Neural and subjective effects of inhaled N,N-dimethyltryptamine in natural settings

Pallavicini and colleagues situate their study within the recent resurgence of human research on serotonergic psychedelics, noting that prior laboratory work with compounds such as psilocybin, LSD and DMT has repeatedly reported broadband reductions in spontaneous oscillatory activity (particularly alpha-band) and increased signal diversity, alongside altered functional connectivity on fMRI. The authors highlight that most existing neurophysiological data come from controlled clinical or imaging settings, which limit ecological validity because psychedelic experiences are highly sensitive to 'set' (internal state) and 'setting' (external environment). They argue that naturalistic field investigations can capture contextual influences that laboratory experiments typically cannot, and that advances in mobile EEG make such recordings feasible. This study set out to characterise the acute neural and subjective effects of inhaled freebase DMT consumed in participants' preferred, self-chosen settings. The investigators combined wireless 24-channel EEG recordings with a battery of validated psychometric instruments to examine spectral power, signal diversity and measures of functional synchrony, and to relate these neural measures to detailed assessments of the subjective experience and of contextual factors such as intention, social setting and preparation.

This observational field study recruited 35 healthy, experienced participants (7 females; mean age 33.1 ± 6 years) between May and December 2019 by word-of-mouth and social media. Inclusion required at least two prior experiences with ayahuasca or DMT, abstinence from psychoactive substances (including alcohol, caffeine and tobacco) for at least 24 hours before the session, and willingness to consume DMT in the presence of four research team members. A non-diagnostic psychiatric interview screened for exclusionary conditions including psychotic or bipolar disorders (including family history), recent substance dependence (except nicotine), recurrent depressive or other specified diagnoses, neurological disorders, elevated baseline anxiety (more than 1 SD above the mean on the State-Trait Anxiety Inventory) and current psychiatric medication. Participants provided informed consent and determined the time and place of their DMT use; the research team did not supply the drug or direct how it was used. Participants consumed DMT freebase vapour derived from Mimosa hostilis recrystallised over non-psychoactive plant leaves (typical facilitator/participant estimate of pipe load ≈ 40 mg freebase). Most subjects were assisted by a facilitator and used their usual contextual elements (music, scents, lighting); four self-administered. Baseline EEG recordings comprised five minutes eyes open and five minutes eyes closed. Post-inhalation EEG began at exhalation and continued until the participant indicated a return to baseline (mean recording duration 6 ± 1.4 minutes). DMT presence in samples was confirmed by HPLC–mass spectrometry for profiling and qualitative analysis. Subjective measures administered immediately before and after the experience included the State-Trait Anxiety Inventory (STAI), a 12-item set assessing set/setting/intentions, the 5D-ASC (94 items), MEQ-30 (mystical experience), NDE (near-death experience) scale, a 27-item post-experience questionnaire addressing set/setting/social/fusion, the Big Five Inventory (BFI) and the Tellegen Absorption Scale (TAS). These instruments were used to quantify altered states, mystical-type experiences, near-death-like features, personality and absorption. EEG acquisition used a 24-channel wireless system (mBrainTrain) with electrodes at standard 10-20 sites; amplifier sampling rate was 500 Hz. Preprocessing in EEGLAB divided data into 2 s epochs, applied 1–90 Hz bandpass and a 47.5–52.5 Hz notch filter, and used automated plus manual inspection to flag and remove bad channels and epochs. Automated detection flagged on average 4.4 ± 1.8 channels per subject; manual inspection led to interpolation of rejected channels (mean 30% rejected channels, maximum 8). Epoch rejection averaged 21.3 ± 13.7 per subject. Infomax ICA was applied to remove eye and muscle components (mean 2.7 ± 1.1 components removed). Six participants were excluded from EEG analyses due to excessive artifact, leaving 29 subjects for spectral and complexity analyses. Spectral analyses computed logarithmic power spectral density (LPSD) in delta (1–4 Hz), theta (4–8 Hz), alpha (8–12 Hz), beta (12–30 Hz) and gamma (30–40 Hz) bands using FFT and time-frequency decomposition with Morlet wavelets. Signal diversity was estimated using the Lempel–Ziv complexity algorithm, and measures of coherence and metastability were derived to characterise synchrony and variability of oscillatory activity. Statistical comparisons used paired t-tests (DMT vs eyes-closed and vs eyes-open as appropriate) with multiple-comparison correction (Benjamini–Hochberg FDR or Bonferroni where stated). Correlations between EEG features and questionnaire scores used Pearson’s r and FDR correction, with significance at p<0.05 after correction.

Participant-reported adequacy ratings of pre-experience factors averaged 71.4% ± 27.4% for 'set', 88.2% ± 10.1% for 'setting' and 53.9% ± 23.8% for 'clear intentions' (means ± SD expressed as % of maximum). On the 5D-ASC, the largest effects were in the visionary restructuralization domain (elementary imagery, complex imagery, audio–visual synesthesia) and in the blissful state facet of oceanic boundlessness; anxious ego dissolution items (impaired control, anxiety) were comparatively low. The NDE scale scored highest on its affective component. Using a 60% cut-off on each MEQ-30 subscale to define a 'complete mystical-type experience', 13 of 35 participants (37%) met this criterion. Comparing pre- to post-experience measures, agreeableness (BFI) and absorption (TAS) increased significantly, while state anxiety (STAI state) decreased. Baseline individual differences related to the acute response: 'clear intentions' correlated with the MEQ-30 mystical factor (R = 0.36, p = 0.029), baseline trait absorption correlated with the MEQ-30 mystical factor (R = 0.50, p = 0.005), and baseline neuroticism predicted the impaired control/cognition item on the 5D-ASC (R = 0.44, p = 0.007). Baseline state anxiety predicted both impaired control/cognition (R = 0.41, p = 0.012) and anxiety (R = 0.65, p = 0.001) during the DMT state. Spectral power analyses showed a widespread and significant attenuation of alpha-band (8–12 Hz) power under DMT compared with the eyes-closed baseline, with all electrodes exhibiting decreases. Relative to eyes-closed baseline, DMT also produced significant increases in low-frequency (< 3 Hz / delta) and high-frequency (> 36 Hz / gamma) power; no consistent changes were observed for theta or beta bands. Spatially, delta increases were most evident in occipital, parietal and anterior-central electrodes, whereas gamma increases were prominent in occipital, parietal and temporal regions. No significant differences were found between DMT and the eyes-open condition for overall power. Time-frequency analysis of the first seven minutes post-inhalation (mean EEG duration 6 ± 1.4 min) revealed that alpha power decreased most strongly in the first 40 s and returned progressively toward baseline, with statistically significant alpha reductions persisting until approximately 3 minutes after onset and LPSD differences approaching zero by about 7 minutes. The temporal evolution of alpha topographies showed non-uniform recovery across the scalp, with occipital and parietal electrodes exhibiting more persistent reductions. Correlational analyses relating subjective measures to band-specific LPSD found significant associations largely in the beta and gamma bands. Central beta power correlated positively with the 'anxiety' item of the 5D-ASC and with the 'cognition' component of the NDE scale across several regions. Occipital gamma power correlated with items including 'experience of unity', 'disembodiment' (5D-ASC), 'mystical' and 'transcendence of time and space' (MEQ-30) and the NDE 'cognition' item. Central, frontal and temporal gamma also showed positive correlations with subsets of mystical-type and imagery items. The temporal profiles of these correlations varied; some peaked early (around the peak subjective intensity) while others peaked later. Measures of functional synchrony and complexity differed under DMT: coherence and metastability decreased in the alpha band and increased in the gamma band when comparing DMT to eyes-closed baseline. Lempel–Ziv complexity increased across channels under DMT relative to eyes-closed, mirroring increases seen when comparing eyes-open to eyes-closed. The authors did not find significant correlations between the psychometric questionnaire scores and EEG coherence, metastability or Lempel–Ziv complexity.

Pallavicini and colleagues interpret their findings as broadly consistent with prior laboratory reports of psychedelic-induced reductions in alpha power and increased signal diversity, while also identifying effects that may reflect the influence of naturalistic context. They emphasise that recruiting participants who had already planned and obtained DMT allowed a larger sample for field EEG recordings than many laboratory studies, and that the profile of subjective effects—especially prominent visionary restructuralization and strong affective NDE-like responses—resembled previous DMT reports. The authors highlight the notable association between increased gamma-band power and multiple questionnaire items indexing mystical-type and near-death-like phenomenology. They discuss possible interpretations carefully, noting that gamma activity can be contaminated by muscle artefacts (jaw clenching, microsaccades) but that the observed gamma effects persisted after conservative preprocessing and were spatially localised (predominantly occipital and central). They suggest that contextual factors related to set and setting might facilitate experiences whose neural correlates manifest in the gamma range, and point to converging but heterogeneous evidence from other psychedelic and meditative studies linking high-frequency synchrony to unitive or peak states. Differences in the temporal dynamics of alpha suppression compared with prior intravenous DMT work are considered in light of administration route and dose uncertainty: inhaled DMT may have faster clearance and shorter duration than intravenous dosing, and effective inhaled dose is difficult to quantify because combustion/inhalation efficiency is variable. The authors note an observed increase in delta power that diverges from some prior reports but accords with certain ayahuasca and animal studies; they attribute cross-study heterogeneity partly to differing pharmacological profiles and measurement approaches. Key limitations acknowledged by the study team include the impossibility of double-blinding and placebo control in this field design, lack of precise dose quantification and variable inhalation efficiency, potential expectation effects, and the decision not to collect follow-up data to preserve participant anonymity. The investigators frame field studies as complementary to controlled trials and large surveys: they argue that naturalistic recordings can illuminate how set, setting and intention interact with neurochemistry, and that such knowledge is relevant given the increasing unregulated therapeutic use of psychedelics. The authors conclude that wireless EEG can successfully record neural correlates of DMT experiences in real-world contexts and recommend future field studies to further probe interactions between context, subjective phenomenology and brain activity.