024) identified modulations in theta and beta band power during an awe-inducing VR experience simulating a rocket launch. Finally, Lee et al. (2025) characterized awe as an ambivalent affective state encompassing both positive and negative emotional components. These components were found to be associated with changes in frontoparietal multi-band oscillatory power.

The integration of EEG with immersive VR presents an unprecedented opportunity to investigate the neural dynamics of awe not only in healthy individuals, but also in clinical populations – such as those with ADs – who could benefit from awe-based experiential interventions. The ecological validity of VR-induced awe combined with the temporal precision of EEG can be used to explore whether experimentally induced awe experiences can modulate dysfunctional neural patterns that characterize emotional processing deficits in ADs. Such investigations are essential for determining whether awe-based interventions could serve as novel adjunctive treatments for the full spectrum of ADs.

In this study, for the first time to our knowledge, we aim to investigate the EEG activity and connectivity patterns underlying different awe experiences induced by validated VR scenarios (Chirico et al., 2018a) in a small sample of individuals with ADs with respect to a group of healthy controls (HCs).

2 Materials and methods2.1 Participants

The sample consisted of 3 outpatients volunteers with ADs – 1 with BD (M, 24 years old) and 2 with MDD (F, 38 years old; F, 20 years old) – and 12 HCs (5 males, 7 females, 27.6 ± 5.0 years). All participants were between 20 and 40 years, had a normal or corrected vision, were not pregnant, and were capable of giving informed consent and complying with the study procedures. Individuals with ADs were diagnosed with affective disorders using the Italian version of the Structured Clinical Interview for DSM-5. Main exclusion criteria for HCs were meeting the DSM-5 criteria for psychotic, mood, or substance use disorders, either currently or in the past, balance or vestibular disorders, pregnancy, severe cognitive deficits, and intellectual disability. A more detailed list of exclusion criteria can be found in the registered protocol within the Open Science Framework (https://osf.io/m47du) and the relative study protocol paper (Bondi et al., 2025). Volunteers with ADs were enrolled from the Mood Disorders Day Hospital service in the Psychiatric Clinic of the Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, while controls were enrolled from the general Italian population. A written informed consent was obtained by all participants prior to the study. This study is compliant with the Helsinki Declaration and has been approved by the competent Ethical Committee on January 26th, 2021 (OSMAMI-26/01/2021-0002688-U).

2.2 Experimental design

The protocol included a resting-state EEG recording (baseline) and a subsequent VR session within which participants were asked to navigate in four different VR scenarios, three awe-inducing and one awe-neutral (also called reference), which were administered in a counterbalanced order. EEG data were recorded during navigation in each of the four VR scenarios.

The three awe-inducing scenarios, previously validated for awe elicitation by Chirico et al. (2018a), featured natural scenes of (1) a forest with tall trees (waterfall), (2) high mountains (mountains), and (3) the earth view from deep space (Earth). The reference scenario represents green grass with few flowers and trees and is used as control (awe-neutral) condition. The awe-inducing scenarios have been designed to feature a moment in which awe should reach its maximum intensity, hereinafter called peak of awe, which occurs when something unexpected happens after the exploration of a standardized navigation path (Bondi et al., 2025). During the EEG data acquisition of awe-inducing VR scenarios, the peak of awe was marked on the EEG as an event.

Beside neurophysiological measures, at baseline and after each VR scenario, the emotional status of participants was assessed using the Short Positive and Negative Affect Scale (PANAS) (Terracciano et al., 2006), and the Single Item Emotion (SIE) scale (Chirico et al., 2018a). Further details on the experimental protocol can be found in Bondi et al. (2025).

2.3 EEG data acquisition and pre-processing

At baseline and during each VR session, EEG data were continuously recorded using a Geodesic EEG system (GES 400, Electrical Geodesic Inc. (EGI), Philips, The Netherlands) equipped with a TMS-compatible 64-channel Geodesic EEG cap (MicroCel Geodesic Sensor Net 100, EGI, Philips, The Netherlands). The EEG sampling frequency was set to 1 kHz, and the reference electrode was placed at the vertex of the head.

The EEG data pre-processing was conducted using Matlab® R2022a (Mathworks, Inc.), the EEGLAB toolbox (Delorme and Makeig, 2004), Matlab in-house scripts, and BrainVision Analyzer 2.0 R © software (Brain Products, Gilching, Germany). Sixty-second-long artifact-free segments were extracted for the baseline and reference VR scenarios, whereas a window of [−30:30] seconds around the peak of awe was used for the awe-inducing ones. The data were filtered using a finite impulse response band-pass filter (1–120 Hz) and a notch filter (50 Hz) to remove powerline noise and were down-sampled to 250 Hz. The BrainVision Analyzer software was then used to perform Independent Component Analysis (ICA) to detect and remove artifacts related to eye blinks, muscle movements, and single electrodes. The EEG signal was then reconstructed using the artifact-free ICs and re-referenced to the channels’ common average. The clean EEG data were divided into 2-s epochs, and for the awe-inducing VR scenarios, only the epochs after the awe peak were retained. Power spectral density (PSD) and functional connectivity (FC) were then computed across the theta, alpha, and beta bands for each epoch.

2.4 Emotional questionnaires

The emotional changes induced by the VR protocol relative to the baseline were evaluated computing the percentage-normalized scores for the PANAS and SIE questionnaires items for each VR scenario using the following formula:

Where VR is the scenario (i.e., reference, Earth, waterfall, and mountains), is the percentage-normalized item score of the VR scenario, is the item score of the VR scenario, and refers to the item score of the questionnaire administered at baseline.

2.5 PSD and FC features

The power analysis was performed on individual EEG epochs using the Fast Fourier Transform method. Spectral power was quantified across three frequency bands: theta (4–8 Hz), alpha (8–13 Hz), and beta (13–30 Hz). For each band, power was computed through trapezoidal integration across the corresponding frequency range. Relative PSD (rPSD) values were obtained by dividing the power in each band by the total integrated power spanning 1 Hz to the Nyquist frequency. These rPSD values were then averaged across epochs for each participant, experimental condition, and frequency band. To quantify the changes induced by the VR compared to baseline levels, the percentage-normalized rPSD was derived via Equation 1.

The imaginary part of coherency (iCOH), which measures the brain functional connectivity by isolating true neural interactions by excluding zero-lag correlations arising from volume conduction artifacts (Nolte et al., 2004), was computed using the FieldTrip toolbox (Oostenveld et al., 2011). The EEG data underwent Fourier decomposition via multitaper methods with Hanning windows across the 1–120 Hz range. Similarly to the rPSD normalization approach, band-specific connectivity was determined through trapezoidal integration and normalized by total iCOH, in a manner similar to the rPSD normalization approach.

Graph-theoretic metrics were extracted using the Brain Connectivity Toolbox (Rubinov et al., 2009) in MATLAB. After normalizing the FC adjacency matrices and setting the diagonal elements to zero, node positive strength (NPS) and global efficiency (GE) were computed for each experimental condition and frequency band. NPS, the sum of weights of links connected to the node, reflects nodal centrality within the network. GE, calculated as the mean inverse shortest path length between all node pairs, with higher values indicating more direct communication pathways between nodes, was used as a measure of network integration. As with spectral analyses, connectivity metrics were expressed as normalized-percentage changes from baseline (Equation 1).

2.6 Statistical analysis

The Wilcoxon rank-sum test was used to evaluate differences in the percentage-normalized rPSD and graph-theory EEG measures between the ADs and HCs groups in each VR scenario and frequency band. The Wilcoxon signed-rank test was used to evaluate differences between awe-inducing VR scenarios and the reference one in each group and frequency band. Significance thresholds were set at p < 0.05 (uncorrected for multiple comparisons due to the exploratory nature of the study). The same statistical tests were performed for the PANAS and SIE percentage-normalized scores. Given the exploratory nature of this study and the constraints of the EEG analytical approach, p-values are reported for descriptive purposes only and should not be interpreted as confirmatory evidence.

3 Results3.1 Emotional questionnaires: within- and between-group comparisons

The PANAS and SIE median scores for each group and condition are reported in Tables 1, 2. No differences (p < 0.05) were found for the PANAS scale, neither between groups nor conditions. Contrarily, differences within the HCs group and between ADs and HCs were found for the Amused, Happy, and Awe items of the SIE scale. HCs showed differences for the Amused item between reference and Earth (p < 0.05) and reference and waterfall (p < 0.01) scenarios; for the Happy item between reference and Earth (p < 0.05) and reference and waterfall (p < 0.05) scenarios; for the Awe item between reference and Earth (p < 0.01) and reference and mountains (p < 0.01) scenarios. Meanwhile, differences between groups were found for the Amused item in the waterfall scenario (p < 0.05); for the Happy item in the Earth (p < 0.05) and waterfall (p < 0.05) scenarios; for the Awe item in the reference (p < 0.05), waterfall (p < 0.05), and mountains (p < 0.05) scenarios.

Baseline HCBaseline ADReference HCReference ADEarth HCEarth ADWaterfall HCWaterfall ADMountain HCMountain ADEnthusiastic333433332,54Determined3,532,53323,5435Attentive4434313,543,55Excited322,54333223Inspired2,5223323224Positive Affect16,5131418141115,5141221Nervous22211,541122Afraid1,5112111111Upset2,511,5112111,52Distressed1112111111Scared1111111111Negative Affect86675,595577

Median PANAS scores for healthy controls (HC) and affective disorders (AD).

Baseline HCBaseline ADReference HCReference ADEarth HCEarth ADWaterfall HCWaterfall ADMountain HCMountain ADAngry1112161111Disgusted111,51131111Afraid1111121111Proud2313221,5124Amused45253,54423,56Sad1211111111Happy3,542,514,514134Awe3134533,544,55

Median SIE scores for healthy controls (HC) and affective disorders (AD).

3.2 rPSD between-group comparison: VR scenarios

As shown in Figures 1, 2, when comparing the ADs and HCs groups, differences (p < 0.05) in terms of rPSD were mostly found in theta and alpha bands. Specifically, in the theta band the mountains scenario showed changes in the bilateral frontal (p < 0.01) and temporal areas, with an increase (with respect to baseline) for ADs and a decrease for HCs. While both groups exhibited negative rPSD in the alpha band in all VR scenarios, with respect to baseline, ADs showed lower values in the occipital regions in both the reference and waterfall scenarios, and in the frontal and frontocentral regions in the reference one only.

The relative power spectral density (rPSD) for each group [healthy controls (HC) and affective disorder (AD)] in the theta, alpha, and beta bands is shown for the reference and Earth scenarios. The topoplot values refer to the median percentage-normalized – with respect to baseline – value of rPSD for the VR scenarios. Difference in median percentage-normalized rPSD between ADs and HCs for electrodes showing group differences (Wilcoxon rank-sum test, p < 0.05).

The relative power spectral density (rPSD) for each group [healthy controls (HC) and affective disorder (AD)] in the theta, alpha and beta bands is shown for the Waterfall and Mountains scenarios. The topoplot values refer to the median percentage-normalized – with respect to baseline – value of rPSD for the VR scenarios. Difference in median percentage-normalized rPSD between ADs and HCs for electrodes showing group differences (Wilcoxon rank-sum test, p < 0.05).

3.3 rPSD within-group comparison: awe-inducing vs. reference scenarios

When comparing awe-inducing VR scenarios (mountains, Earth, and waterfalls) rPSD with the reference one, differences were observed in all frequency bands, solely in the HCs group (Figure 3). In the theta band, a reduction (p < 0.05) of rPSD is observed predominantly in the left temporal area in the Earth environment and the right frontal area in the waterfall environment. In the alpha band, lower rPSD values were observed bilaterally in the frontal and temporal regions, with the latter being more widespread in the Earth scenario. Finally, the beta band showed lower rPSD values in bilateral temporal regions for the Earth scenario, bilateral frontal for the Waterfall, and central regions for the mountains. A higher rPSD value was observed in the right frontal region for the mountains scenario.