Topographic pharmaco-EEG mapping of the effects of the South American psychoactive beverage ayahuasca in healthy volunteers

Ayahuasca is a traditional Amazonian psychoactive brew prepared from Banisteriopsis caapi together with leaves of plants such as Psychotria viridis, combining b-carboline MAO inhibitors and the psychedelic N,N-dimethyltryptamine (DMT). Prior work by the study team and others showed that oral ayahuasca produces a mixed stimulant–psychedelic subjective profile that is longer than intravenous DMT but shorter than LSD or mescaline; however, objective neurophysiological characterisation in controlled conditions remained limited. Concerns about adverse reactions in contemporary users and the growing international use of ayahuasca motivated further clinical characterisation of its central nervous system effects. Riba and colleagues set out to assess the cerebral bioavailability and time course of ayahuasca’s central effects using topographic quantitative electroencephalography (q-EEG) in a double-blind, randomised, placebo-controlled crossover design. The specific aims were to determine whether EEG can objectively detect dose-dependent changes after two oral doses of encapsulated freeze-dried ayahuasca in healthy volunteers with prior psychedelic experience, and to compare the q-EEG profile with subjective effects measured by the Hallucinogen Rating Scale (HRS).

Eighteen healthy volunteers (15 men, 3 women; mean age 25.7 years) with no personal or first-degree family history of Axis I psychiatric disorder and with prior recreational psychedelic use (at least five occasions) were enrolled. Exclusion criteria included current or past Axis I disorders, dependence, and high trait anxiety. Two volunteers had previously used ayahuasca; all participants had prior cannabis and cocaine exposure. Screening included structured psychiatric interview, trait-anxiety assessment, physical exam, laboratory tests, ECG and urine drug screening. Written informed consent and ethical approval were obtained. The investigational product was lyophilised ayahuasca administered in gelatin capsules. Two active doses were used, calibrated to deliver the equivalent of 0.6 mg DMT kg and 0.85 mg DMT kg body weight. Analysis of the freeze-dried material reported concentrations per gram of approximately 8.33 mg DMT, 14.13 mg harmine, 0.96 mg harmaline and 11.36 mg tetrahydroharmine (THH), with corresponding concentrations estimated for the original tea. Capsules of varying amounts of freeze-dried material plus lactose placebo were combined so each volunteer swallowed the same number of capsules on each experimental day. The study used four experimental sessions per volunteer. All participants received a single-blind placebo on the first day to familiarise them with the setting; on the following three sessions they received, in a double-blind balanced order, randomized placebo, the low ayahuasca dose and the high ayahuasca dose, with one treatment repeated. A 7 day washout separated sessions. Participants fasted on experimental mornings, took the capsules about 10:00 with 250 ml water, and remained in a quiet, dim room reclined for recordings. Subjective questionnaires were completed at 4 h, and the last EEG recording occurred at 8 h; discharge was about 9 h post-dose. EEG was recorded from 19 electrodes (10/20 system) referenced to averaged mastoids, plus vertical and horizontal EOG, using a 0.3–30 Hz bandpass and 100 Hz sampling. Vigilance-controlled eyes-closed 3 min EEG epochs were recorded at multiple time points: two predrug baselines and at +15, +30, +45, +60, +90, +120, +150, +180, +210, +240, +360 and +480 min post-administration. Artefact handling combined regression-based ocular correction and automatic time/frequency-domain rejection, followed by visual inspection. Spectral analysis (1.3–30 Hz, 0.2 Hz resolution) quantified 34 target variables including total, absolute and relative power in multiple frequency bands (delta/theta, alpha-1/2, beta-1 to beta-4), dominant frequency measures, centroids and centroid deviations, and an alpha/delta-theta vigilance index; topographic maps used cubic interpolation. Subjective effects were assessed with a Spanish adaptation of the Hallucinogen Rating Scale (HRS), which yields six subscales (Somaesthesia, Affect, Volition, Cognition, Perception, Intensity). Statistical analysis of EEG followed International Pharmaco-EEG Group guidance and used a descriptive data analysis (DDA) framework: multivariate tests (Hotelling T2), paired t-tests for univariate comparisons, and nonparametric Friedman and Wilcoxon sign-adjusted tests on multiple leads and variables to explore dose and time effects. Significance probability maps were used to display spatial–temporal patterns of P values rather than applying strict multiplicity corrections. HRS scores were analysed by repeated-measures ANOVA with Greenhouse-Geisser correction and Bonferroni-corrected pairwise tests where appropriate.

Multivariate topographic analyses showed dose-dependent central effects on EEG. After the low ayahuasca dose (0.6 mg DMT kg) statistically significant differences from placebo were limited and spatially isolated between about 45 min and 2.5 h. The high dose (0.85 mg DMT kg) produced widespread EEG changes first reaching significance at about 1 h, peaking between 1.5 and 2 h, and progressively declining to baseline by 6–8 h. At peak pharmacodynamic effect the strongest changes were evident centrally and over right temporo‑occipital electrodes. Total power and centroids: Ayahuasca induced a significant, dose-dependent reduction in total absolute power across the scalp, with a temporal maximum around 90 min after the high dose. Concurrently, the centroid (centre-of-gravity frequency) of total activity and its deviation increased, peaking at roughly 120 and 90 min respectively. Slow activity (delta/theta): Absolute power of the combined delta–theta band decreased dose‑dependently, with maximal reductions at 90 min for the low dose and between 90–120 min for the high dose. When separated, theta showed the most pronounced absolute decreases (dose-dependent, peak 90–120 min), while relative power decreases were especially prominent and widespread for theta at about 120 min; delta relative-power decreases were more marginal. The centroid of the combined delta–theta band showed a modest deceleration with increased variability at selected leads. Alpha activity: Absolute alpha power fell in a dose-dependent manner, most notably in alpha-2 and at left-temporal and centro‑parieto‑occipital sites, with maxima at about 60–90 min depending on subband. Relative alpha activity paradoxically increased at 120 min (primarily in alpha‑1), showing an inverse dose–response pattern with the largest relative increase after the low dose; alpha-2 relative power tended to decrease, peaking at 60 min. Dominant-frequency measures showed no consistent shifts. Fast activity (beta): Absolute global beta power decreased dose-dependently, most strongly in beta-1, with peak decrements around 90 min. In contrast, relative beta power increased significantly and dose-dependently, more intensely and for longer duration after the high dose; the increases were greatest in the faster beta-3 and beta-4 subbands (maximal between 45 and 90 min) and showed central and parietotemporal predominance. The centroid of the beta range shifted significantly towards higher frequencies at about 90 min. Nonparametric multilead tests corroborated the dose and time effects: at the low dose ayahuasca could be differentiated from placebo only at about 45–60 min, whereas at the high dose statistically significant differences from placebo were present from 45 through 120 min. Rank‑sum plots indicated EEG changes beginning between 15 and 30 min, a steep rise at 45 min, a pharmacodynamic peak between 45 and 90 min (high dose), and a return towards baseline by 4–6 h post-dose. Subjective effects: HRS scores increased dose‑dependently across all six subscales. The low ayahuasca dose produced statistically significant increases on all HRS subscales except Volition; the high dose produced significant increases on all subscales including Volition. Reported subjective phenomena included paresthesias, predominantly visual perceptual changes (less so auditory), and alterations in thought and emotion described qualitatively as dream-like. The temporal course of subjective effects paralleled the q-EEG time course.

Riba and colleagues conclude that ayahuasca’s central effects can be objectively measured by q-EEG and that these neurophysiological changes closely parallel the time course of subjective psychedelic effects. The q-EEG signature was characterised by a global reduction in absolute power across frequency bands, most markedly in theta and delta, together with a shift of centre‑of‑gravity frequencies to higher values and relative increases in fast activity (beta), particularly in the beta‑3 and beta‑4 ranges. Effects were dose-dependent, emerged within 15–30 min, peaked between roughly 45 and 120 min and largely resolved by 4–8 h. When compared with older and more limited EEG reports of classical serotonergic psychedelics, the observed pattern—reduction of slow activity and a relative enhancement of fast rhythms—bears similarities to prior qualitative descriptions of LSD and related agents. The authors note, however, that the detailed pharmaco‑EEG profile does not map cleanly onto any single psychotropic drug class. In particular, the combination of theta suppression and relative beta enhancement shows overlap both with stimulant or serotonin‑releasing agents (which reduce delta/theta) and with monoaminergic antidepressants (which increase relative beta), suggesting mixed pro‑serotonergic and pro‑dopaminergic features in ayahuasca’s CNS effects. Mechanistically, the investigators attribute primary importance to DMT’s agonism at 5‑HT2A/2C receptors as a driver of the psychedelic profile. They reason that the marked theta decreases argue against a dominant 5‑HT1A contribution, since selective 5‑HT1A agonists (for example buspirone) tend to increase theta power. The authors also propose that MAO‑inhibiting b‑carbolines in the brew (harmine, harmaline, THH) may augment synaptic monoamines and contribute to the relative beta‑promoting effects, and they discuss evidence that dopaminergic release may accompany classical psychedelic administration. On this basis the q‑EEG findings are interpreted as supporting involvement of 5‑HT2 and dopamine D2 signalling in ayahuasca’s central effects. The authors acknowledge limitations in comparing their results with older EEG literature and with a single prior field study conducted without placebo control and without the same frequency bandwidths. They also note the complexity of attributing EEG signatures to discrete receptor actions given the mixed pharmacology of the brew. Nevertheless, they emphasise that the controlled double‑blind design, multi‑lead q‑EEG approach and concordance with subjective measures provide convergent evidence for ayahuasca’s cerebral bioavailability and transient psychoactive effects.