The current state of research on ayahuasca: A systematic review of human studies assessing psychiatric symptoms, neuropsychological functioning, and neuroimaging

Ayahuasca (AYA) is a botanical hallucinogen traditionally used in Amazonian ritual contexts and, more recently, as a sacrament within syncretic Brazilian religions such as Santo Daime and União do Vegetal (UDV). The brew combines Banisteriopsis caapi, which contains reversible monoamine oxidase A (MAO-A) inhibitors (β-carbolines such as harmine and tetrahydroharmine), with Psychotria viridis, a source of the serotonergic agonist dimethyltryptamine (DMT). When taken together, the β-carbolines allow orally ingested DMT to reach the central nervous system and produce psychoactive effects that typically peak around 1.5–2 hours and last roughly 4–6 hours. Previous human studies and reviews suggested AYA has moderate, short-lived physiological effects and that ritual consumption is not generally associated with psychopathology or cognitive deficits, but many earlier studies were limited by small samples, a focus on experienced ritual participants, or a lack of neuroimaging data. Dos Santos and colleagues set out to update and extend prior reviews by performing a systematic review of human studies published up to 16 December 2015 that assessed acute, subacute, and long-term effects of ayahuasca on psychiatric symptoms, neuropsychological functioning, and neuroimaging. The objective was to synthesise evidence on safety, tolerability, cognitive effects, neural correlates and preliminary therapeutic signals (for example antidepressive or antiaddictive effects), while noting methodological limitations that constrain causal inference and generalisability.

The review was conducted following PRISMA guidelines. Electronic searches were performed in PubMed, LILACS and SciELO up to 16 December 2015 using terms combining "ayahuasca" with keywords covering subjective/psychological effects, psychiatry, cognition, neuropsychology, neuroimaging and related modalities (for example SPECT and fMRI). The authors also checked reference lists of identified papers. Only studies published in English were considered. Pre-specified inclusion criteria limited the review to human experimental studies in healthy volunteers, observational studies of experienced AYA consumers, and clinical trials in patients diagnosed using structured interviews (DSM criteria). Excluded materials comprised animal studies, reviews, qualitative reports, case reports, opinion pieces, letters, conference abstracts, books and chapters. Eligible study designs included experimental AYA administration assessing psychiatric symptoms, neuropsychological performance or neuroimaging with validated instruments; observational studies comparing experienced users with controls or normative data; and clinical trials. Comparators, when available, were placebo (in experimental and clinical trials) or control groups/normative data in observational work. Study selection and data extraction were performed by two independent reviewers with adjudication by a third where necessary. Extracted information included authors, year, design, participant characteristics (healthy volunteers, experienced users, clinical patients and sample sizes), outcome measures (psychometric scales, neuropsychological tests, neuroimaging parameters) and timing of assessment. To aid synthesis the sample of studies was organised by outcome domain (psychiatric symptoms, neuropsychological functioning, neuroimaging) and by time-frame (acute, subacute, long-term). The search process initially returned 68 references for abstract screening, 27 were taken forward for full-text review, and a final 28 citations met the inclusion criteria. No pooled meta-analytic statistics were reported; the heterogeneity of designs, small sample sizes and variable control conditions led to a narrative synthesis of findings.

The review included 28 human studies, with most addressing subjective/psychological effects or psychiatric symptoms (25 studies), five evaluating neuropsychological functioning and five employing neuroimaging techniques (some studies contributed to multiple domains). Psychiatric symptoms and subjective effects: Across experimental studies in healthy volunteers, acute AYA produced dose-dependent subjective effects measured by scales such as the Hallucinogen Rating Scale (HRS) and the Addiction Research Center Inventory (ARCI). Reported acute effects included perceptual alterations, enhanced introspection, intense emotional changes and increased positive mood and activation; peak effects occurred around 1.5–2 hours and resolved within about 4–6 hours. In small controlled and crossover trials (for example encapsulated freeze-dried AYA doses up to 1 mg DMT/kg), most HRS subscales increased significantly except for volition in some reports. Open-label clinical studies in patients with depression found rapid and sustained antidepressant and anxiolytic effects after a single oral AYA dose: a pilot study (n=6) reported up to 82% reductions in depressive scores on HAM-D and MADRS at 1, 7 and 21 days, and a subsequent larger open-label sample (n=17, including the pilot participants) found significant score reductions from 80 minutes through day 21. Vomiting was the most common adverse effect in these patient trials (reported by 47–50% of participants). Subacute effects observed in naturalistic ritual contexts included short-term reductions in minor psychiatric symptoms after first-time use (assessed with CIS-R) and increases in measures of visual creativity and mindfulness-related capacities (Five Facet Mindfulness Questionnaire, Experiences Questionnaire), with some post-session scores comparable to those seen after extensive mindfulness practice. These findings, however, derive from non-randomised, often uncontrolled designs. Long-term effects from observational studies of ritual users generally showed no increase in psychopathology or cognitive deficits. Studies of UDV and Santo Daime members reported no current psychiatric diagnoses in many ritual users and, in several cases, remission of prior disorders after church participation. Personality and attitude changes associated with regular use included lower Novelty Seeking and Harm Avoidance, higher Self-Transcendence, greater spirituality and psychosocial well-being, and reduced impulsivity; some of these differences persisted at one-year follow-up. Adolescent UDV members (n=40) were comparable to matched controls on multiple mental health measures. Neuropsychological functioning: Under acute AYA influence, studies reported impaired working memory (for example lower performance on the Sternberg task) but improved inhibitory control (reduced response latency on the Stroop test without loss of accuracy). Occasional users sometimes showed worse planning performance (Tower of London) relative to experienced users, with planning impairment inversely correlated with lifetime AYA exposure. Long-term neuropsychological assessments generally found no evidence of cognitive deterioration; some studies reported superior performance among regular AYA users on executive tasks (Wisconsin Card Sorting Test, two-back working memory, task-switching) and these differences were largely maintained at one-year follow-up. Neuroimaging: A randomized double-blind SPECT study in 15 volunteers given encapsulated freeze-dried AYA (1.0 mg DMT/kg) versus placebo found bilateral activation of the anterior insula/inferior frontal gyrus (more pronounced on the right) and increased perfusion in right anterior cingulate/medial frontal gyrus; left amygdala/parahippocampal activation was also reported. An fMRI imagery task in nine frequent users given a single oral dose (2.2 mL/kg, 0.8 mg DMT/mL) showed increased activation across occipital, parietal, temporal and frontal regions during imagery, with primary visual cortex activation comparable to viewing natural images and altered fronto-occipital effective connectivity. A separate fMRI study of ten experienced users reported decreased activity and reduced functional connectivity within key default mode network (DMN) hubs, including posterior cingulate/precuneus and medial prefrontal cortex. In a SPECT study of 17 patients with recurrent depression, AYA administration was associated with increased perfusion in the left nucleus accumbens, right insula and left subgenual area. Structural MRI in 22 Santo Daime members versus 22 matched controls identified cortical thinning in mesotemporal and posterior cingulate regions and thickening in anterior cingulate and precentral gyri; posterior cingulate thickness was inversely correlated with age of onset, intensity of prior AYA use and scores on the Self-Transcendence temperament scale. No associations were found between these morphometric differences and psychopathology scales. Adverse events: Across controlled and observational studies, AYA was generally well tolerated. Short-lived side effects included transient anxiety, nausea and vomiting; no persistent side-effects were consistently reported. Reported severe psychiatric adverse events were rare. In the UDV monitoring data, 13–24 psychotic incidents were reported over an estimated 25,000 sessions (a rate reported as less than 0.1%, specifically 0.052–0.096%), but case reviews often identified pre-morbid risk factors or substance co-use. Isolated case reports described psychotic or manic episodes with potential confounders such as heavy cannabis or prior bipolar history. The review also flagged the lack of human data on pregnant women and children, and cautioned about potential pharmacological interactions between AYA (β-carbolines) and monoaminergic drugs that could theoretically precipitate serotonin syndrome.

Dos Santos and colleagues interpret the assembled evidence as indicating that acute AYA administration in controlled settings produces subjective effects similar to other classic 5-HT2A agonist hallucinogens—notably perceptual alterations, enhanced introspection, salient emotional changes and improved positive mood—with peak effects at 1.5–2 hours and resolution several hours later. The authors highlight preliminary clinical signals of therapeutic utility: open-label and observational data suggest rapid and sustained antidepressive effects and reductions in substance use in some samples, findings that are consonant with emerging clinical data for other psychedelics such as psilocybin and LSD. Nevertheless, they emphasise that the clinical trials reviewed were small, uncontrolled or unblinded, limiting causal inference. Neurobiological hypotheses discussed include 5-HT2A receptor agonism in frontal and paralimbic circuits involved in emotion, interoception, memory and self-processing, reductions in DMN activity (potentially related to decreased rumination), and acute increases in perfusion in mood-regulatory regions. The authors argue that AYA’s β-carbolines and harmine may contribute to neuroplastic effects via glutamatergic mechanisms and upregulation of transcriptional and neurotrophic factors (for example BDNF), citing animal data where harmine increased hippocampal BDNF and produced antidepressant-like behaviours. Structural MRI findings of anterior cingulate thickening and posterior cingulate thinning are considered potentially compatible with neuroplastic adaptations, and may relate to observed changes in attention/executive control and self-transcendence, although causality is not established. Key limitations acknowledged by the authors include small sample sizes, high heterogeneity across studies, frequent absence of placebo or control groups, predominance of experienced ritual users in observational work (raising selection bias), and lack of randomisation and blinding in clinical trials. They also note the scarcity of data on vulnerable populations (pregnant women, children) and on pharmacological interactions with MAO inhibitors or serotonergic medications. As a result, the authors call for more rigorous controlled trials, including studies in naïve users and clinical populations, and systematic investigation of safety, mechanisms and long-term consequences before therapeutic claims can be robustly supported.

The authors conclude that—based on the available human studies up to December 2015—acute, subacute and long-term controlled or ritualistic use of ayahuasca appears to have a favourable safety profile in the populations studied, with nausea and vomiting the most common adverse effects and rare serious psychiatric reactions that are often confounded by pre-morbid factors or concomitant substance use. Neuropsychological data indicate transient working memory impairments during acute intoxication alongside improvements in some executive functions (planning, inhibitory control) in experienced users. Neuroimaging studies point to activation of frontal and paralimbic regions, reductions in default mode network activity, and structural differences such as anterior cingulate thickening and posterior cingulate thinning. Preliminary evidence suggests potential antidepressive and antiaddictive effects, but these findings stem from small, uncontrolled studies and require replication in randomised, placebo-controlled trials and broader populations (including naïve users, clinical samples, pregnant women and children) to clarify efficacy, mechanisms and long-term risks.