Effects of Ayahuasca on the Recognition of Facial Expressions of Emotions in Naive Healthy Volunteers: A Pilot, Proof-of-Concept, Randomized Controlled Trial

Recognition of emotions in facial expressions (REFE) is central to social cognition and has been used to probe the effects of serotonergic hallucinogens on emotional processing. Earlier experimental work with LSD and psilocybin reported mixed findings, with some studies showing reduced recognition of negative emotions (notably fear) and others showing no effect or variable effects on reaction time. To date, the effects of ayahuasca, a traditionally used brew containing N,N-dimethyltryptamine (DMT) and β-carbolines, on REFE had not been assessed in controlled trials. Mendes Rocha and colleagues designed a pilot, proof-of-concept, randomised, double-blind, placebo-controlled, parallel-group trial to test whether a single dose of ayahuasca modifies REFE in healthy, ayahuasca-naive volunteers. Secondary objectives included characterising subjective effects, tolerability (cardiovascular measures and adverse events), plasma brain-derived neurotrophic factor (BDNF) as a proxy for neuroplasticity, and the chemical stability of ayahuasca alkaloids over the 18-month study period.

The study used a randomised, double-blind, placebo-controlled, parallel-group design with simple randomisation performed by a researcher not involved in experimental sessions. Participants were recruited between November 2017 and May 2019; 49 people were screened and 22 were randomised (11 to ayahuasca, 11 to placebo). Two participants were excluded from analysis for protocol-related reasons (loss of baseline REFE data; hypoglycaemia during blood collection), leaving 20 participants (10 per group) in the final analysis. The mean age was 31.8 years (range 21–55). Eligibility required age 18–65, no prior ayahuasca use and ≤2 lifetime uses of other hallucinogens; exclusions included current or past psychiatric, cardiovascular, liver or neurological disease, use of psychotropic medications, recurrent drug use, and pregnancy or lactation. All participants were naive to hallucinogens and provided written informed consent. Ayahuasca and placebo were administered at 1 mL/kg in opaque brown bottles to limit organoleptic identification. The ayahuasca batch, prepared by a Santo Daime church in October 2017, contained DMT and β-carbolines (harmine, tetrahydroharmine [THH], harmaline); alkaloid concentrations were monitored by ultraperformance liquid chromatography–tandem mass spectrometry over 18 months. Placebo was prepared to mimic taste and smell (mineral water, glycerin, propylene glycol, methylparaben). Participants fasted and avoided certain foods and substances before sessions; sessions followed a nondirective, supportive approach without formal psychotherapeutic integration. Primary cognitive testing employed a dynamic, computerised REFE task derived from Ekman and Friesen depicting six basic emotions (happiness, sadness, fear, disgust, anger, surprise) with gradually increasing intensity from 0% to 100%. Participants labelled emotions as soon as recognised; outcomes were accuracy and reaction time. Subjective effects were recorded via spontaneous participant reports and observer impressions. Tolerability measures included systolic and diastolic blood pressure and heart rate at baseline and multiple post-dose time points (up to 240 minutes); adverse events were collected by spontaneous report. Plasma BDNF was sampled at baseline, 120 minutes and 240 minutes and analysed by ELISA. Blinding efficacy was assessed by asking participants and session researchers to guess treatment allocation at session end. Statistical analyses used two-way repeated-measures analysis of variance (ANOVA) with time and group as factors for quantitative outcomes (REFE variables, cardiovascular measures, BDNF). Pairwise comparisons used Student’s t tests with Bonferroni correction; significance was P < 0.05. Missing data were imputed with group medians. No formal sample size calculation was performed because this was a pilot study; sample size aimed to be comparable with prior psychedelic REFE studies (n = 16–24).

Recruitment and adherence were high: 22 participants were randomised, 21 completed the experimental session, and most attended follow-ups; final analyses included 20 participants (10 per group). No experimental ayahuasca session required interruption, and no serious adverse events occurred in the ayahuasca group; one serious event (hypoglycaemia with fainting) occurred in the placebo group and led to unblinding for that participant. Primary outcome—REFE: Compared with placebo, ayahuasca did not produce significant changes in recognition accuracy or reaction time. Repeated-measures ANOVA revealed significant main effects of time for overall accuracy (F6,84 = 5.49, P ≤ 0.001) and for reaction time (F2.71,84 = 7.59, P = 0.001), but there were no significant time × group interactions and no between-group differences. Emotion-specific analyses likewise showed significant effects of time for some emotions (for example, reaction time for happiness F6,84 = 6.84, P ≤ 0.001; accuracy for anger F4.7,84 = 2.760, P = 0.028; accuracy for disgust F6,84 = 3.4, P = 0.005), yet none of these effects differed by treatment group. No significant effects were seen for fear or surprise in accuracy or reaction time. The authors report increased performance across study visits, consistent with a learning effect. Subjective effects and blinding: In the ayahuasca group, 40% of volunteers reported visual effects during peak intoxication (90–180 minutes). Reported experiences included visual imagery (for example, "a white dove in a blue sky"), geometric forms, enhanced colour perception, increased creativity, improved sleep or reduced pain in one case, and feelings of tranquility, relaxation and well-being. Placebo participants typically reported mild tranquillity or other non-specific changes. Blinding success was limited: participants correctly guessed their treatment in 65% (13/20) of sessions and researchers in 75% (15/20). Tolerability and physiological measures: Adverse events in the ayahuasca group were mostly gastrointestinal and transient: vomiting (n = 3), nausea/gastrointestinal discomfort (n = 3), plus isolated reports of headache, drowsiness and concentration difficulty. No serious adverse events occurred in this group. Cardiovascular analyses showed a significant main effect of time for heart rate (F5,90 = 2.57, P = 0.03) but pairwise differences did not remain significant after Bonferroni correction; systolic and diastolic blood pressure showed no significant time or group effects. BDNF and alkaloid stability: Plasma BDNF did not differ between groups; analyses were constrained because five ayahuasca-group samples were missing due to participant refusal or technical issues. Alkaloid monitoring over 18 months showed progressive degradation, especially for DMT (average concentrations across time: DMT 1.58 mg/mL, range 0.45–3.04; harmine 1.15 mg/mL, range 0.30–1.57; THH 0.73 mg/mL, range 0.59–0.81; harmaline 7.38 mg/mL, range 2.07–11.80). Based on group mean weight (69 kg) and 1 mL/kg dosing, estimated mean ingested doses were 109.36 mg DMT (range 31.05–209.55), 79.62 mg harmine, 50.71 mg THH and 7.38 mg harmaline.

Mendes Rocha and colleagues report that their protocol was feasible and well tolerated in healthy ayahuasca-naive volunteers, but contrary to their hypothesis, a single ayahuasca dose did not alter performance on a dynamic REFE task compared with placebo. The researchers situate this null finding within a heterogeneous literature: prior studies with LSD and psilocybin have shown inconsistent effects on REFE accuracy and reaction time, with some evidence for reduced fear recognition or increased reaction times but no consensus. Several explanations for the null result are offered. First, the dose or pharmacological potency of the ayahuasca at the time of each session may have been insufficient because of significant alkaloid degradation over the 18-month storage period, particularly for DMT. Second, the study used a parallel-arm design rather than a crossover design, which may have reduced statistical power and increased the risk of type II error in the small sample. Third, practice or learning effects were evident: performance increased across repeated assessments, potentially producing ceiling effects that masked between-group differences. Fourth, participant characteristics may have limited sensitivity: the sample was highly educated and inexperienced with psychedelics, factors associated with higher baseline REFE performance and with challenges to blinding. Task features also mattered; the dynamic task allowed unlimited response time and full emotion intensities, which could make the task easier and less sensitive to acute drug-induced changes at subperceptual intensity levels. The authors further address secondary outcomes. Ayahuasca induced subjective psychedelic effects consistent with previous reports and was generally well tolerated, producing transient nausea, gastrointestinal discomfort and vomiting but no serious adverse events in this cohort. Cardiovascular measures showed no clinically significant changes. Plasma BDNF did not differ between groups; the investigators note that small sample size, missing samples, and methodological differences across studies (sample handling, centrifugation, assay timing) limit interpretation and comparison with prior reports that have found acute or delayed BDNF changes after psychedelic administration. Key limitations acknowledged include small sample size without formal power calculation, parallel-group design versus crossover alternatives, limited blinding efficacy, missing biological samples, and the observed alkaloid degradation. Practical implications recommended by the authors include storing natural ayahuasca under refrigerated conditions to preserve alkaloid stability, considering active placebos to enhance blinding, using larger samples or crossover designs to improve power, and standardising BDNF measurement methods. They conclude that further controlled trials—including repeated-dose regimens and studies in clinical populations—are required to clarify ayahuasca's effects on REFE and other aspects of social cognition.

In this pilot randomised, double-blind, placebo-controlled trial in healthy, ayahuasca-naive volunteers, a single dose of ayahuasca did not produce significant changes in recognition of facial expressions of emotions compared with placebo. No significant effects were observed on cardiovascular measures or plasma BDNF levels. The authors attribute the absence of between-group differences to possible factors including alkaloid degradation over the study, task learning and ceiling effects, high educational level of participants, dose considerations, and limited power due to sample size and parallel design. Ayahuasca produced expected subjective psychedelic effects and was generally well tolerated. The investigators recommend future trials with improved alkaloid storage, larger or crossover designs, active placebos to enhance blinding, and studies in clinical populations to better characterise effects on social cognition.