Safety and Side Effects of Ayahuasca in Humans-An Overview Focusing on Developmental Toxicology

Ayahuasca is a traditional Amazonian brew combining DMT-containing plants with monoamine oxidase A (MAO-A) inhibitors from Banisteriopsis caapi (β-carbolines such as harmine, harmaline and tetrahydroharmine). These β-carbolines prevent gastrointestinal breakdown of DMT, allowing central nervous system exposure. Although the botanical, chemical and acute pharmacology of ayahuasca have been relatively well studied, the potential toxic effects of acute and long-term consumption—particularly in pregnant women, adolescents and their offspring—remain poorly characterised. This paper aims to provide an overview of the human toxicological literature on ayahuasca up to January 2013, and to discuss the relevance of preclinical developmental toxicology studies to human ritual contexts. The investigators set out to review controlled clinical trials of acute administration, observational studies of long-term ritual consumption, case reports and preclinical developmental studies, with particular attention to safety, adverse events and developmental outcomes.

A bibliographical review was conducted for human ayahuasca research and for preclinical developmental toxicology studies published up to January 2013. The reviewers prioritised double-blind, placebo-controlled, randomised clinical trials of acute ayahuasca administration when available, but also included uncontrolled human studies, cross-sectional investigations of long-term users, case reports and relevant preclinical animal studies. Searches were limited to articles in Portuguese, English and Spanish and used PubMed and the Brazilian SciELO database, supplemented by specialised books, book chapters and reference lists of identified papers. For the preclinical literature the same date limit and languages were applied. The methods emphasise a narrative synthesis of the clinical pharmacology, neurophysiology and reported adverse effects in humans, and a discussion of animal developmental toxicology findings in relation to human ritual use patterns.

Acute administration: Controlled clinical trials, notably by Jordi Riba's group in Barcelona, used encapsulated freeze-dried ayahuasca imported from Brazil at doses that delivered approximately 0.5–1 mg DMT/kg body weight in single-dose or two-dose crossover, placebo-controlled designs. Pharmacokinetic, neurophysiological, biochemical and neuroimaging measures were obtained. Subjective effects included dose-dependent perceptual, cognitive and affective changes. Plasma DMT time-courses matched subjective effect profiles, and urinary normetanephrine increased consistent with MAO inhibition. Standard blood tests showed no clinically relevant haematological or liver-function abnormalities in these trials. Neurophysiological and neuroimaging findings included dose-dependent increases in EEG beta power and reduced sensory gating (P50 suppression). LORETA analyses reported decreased power density over temporo-parieto-occipital junctions, temporomedial and frontomedial regions. SPECT showed increased perfusion bilaterally in anterior insula/inferior frontal gyrus (right greater than left), and in right anterior cingulate/frontomedial cortex, with smaller clusters in ventral anterior cingulate/subcallosal gyrus and increases in left amygdala/parahippocampal gyrus. Ayahuasca inhibited REM sleep but did not produce subjective deterioration of sleep quality. Moderate increases in systolic and diastolic blood pressure and heart rate were observed; across pilot and final studies two volunteers had systolic BP > 140 mm Hg, four had diastolic BP > 90 mm Hg and one had heart rate > 100 bpm. Prolactin, cortisol and growth hormone rose after dosing; autonomic measures (temperature, respiration, pupil size) changed moderately. Time-dependent changes in lymphocyte subsets were reported (percent CD4 and CD3 decreased, natural killer cells increased). The commonest adverse effects were nausea and vomiting. One participant experienced a transient intensely dysphoric reaction with disorientation and anxiety lasting ~20 minutes and withdrew voluntarily; other transient paranoid-type feelings have been reported but resolved without medical intervention. Other human experimental and imaging work without placebo controls (fMRI, EEG) in frequent users found activation of occipital, temporal and frontal regions associated with visual imagery and internal perceptual processes, increased gamma coherence in experienced users, and alterations in task-related activation during verbal fluency. Studies measuring acute effects in long-term users reported broadly similar subjective and neuroendocrine patterns to those seen in laboratory trials. Clinical populations and short-term follow-up: An exploratory, uncontrolled study of three women with recurring depressive disorder who received a single ayahuasca dose reported a significant decrease in Hamilton Depression Rating Scale scores beginning 40 minutes post-dose and sustained reductions from day 1 (around 79% reduction from baseline) to day 14 (around 66% below baseline), with scores rising toward baseline by day 14. These data are preliminary and uncontrolled. Long-term ritual consumption: Cross-sectional studies of adult long-term ritual users (members of Santo Daime, União do Vegetal, Barquinha and similar groups) found no clear evidence of physiological toxicity, and generally no consistent psychological, neuropsychological or psychiatric harm attributable to ritual ayahuasca use. Psychotic-like adverse reactions have been described in a small number of cases and appear to be rare according to the reviewed literature. Pregnancy, adolescents and developmental outcomes: Human data are sparse. The limited literature that evaluated adolescents exposed to ayahuasca, including in utero exposure reported by mothers, found no evidence of psychiatric, psychological or neuropsychological abnormalities attributable to exposure; the reviewers found no scientific publications reporting toxic effects in pregnant women or children born to women who drank ayahuasca. Preclinical developmental toxicology is more mixed: early animal studies (1960s–1970s) showed inconsistent results across species, with some reports of abortifacient effects or reproductive toxicity at particular doses. A study administering ayahuasca throughout gestation in rats reported maternal toxicity at the highest dose and dose-dependent visceral (dilated brain ventricles, renal pelvis) and skeletal (incomplete ossification) fetal findings, plus decreased fetal body weight at the highest dose. The reviewers emphasised that the high concentrations of specific alkaloids used in some rodent studies (particularly harmaline) and the frequency/duration of dosing are not representative of human ritual patterns. A later study by Oliveira et al. (2011) gave lower, more human-relevant doses to pregnant rats from gestational day 6 to lactation day 10 and found no maternal toxicity, no changes in physical or reflex development or body weight, no catalepsy or stereotypy, but an increased sensitivity to convulsant stimuli and altered cortical dopamine levels. That study also reported anxiolytic-like behaviour in offspring on the elevated plus-maze. Overall, the animal literature shows species-, dose- and schedule-dependent effects, limiting direct extrapolation to humans. Life-threatening and fatal reports: The review identified sensational media reports linking ayahuasca to coma and death, but noted absence of forensic confirmation in many cases and no published scientific case attributed conclusively to ayahuasca per se. The bibliographical search did not find scientific literature demonstrating life-threatening or lethal outcomes directly caused by ayahuasca. Potential interactions: The reviewers highlighted pharmacological interaction risks: combining MAO inhibitors in ayahuasca with serotonergic drugs (SSRIs), tryptophan, dextromethorphan, certain herbal products (St John's wort), amphetamines or MDMA could risk serotonin syndrome, a potentially serious condition of excessive serotonergic activity with autonomic, neuromuscular and cognitive symptoms. Harmine inhibits CYP2D6, raising the potential for drug–drug interactions. Foods high in tyramine combined with MAO inhibition could, in theory, precipitate hypertensive episodes.

Guimarães and colleagues interpret the assembled evidence as indicating that acute ayahuasca administration in controlled clinical settings appears to have acceptable tolerability in healthy volunteers, with commonly reported transient effects (perceptual changes, nausea, vomiting, moderate cardiovascular activation and occasional psychological distress) but no consistent biochemical or haematological toxicity. They position the Riba-led controlled trials and subsequent imaging and psychophysiological studies as key sources supporting relative short-term safety in adults. With respect to long-term ritual consumption, the authors state that cross-sectional studies of adult and adolescent ritual users provide no clear signal of physiological or psychiatric harm attributable to ayahuasca, though rare cases of persistent psychotic-like reactions have been reported. For pregnancy and developmental outcomes the reviewers emphasise the paucity of human data: the limited observational reports and interviews reviewed did not identify congenital abnormalities or clear developmental toxicity, but these data are sparse and not definitive. The authors caution against straightforward extrapolation from animal developmental-toxicology studies to humans. They note species differences in drug-metabolising enzymes (for example CYP2D6 in humans versus analogous but non-identical enzymes in rodents), receptor differences and the use of doses or dosing schedules in animal studies that do not reflect ritual human consumption. High-dose rodent findings (maternal toxicity, fetal visceral and skeletal changes) are interpreted as dose-dependent and not necessarily relevant to typical human use, whereas lower-dose animal work (Oliveira et al.) yielded fewer adverse findings but did note increased convulsant sensitivity and altered dopamine parameters. Key limitations acknowledged include the scarcity of controlled clinical data in clinical populations and pregnant women, reliance in places on uncontrolled or small-sample studies, the difficulty of reconciling heterogeneous animal protocols with human ritual practices, and the presence of only anecdotal or media-sourced reports for serious adverse events lacking forensic substantiation. The authors call for more basic and human research to resolve uncertainties, including additional preclinical studies using multiple species and dosing regimens more comparable to human ritual use, and carefully designed human studies to better define safety in pregnant women, adolescents and clinical populations. They also advise scientists to present findings with sensitivity to social consequences, warning against over-interpretation of animal data in ways that could increase prejudice against religious minorities who use ayahuasca.

In their final considerations the reviewers conclude that the available scientific literature up to January 2013 suggests that acute ayahuasca administration to healthy volunteers is relatively safe and that long-term ritual consumption by adults and adolescents does not appear to be seriously toxic. Human observational data regarding pregnancy and offspring are very limited but, in the reviewed reports, do not indicate clear toxic effects. The authors recommend further animal studies across species, doses and administration schedules that better model human ritual use and urge caution when extrapolating nonhuman preclinical data to the human context.