Acute Biphasic Effects of Ayahuasca

Ayahuasca is an Amazonian plant brew containing beta-carbolines (harmine, harmaline, tetrahydroharmine or THH) and the psychedelic tryptamine DMT. Earlier pharmacological work established that the beta-carbolines act as monoamine-oxidase A (MAO-A) inhibitors in the periphery, enabling orally ingested DMT to reach the brain, and prior neurophysiological studies using EEG have reported a mixture of findings including alpha decreases, theta/delta changes, beta increases and occasional reports of gamma-band effects. These prior studies varied in setting (ritual versus laboratory), timing after ingestion, dose regimens and the use of lyophilised versus liquid preparations, leaving unresolved questions about the temporal dynamics and pharmacological correlates of ayahuasca's effects on brain oscillations. Schenberg and colleagues therefore set out to record continuous EEG alongside repeated plasma sampling after a standardised dose of liquid ayahuasca in experienced volunteers in a controlled laboratory setting. The principal aim was to characterise the time course of oscillatory changes across frequency bands and to relate these changes to circulating concentrations of DMT, beta-carbolines and metabolites, thereby clarifying whether distinct early and late electrophysiological phases of the ayahuasca experience exist and which compounds are associated with each phase. The study combines spectral EEG analysis with pharmacokinetic measures to investigate neural and pharmacological correlates of the acute ayahuasca state.

Twenty healthy adult volunteers with prior ayahuasca experience were recruited and screened; exclusion criteria included age under 21, personal history of psychiatric illness, current psychiatric medication, cardiovascular disease and past neurological injury. After informed consent and psychiatric assessment, participants attended EEG sessions in a private room within a psychiatric facility where they remained seated, eyes closed and in a resting-state posture for the recordings. One male volunteer did not drink ayahuasca due to baseline EEG noise and two subjects had their entire EEG data discarded for excessive movement, leaving EEG data from the remaining participants. Two volunteers declined blood sampling and one blood sample set was discarded for hemolysis, resulting in plasma data from 16 individuals. A 6-litre sample of liquid ayahuasca (Hoasca) donated by a church was characterised by LC-MS/MS; the liquid was concentrated to reduce volume consumed. The measured concentrations in the brew were reported as 0.328 mg/ml DMT, 1.08 mg/ml harmine, 0.18 mg/ml harmaline and 1.28 mg/ml THH, yielding approximate per‑kg doses (reported in the paper) of 1.39 mg/kg DMT, 4.58 mg/kg harmine, 0.75 mg/kg harmaline and 5.43 mg/kg THH. Blood was collected via an indwelling catheter at ten time points: 25 minutes before ingestion (baseline) and then every 25 minutes up to 200 minutes after ingestion; plasma was frozen and analysed for a panel of tryptamines, beta-carbolines and metabolites. EEG was recorded for two hours with a 64-channel system referenced to Cz and re-referenced offline to the common average. Data were sampled at 2 kHz and later downsampled to 500 Hz; bandpass filtering (0.5–150 Hz) was applied during preprocessing. Data were divided into ten-minute windows aligned with blood draws (baseline, 25, 50, 75, 100 and 125 minutes post-ingestion were analysed) and subjected to visual inspection, channel interpolation, epoching (3 s), and independent component analysis (infomax) to remove line noise, muscle and ocular artefacts. Spectral estimation used a Hanning window for 1–30 Hz and Slepian multitapers for 30–100 Hz with ±10 Hz smoothing. Six frequency bands were defined a priori: delta (1–4 Hz), theta (4–8 Hz), alpha (8–13 Hz), beta (13–30 Hz), slow-gamma (30–50 Hz) and fast-gamma (50–100 Hz). Statistical analysis employed a non-parametric cluster-based approach: permuted F statistics with 5,000 randomisations to detect spectral changes across time, followed by Dunnett's test for post-hoc comparisons to baseline. Relationships between plasma compound concentrations and spectral changes were evaluated using generalised linear models (MATLAB glmfit), with the resulting maps corrected by the same cluster-based permutation procedure (5,000 permutations). Significance was taken at p < 0.05; effect sizes reported included partial eta squared for main clusters and Cohen's d for GLM results.

The primary electrophysiological finding was a biphasic pattern of spectral change. The first phase comprised a reduction in alpha-band (8–13 Hz) power peaking around 50 minutes after ingestion, localised most prominently to the left parieto‑occipital / centro-parieto‑occipital electrodes. The second phase comprised increases in both slow‑gamma (30–50 Hz) and fast‑gamma (50–100 Hz) power emerging between 75 and 125 minutes post-ingestion, with significant clusters in left centro‑parieto‑occipital, left fronto‑temporal and right frontal cortices and, for fast‑gamma, additional right parieto‑occipital sites. Post-hoc tests showed these gamma increases were significant at the 75, 100 and 125 minute windows in several clusters. Plasma analyses (available for 16 participants) identified detectable peripheral levels of DMT, NMT, DMT‑N‑oxide and a range of beta‑carbolines and some metabolites; several compounds (e.g. 5‑MeO‑DMT, 2‑MTHBC, DMT‑NO not detected in the brew) were absent. Time‑concentration measures (Cmax, Tmax, AUC) were summarised in a table in the paper; harmine and DMT peaked earlier than THH and harmaline. Age correlated positively with indices of indoleacetic acid (IAA) exposure and showed marginal positive correlations with DMT exposure. Two volunteers did not provide blood and one sample set was excluded for hemolysis. Linking EEG and pharmacokinetics, the authors report multiple significant associations between plasma concentrations and spectral changes. DMT concentration correlated strongly with the alpha decrease and also showed clusters in the gamma bands, suggesting contribution to both early alpha suppression and later high‑frequency increases. DMT‑N‑oxide and NMT were associated mainly with gamma‑band increases in frontal and parieto‑occipital regions. Beta‑carbolines (harmine, harmol, harmaline, harmalol and THH) showed widespread associations particularly with gamma‑band power increases; harmine and THH also correlated with alpha reductions in left centro‑parieto‑occipital areas. Harmaline showed especially strong associations with fast‑gamma increases and its Tmax correlated with time to vomiting. Subjectively, all participants reported noticeable alterations (visual imagery with eyes closed, heightened sensitivity to ambient sounds, distortions of time and space, introspection and emotional arousal). Almost all participants vomited (reported mean time to vomit approximately 95 minutes post‑ingestion, range 25–242 minutes), but vomiting time did not correlate with compound Cmax or AUC except for a positive correlation between harmaline Tmax and time to vomit. Psychometric scores on the Hallucinogen Rating Scale (Brazilian version) did not show significant correlations with spectral alterations or pharmacokinetic parameters.

Schenberg and colleagues interpret their results as evidence that ayahuasca produces an acute biphasic electrophysiological profile: an early phase (around 50 minutes) characterised by alpha‑band suppression, and a later phase (75–125 minutes) marked by increased slow‑ and fast‑gamma power. They link the early alpha decreases chiefly to DMT and harmine, which peaked earlier in plasma, and attribute the later gamma enhancements to harmaline and THH, which displayed slower pharmacokinetics. The authors note that these temporal associations are consistent with known absorption dynamics of the brew and provide a pharmacokinetic explanation for why sequential doses in ritual contexts might produce overlapping effects. In considering mechanisms, the paper highlights the likely involvement of cortical 5‑HT2A receptor activation in alpha suppression, drawing parallels with psilocybin studies that implicate 5‑HT2A signalling and posterior cortical effects. The observed gamma increases are discussed in the context of theories that associate gamma synchrony with the binding of distributed neural representations, attention, memory and conscious reportability; the authors suggest these increases could be related to enhanced awareness of internal psychological states and imagery. They explicitly address concerns that high‑frequency EEG effects might arise from muscle artefact and offer four reasons supporting a neural origin: (1) data‑loss due to movement did not increase over time, (2) muscle artefact typically spans a broader frequency range including beta, where effects were absent, (3) the spatial topography of gamma effects (left‑dominant slow‑gamma) is inconsistent with global muscle contamination, and (4) careful preprocessing including ICA and visual inspection detected neural‑like high‑frequency tracings while subjects were still. The authors further discuss the roles of individual compounds beyond peripheral MAO inhibition. Harmine is emphasised as a centrally active MAO‑A inhibitor with some 5‑HT2A affinity and broad associations with EEG changes; harmaline is noted both for strong electrophysiological correlations and for its likely contribution to emetic effects. THH showed associations with alpha and gamma changes similar to harmine. Metabolites such as DMT‑N‑oxide and NMT also correlated with gamma activity, raising the possibility of central effects for metabolites though the authors remain cautious because psychoactivity of those metabolites is not established. Emesis is considered both a physiological side‑effect and, in ritual contexts, a culturally meaningful cleansing process; the authors report harmaline’s Tmax correlation with vomiting latency and note that vomiting did not reliably disrupt early DMT and harmine exposure. Limitations acknowledged by the authors include the lack of blinding and placebo control, which they argue is difficult to implement with experienced participants given the unmistakeable subjective effects, and the artificial lab setting that omits ritual elements such as music and communal interaction. They also note that the Hallucinogen Rating Scale used has not been fully validated in large samples, which may account for the absence of correlations between psychometrics and EEG/pharmacokinetic measures. Finally, the authors caution that while their preprocessing reduces the likelihood of artefact-driven gamma findings, residual confounds cannot be excluded entirely.

The study reports acute, biphasic effects of a standardised liquid ayahuasca preparation on human brain oscillations: left centro‑parieto‑occipital alpha decreases around 50 minutes post‑ingestion followed by widespread slow‑ and fast‑gamma increases between 75 and 125 minutes. These electrophysiological phases covaried with circulating concentrations of DMT and multiple beta‑carbolines, supporting a view that interactions among several ayahuasca constituents shape the temporal profile of the altered state. The authors present these findings as relevant both for understanding ritual ayahuasca use and for considering therapeutic potentials in conditions such as depression, anxiety, PTSD and substance dependence, while noting the need for further research that integrates pharmacology, neural dynamics and contextual factors.