Neural Correlates of the Shamanic State of Consciousness

Psychedelic compounds have been used as model interventions to study altered states of consciousness, but pharmacologic induction introduces potential confounds because drugs act on multiple receptor sites. The shamanic state of consciousness—typically entered during repetitive drumming and practised both in indigenous and neo‑shamanic contexts—produces phenomenology that is anecdotally similar to psychedelic experiences (mystical experiences, disembodiment, ego dissolution), yet little empirical work has examined its neural basis in non‑pharmacologic settings. This study set out to characterise the shamanic state of consciousness using high‑density electroencephalography (EEG) and a validated psychometric measure of altered states (the OAV scale). The investigators compared experienced shamanic practitioners with matched control participants across baseline rest, a 25‑minute drumming period intended to elicit trance, and a classical music control, and applied multiple computational EEG metrics—absolute spectral power, functional connectivity (wPLI), neural signal diversity (Lempel‑Ziv complexity, LZc), and criticality (pair correlation function, PCF)—to identify brain changes associated with the shamanic state and their relationship to subjective reports, and to compare those reports with published data from psychedelic drug studies.

The study recruited two groups: 24 experienced shamanic practitioners and 24 age‑ and sex‑matched control participants. Inclusion criteria for practitioners required at least five years of shamanic practice, a minimum of 40 healing sessions in the past five years, training under expert supervision, habitual use of rhythmic drumming to enter trance, and the ability to enter trance while seated within 15 minutes. Controls were screened to exclude prior shamanic, trance, meditation, spiritual healing, or psychedelic use, as well as serious psychiatric or neurological history. After exclusions for EEG technical problems, data quality, or missing questionnaires, the final analysed sample comprised 18 practitioners and 19 controls. Before the lab visit, practitioners practised entering trance to a 25‑minute pre‑recorded drumming piece; controls listened to the same recording but were instructed not to attempt trance. In the lab, participants underwent baseline EEG (5 minutes eyes open, 5 minutes eyes closed) and three experimental blocks presented in counterbalanced order: cognitive testing (not reported here), shamanic drumming (25 minutes, eyes closed), and classical music listening (15 minutes, two instrumental movements chosen for some repetition and absence of lyrics). The OAV questionnaire (66 items) was completed after the drumming and after the classical music block; OAV item scores were pooled into 11 domains (e.g. experience of unity, complex imagery, elementary visual alterations) and expressed as percent of theoretical maximum. EEG was recorded with a 129‑channel HydroCel sensor net at 500–1,000 Hz, referenced to the vertex, then preprocessed in EEGLAB: resampled to 500 Hz as needed, bandpass filtered 1–45 Hz, re‑referenced to the global average, visually inspected for artefact and epoched into 10‑s non‑overlapping windows. Analyses focused on canonical frequency bands (delta 1–4 Hz; theta 4–8 Hz; low alpha 8–10 Hz; high alpha 10–13 Hz; low beta 13–20 Hz; high beta 20–30 Hz; gamma 30–45 Hz), with connectivity limited to 30–35 Hz for reliability. Four analytic streams were applied: (1) absolute spectral power via short‑time Fourier transform, (2) functional connectivity using weighted Phase Lag Index (wPLI) averaged across time and channels, (3) neural signal diversity computed with Lempel‑Ziv complexity (LZc) on binarised epochs and calculated per band and broadband (1–45 Hz), and (4) criticality assessed via the pair correlation function (PCF), a measure of variance in network synchronisation which is maximal at criticality. Statistical comparisons used two‑way mixed ANOVAs to test group (practitioner vs control) by condition (eyes closed, classical music, drumming) interactions for EEG measures, with exploratory post‑hoc unpaired t‑tests within conditions. Non‑parametric Wilcoxon tests compared OAV domain scores between paired conditions and between groups. Spearman correlations examined relationships between EEG metrics that differed between groups during drumming and OAV domains that differed between groups during drumming. The authors report unadjusted p‑values for exploratory post‑hoc tests and correlations, and compared OAV scores with aggregated published data from 43 psychedelic studies (psilocybin, ketamine, MDMA) using unpaired t‑tests; the aggregated nature of those data was noted as a limitation.

Subjective experience: Shamanic practitioners reported a clear altered state of consciousness during the drumming condition. Compared to their own scores during classical music, practitioners had significantly higher OAV domain scores in the majority of domains assessed (all except elementary visual alterations). When comparing practitioners’ OAV scores during drumming to pooled data from healthy volunteers under psilocybin, ketamine or MDMA from prior studies, practitioners scored at or above those drug conditions on several domains. Specifically, practitioners exceeded the drug‑condition means for complex imagery (psilocybin, ketamine, MDMA; p ≤ 0.0077 to p < 0.001), experience of unity (p ≤ 0.0052 to p < 0.001), spiritual experience (p < 0.001 across drugs), and insightfulness (p < 0.001 across drugs). Other domains showed similarity to one or more drugs (for example, blissful state similar to MDMA, disembodiment similar to ketamine), and practitioners had lower anxiety than the ketamine group (p = 0.037). The authors note that the psychedelic comparison used aggregated data of varying doses and methods. Spectral power: There was a significant group × condition interaction for gamma power (F(1.7,59.1) = 6.3, p = 0.005). Post‑hoc tests showed shamanic practitioners had greater absolute gamma (30–45 Hz) power than controls during drumming [t(29) = 2.16, p = 0.039, 95% CI 0.040–1.43]. Within practitioners, gamma power during drumming correlated with elementary visual alterations (Spearman r s = 0.52, p = 0.025). Functional connectivity: Practitioners differed from controls in low alpha connectivity (main effect of group: F(1,35) = 5.6, p = 0.024) and showed increased low beta connectivity during music (post‑hoc t(24.9) = 2.25, p = 0.033). The extracted text indicates practitioners had decreased low alpha and increased low beta connectivity across drumming and classical music; however, these connectivity differences did not correlate with OAV domain scores. Some connectivity differences also appeared at baseline, suggesting possible trait‑level effects. Neural signal diversity (LZc): There was a main effect of group on gamma‑band LZc (F(1,35) = 4.79, p = 0.035). Post‑hoc testing found decreased gamma LZc in practitioners during drumming (t(19.4) = -2.53, p = 0.02, 95% CI [-0.0274, -0.00262]). The authors further report a statistically significant association between LZc in the gamma band and feelings of insightfulness (p = 0.034). The extracted text describes this relationship as negative (an inverse correlation) but the reported Spearman value in the extraction appears inconsistent (r s = 0.5); the direction of the association is not unambiguously clear from the provided text. Criticality (PCF): PCF was increased in practitioners during drumming in the low beta, high beta, and gamma bands. Specifically, increases in low beta PCF correlated with complex imagery (r s = 0.53, p = 0.023) and elementary visual alterations (r s = 0.56, p = 0.015). Practitioners also showed greater PCF in the high beta band during classical music (t(35) = 2.1, p = 0.043) and in the gamma band during classical music (t(31.5) = 2.75, p = 0.00987) and eyes‑closed rest (t(31) = 2.16, p = 0.039). The authors interpret some PCF increases present during music and rest as possible long‑term (trait) shifts in criticality in practitioners. Summary of associations: EEG measures that were specific to drumming and correlated with subjective domains included increased gamma power and decreased gamma LZc (both associated with visual/insight domains) and increased low beta PCF (associated with complex imagery and elementary visual alterations). Connectivity changes did not show direct correlations with OAV scores.

Pasquini and colleagues interpret their results as showing that experienced shamanic practitioners reliably enter an altered state during rhythmic drumming, with subjective effects that overlap substantially with psychedelic experiences yet show distinct neurophysiological signatures. The principal EEG finding was increased gamma power during drumming in practitioners, which correlated with elementary visual alterations; this aligns with prior EEG work reporting elevated beta/gamma activity during shamanic trance and with literature linking gamma oscillations to visual imagery, working memory and some meditative states. In contrast to the psychedelic literature—where many studies report broadband decreases in power, especially in alpha, and increases in broadband signal diversity—the shamanic state here lacked alpha decreases, showed decreased gamma‑band LZc, and demonstrated increased criticality (PCF) in beta and gamma bands. The authors note that prior studies of psychedelics generally report increased LZc that correlates with subjective intensity, so the decrease in gamma LZc observed here and its association with insightfulness suggests a different mechanistic profile. They caution that methodological differences (narrow frequency‑band versus broadband LZc, different entropy measures) and the possibility that shamanic trance and meditation are distinct absorptive states may underlie these discrepancies. The increase in PCF is discussed in the context of the entropic brain hypothesis: elevated PCF suggests a shift toward the upper bounds of an extended critical zone, implying greater network susceptibility and metastability that could support richer conscious contents. Because some PCF increases appeared during non‑drumming conditions, the authors propose that shamanic practice may induce enduring trait‑like changes in criticality. They also compare the shamanic EEG profile to findings from religious and contemplative practices, noting partial overlaps (for example, gamma increases in some mystical recall studies) but overall heterogeneity across spiritual traditions. Key limitations acknowledged by the investigators include absence of electromyography to rule out muscle contamination—particularly relevant for gamma-band effects—relatively small sample size, heterogeneity of shamanic traditions among practitioners, lack of blinding and potential demand characteristics, and the control task difference whereby practitioners were actively engaged in trance while controls passively listened. The authors also caution about their comparison with aggregated psychedelic OAV data, which pooled heterogeneous doses and methodologies from 43 studies. They recommend larger, longitudinal, and ideally blinded studies, and suggest examining shamanic trance with and without pharmacologic agents to further clarify similarities and differences with psychedelic states.

Shamanic practitioners entered a pronounced altered state of consciousness during drumming, with self‑reported changes that matched or exceeded many domains reported in aggregated psychedelic studies. EEG changes accompanying shamanic trance included increased gamma power (correlated with visual alterations), decreased gamma‑band signal diversity, and increased criticality in beta and gamma bands, with some metrics suggesting trait‑like alterations. The investigators conclude that shamanic trance and psychedelic drug‑induced states are distinct but overlapping altered states of consciousness, implying non‑pharmacologic, endogenous mechanisms that produce large shifts in phenomenology via neural processes different from those engaged by psychedelics.