Interactive Effects of Ayahuasca and Cannabidiol in Social Cognition in Healthy Volunteers: A Pilot, Proof-of-Concept, Feasibility, Randomized-Controlled Trial

Earlier research indicates that serotonergic hallucinogens and the endocannabinoid system can alter recognition of emotions in facial expressions (REFE) and empathy, but the role of cannabidiol (CBD) in moderating such effects is unclear. CBD is a negative allosteric modulator at CB1 and has activity at serotonergic receptors, while ayahuasca is a dimethyltryptamine (DMT)‑containing decoction whose effects on endocannabinoids have been sparsely described. The clinical relevance of cannabinoid–psychedelic interactions for social cognition and for therapeutic outcomes in conditions such as depression and anxiety therefore remains uncertain. Rossi and colleagues designed this pilot, proof‑of‑concept randomised controlled trial to evaluate whether oral CBD (600 mg) alters the acute and short‑term effects of a single oral dose of ayahuasca (1 mL/kg) on social cognition in healthy volunteers. Primary outcomes were performance on REFE and an empathy task; secondary outcomes included subjective effects, tolerability, plasma ayahuasca alkaloids, brain‑derived neurotrophic factor (BDNF), and cardiovascular and liver safety measures. The investigators hypothesised that ayahuasca would reduce recognition of negative emotions and produce anxiolysis, that CBD could moderate REFE effects, and that combined CBD–ayahuasca would yield greater anxiolysis and attenuated psychedelic effects compared with placebo plus ayahuasca.

This was a single‑session, parallel‑arm, randomised, double‑blind, placebo‑controlled proof‑of‑concept trial conducted between September 2018 and March 2020. Seventeen healthy volunteers were enrolled after telephone screening with a Structured Clinical Interview for DSM‑5 and in‑person assessment; 9 were randomised to receive 600 mg oral CBD and 8 to placebo. Eligibility required age 18–65, literacy, and no more than two lifetime uses of classic psychedelics; exclusions included current or past cardiovascular, liver or neurological disease, psychiatric diagnoses, psychoactive medication use, recurrent substance misuse, and pregnancy or lactation. All participants provided written informed consent and the protocol had ethics approval. Participants received a standardised breakfast to enhance oral CBD absorption, then two identical capsules containing either 600 mg CBD (300 mg per capsule) or sunflower oil placebo at around 08:00. Ninety minutes later, all volunteers ingested a single batch of ayahuasca at 1 mL/kg (single batch donated by a Santo Daime church), and the session included eight in‑session data points from baseline to +390 minutes and follow‑ups at 1 and 7 days. The investigators maintained a neutral, supportive setting without formal psychotherapeutic interventions; researchers were blinded to CBD/placebo allocation but not to ayahuasca administration (all participants received ayahuasca). Primary cognitive measures were a dynamic computerized REFE task (six basic emotions) and the Multifaceted Empathy Test (MET) assessing cognitive and emotional empathy, both administered at baseline and repeated during and after the session per protocol. Subjective effects were measured repeatedly with the Visual Analog Mood Scale (VAMS) at all eight in‑session time points; spontaneous participant reports and observer impressions were also recorded. Tolerability outcomes included spontaneous adverse reports, heart rate and blood pressure at all time points, and hepatic enzyme panels at baseline and +390 minutes. Four venous blood samples (baseline, +90, +180, +390 minutes) were taken for plasma ayahuasca alkaloids (DMT, harmine, THH, harmaline) quantified by ultraperformance liquid chromatography–tandem mass spectrometry, and for total BDNF by ELISA. Because this was exploratory, no formal sample size calculation was performed. Quantitative outcomes (MET, REFE, VAMS, cardiovascular variables) were analysed with two‑way repeated‑measures ANOVA with Time and Group factors; significant effects were followed by pairwise t tests with Bonferroni correction. BDNF analyses used nonparametric tests due to small group sizes. A post hoc analysis tested correlations between DMT plasma levels and VAMS scores using Kendall rank correlations at individual times and Pearson or nonparametric correlations when pooling observations. Significance was set at P < 0.05 and effect sizes reported as eta‑squared (ƞ2).

Seventeen volunteers were randomised and all completed the experimental session and both follow‑ups, yielding 100% adherence and no dropouts or serious adverse events. Groups were comparable demographically (mean age 25.3 years, mean weight 68.0 kg). No experimental sessions were interrupted. Primary social cognition outcomes: For the REFE task there were significant main effects of Time on general reaction time and on reaction times for sadness, fear and disgust, indicating reductions in reaction time over repeated measurements (e.g. general reaction time F(2.875,37.379)=6.479, P=0.001, ƞ2=0.279). There were no significant Time × Group interactions and no significant effects on accuracy overall or for individual emotions. For the MET, mean reaction times for general cognitive empathy and affective empathy and for both positive and negative emotions decreased over time (for example, general cognitive empathy reaction time F(2,30)=10.916, P<0.001, ƞ2=0.421), again with no Time × Group interactions and no changes in accuracy or intensity measures. Subjective effects: VAMS factors showed significant Time effects, with reductions over the session in Anxiety (F(5.142,77.129)=5.383, P<0.001, ƞ2=0.262), Sedation (F(7,105)=5.471, P<0.001, ƞ2=0.267), Cognitive deterioration (F(3.301,49.517)=4.984, P=0.003, ƞ2=0.249), and Discomfort (F(3.558,53.337)=2.652, P=0.049, ƞ2=0.150). There were no significant Group or Time × Group effects for these subjective measures, indicating similar time courses in CBD and placebo groups. Self‑reports and observer impressions: Acute positive effects typical of ayahuasca were commonly reported in both groups, including visual/perceptual changes (62%–66%), tranquillity (62%–77%) and well‑being (44%–75%). Acute adverse effects included gastrointestinal discomfort and nausea (44%–62%); vomiting occurred in a single volunteer (placebo group). Drowsiness was reported only in the CBD group (12%). One transient dysphoric episode with agitation, anxiety, confusion and dissociation occurred (placebo group) and was managed without medical intervention; two volunteers reported delayed adverse events after the session (a headache in a CBD participant who later vomited after taking ibuprofen, and transient distraction in another participant at day 7). Tolerability and safety physiology: Heart rate and diastolic blood pressure showed transient increases over time (heart rate F(7,105)=6.201, P<0.001, ƞ2=0.293; diastolic BP F(7,105)=3.0, P=0.007, ƞ2=0.167), while systolic blood pressure showed a significant reduction over time in the ANOVA model. Transient moderate hypertension (140–160/92–105 mm Hg) was observed after baseline in seven of 17 volunteers (41%); episodes were short‑lived. No clinically significant cardiovascular complications occurred. Hepatic enzymes: Blood draws for liver enzyme analysis were available from 15 participants. Seventy­six percent (23 of 30 assays) were within reference norms. Seven assays were outside normative ranges (four at baseline, three at +390 minutes); five volunteers (29% of analysed sample) had one or more abnormal results but deviations were not sufficient to characterise liver damage. Plasma alkaloids and BDNF: Alkaloid assays were available for 13 volunteers; all had undetectable alkaloids at baseline and immediately before ayahuasca. Measured mean DMT concentrations were 5.40 ± 3.71 ng/mL (range 0.7–12.4) at ~90 minutes after ayahuasca and 1.65 ± 1.35 ng/mL (range 0.0–4.3) at ~300 minutes. Mean values at the later time point for harmine, THH and harmaline were reported (harmine 8.52 ± 12.38 ng/mL; THH 62.32 ± 35.01 ng/mL; harmaline 1.31 ± 0.50 ng/mL) and did not differ significantly between groups. BDNF analyses yielded no significant within‑ or between‑group differences; visual inspection suggested opposing trends (BDNF decreased from baseline to +180 minutes in the CBD–ayahuasca condition and increased in the placebo–ayahuasca condition), but sample sizes were small. Correlations: A negative correlation was observed between DMT plasma levels and the VAMS Sedation factor at +180 minutes (P=0.033, τ = -0.683), but this relationship was not present at +390 minutes nor when pooling observations. Overall, there was no evidence of CBD moderating ayahuasca effects on the primary cognitive or subjective endpoints.

Rossi and colleagues report that concomitant administration of 600 mg oral CBD with a single dose of ayahuasca did not produce interactive effects on social cognition in healthy volunteers. The principal behavioural findings were reductions in reaction times on REFE and MET tasks over repeated measurements, interpreted by the authors as faster emotional processing; however, accuracy measures did not change and there were no differences between CBD and placebo groups, so the investigators caution that changes may partly reflect learning effects from repeated testing. Subjectively, ayahuasca produced typical psychedelic effects and an apparent acute anxiolytic signal on VAMS factors; again, no between‑group differences were observed, leading the authors to attribute these effects to ayahuasca rather than to CBD. In terms of safety and tolerability, the study reproduced expected acute adverse effects of ayahuasca (gastrointestinal discomfort, nausea, transient increases in heart rate and blood pressure, visual changes). Vomiting was rare (one case). Liver enzyme assays showed some out‑of‑range values in a minority of participants but no pattern consistent with acute hepatic injury. The researchers observed no clinically significant cardiovascular or hepatic harm from combined administration in this small sample and no serious adverse events. The authors position their results within prior literature showing mixed effects of psychedelics and CBD on emotion recognition and empathy: some previous studies detected reductions in recognition of fear with psilocybin or LSD, while others found null results for accuracy; earlier CBD studies also show inconsistent effects dependent on dose and route. They note that healthy volunteers may be less likely to show improvements in social cognition than clinical samples. The pharmacokinetic data confirmed absorption of ayahuasca alkaloids, although measured DMT concentrations were lower than in some prior reports, possibly due to lower β‑carboline content in the batch used and variability in ayahuasca preparations. Key limitations acknowledged by the investigators include the small sample size (the planned sample was larger but recruitment was curtailed by the COVID‑19 pandemic), the lack of a true ayahuasca placebo control (all participants received ayahuasca), potential learning effects from repeated cognitive testing, absence of a psychedelic‑specific validated scale in Portuguese to quantify subjective psychedelic phenomena, and underpowered biochemical analyses (e.g. BDNF). The authors recommend that future trials use larger samples, consider alternative CBD dosing regimens, measure CBD and endocannabinoid levels, standardise alkaloid dosing where possible, and include validated instruments to assess dissociation and psychedelic effects.

In this randomised, double‑blind, placebo‑controlled pilot trial of 17 healthy volunteers, the combination of oral CBD (600 mg) with a single dose of ayahuasca was feasible, generally well tolerated, and produced no evidence that CBD moderates ayahuasca's effects on social cognition, subjective measures of anxiety or standard safety parameters. Ayahuasca, with or without CBD, was associated with faster reaction times on emotion recognition and empathy tasks and with transient anxiolytic and expected psychedelic subjective effects. The authors conclude that the protocol appears safe for further research and suggest that future studies should use larger samples, explore different CBD doses, quantify CBD and endocannabinoid levels, and adopt specific scales to assess dissociative and psychedelic phenomena.