Effects of ayahuasca and its alkaloids on substance use disorders: an updated (2016-2020) systematic review of preclinical and human studies

Substance use disorders (SUDs) are common, chronic psychiatric conditions marked by relapse, compulsive seeking and consumption, impaired control, and negative emotional states such as irritability and dysphoria. Current pharmacological treatments (for example, nicotine replacement, methadone/buprenorphine, naltrexone/acamprosate/disulfiram) are only partially effective and there are no approved drugs for some substances such as cannabis and cocaine. Consequently, there is ongoing interest in alternative therapeutic approaches. Classic serotonergic psychedelics (including ayahuasca, DMT, psilocybin, and LSD) act primarily via cortical 5-HT2A receptor agonism and have been associated with increased neuroplasticity in regions involved in emotion, introspection, and social cognition; these neurobiological effects are hypothesised to underlie their therapeutic potential for SUDs and comorbid mood and anxiety disorders. This paper updates an earlier systematic review by assessing quantitative preclinical and human studies published between 2016 and 2020 that examined the effects of ayahuasca and its principal alkaloids on drug use (primary outcome) and related psychological and cognitive measures (secondary outcomes). Rodrigues and colleagues aimed to synthesise evidence from animal experiments and observational human studies to determine whether more recent data replicate earlier signals that ayahuasca or its constituents may reduce substance use and improve related symptoms, while also evaluating study quality and gaps that remain for clinical translation.

The investigators conducted a systematic review following PRISMA guidance. Electronic searches covered publications from 1 January 2016 to 8 November 2020 in PubMed, PsycINFO, SciELO, and LILACS. Search terms combined ayahuasca and its major alkaloids (dimethyltryptamine, harmine, tetrahydroharmine, harmaline) with dependence-related terms (dependence, addiction, substance use disorder) and animals for preclinical studies. Inclusion criteria encompassed randomized and open-label clinical studies using ayahuasca or its alkaloids for SUDs, observational studies of ayahuasca use in SUDs, quantitative studies with statistical comparisons (between groups or across time), use of standardised instruments for substance use or symptoms, preclinical SUD models employing ayahuasca or its alkaloids, publications in English, Spanish or Portuguese, and publication dates within the 2016–2020 window. Studies already included in the prior review were excluded. For data extraction, preclinical variables included year, design, animal species and sample characteristics, substances and routes of administration. Human-study variables included year and country, design, sample size and demographics, and ayahuasca dose/treatment details. Study quality was assessed with SYRCLES (adapted from the Cochrane RoB tool) for animal studies and NIH study-quality checklists for human studies. Items on the checklists were scored as present (1) or absent (0) and converted to overall percentages. The review process yielded nine eligible studies (four preclinical, five human observational), selected after screening titles/abstracts, full-text review and a hand-search of bibliographies.

Selection and overall characteristics: The search returned 259 records, 242 of which were excluded for reasons such as being preclinical but not SUD-related, not using ayahuasca or its alkaloids, or being reviews, qualitative reports, case reports, or letters. After full-text review and hand-searching, nine studies met inclusion criteria: four preclinical and five observational human studies. Human observational studies: Five human studies were heterogeneous in design and population. Lawn et al. reported a large international online survey (n = 96,901) with three groups: ayahuasca users (n = 527), classic-psychedelic users (LSD/psilocybin; n = 18,137), and non-users (n = 78,236). Ayahuasca users had higher personal well-being (PWI) than both comparator groups (p < 0.001), lower scores on problematic alcohol consumption than classic-psychedelic users (AUDIT, p = 0.002) but higher than non-users (p = 0.001), and a higher self-reported lifetime mental illness prevalence than non-users (24% vs 18.8%, p < 0.001). Barbosa and colleagues conducted cross-sectional analyses among União do Vegetal (UDV) members (n = 1,947) versus a general-population control sample (n = 7,939); UDV members reported higher lifetime alcohol and cannabis use but lower recent (last 30 days/12 months) prevalence of alcohol and tobacco disorders, and longer association with UDV (>3 years) was associated with less recent alcohol and tobacco use (p < 0.005). Two prospective observational cohort studies from the Takiwasi centre in Peru assessed patients admitted for 3–12-month treatment that included ayahuasca: Berlowitz et al. reported increases in quality of life (WHOQOL-BREF, p < 0.001) and readiness to change (RCQ), and Giovannetti et al. (male sample, n = 32, mean age 32) found significant reductions in depression (BDI, p < 0.001) and anxiety (BAI, p < 0.002) over a mean 6.5-month treatment, with greater decreases among those with opioid dependence (p < 0.05) and correlations between symptom reduction and treatment satisfaction and increases in quality-of-life and spirituality measures (p < 0.05). Barbosa et al. (another cross-sectional study) also reported UDV members scoring higher on some personality traits (agreeableness, openness) and on certain neuropsychological tests; no adverse effects were reported in that sample. Overall, the human studies reported mostly positive associations between ritual or clinical ayahuasca use and reduced recent substance use, lower anxiety/depression, improved quality of life and well-being, though some studies noted greater lifetime substance use and mental-health diagnoses among ayahuasca users. The authors rated human-study quality on average about 60% (medium to high) and identified common biases including sample size/power limitations, lack of blinding during analysis, single time-point data collection, absence of control groups in some cohorts, and confounding by therapeutic or religious contexts. Preclinical studies: Four animal experiments used rodent or other models and examined effects on self-administration, conditioned place preference (CPP), locomotion, anxiety, and neuronal activation markers (cFos/Fos). An amphetamine self-administration study in rats (four groups, n = 15 per group) tested ayahuasca 2 mL/kg by gavage and found that ayahuasca reduced preference for amphetamine-containing water (p < 0.05), attenuated amphetamine-induced hyperlocomotion, and reduced anxiety-like behaviour in the elevated plus-maze compared with amphetamine alone (p < 0.05). Cata-Preta et al. used a mouse alcohol-CPP paradigm and reported that ayahuasca at 100 mg/kg induced CPP by itself (p < 0.05) but, importantly, pretreatment with ayahuasca (100 and 300 mg/kg), Banisteriopsis caapi extract, and Psychotria viridis extract blocked or decreased alcohol-induced CPP (various p < 0.05–0.01). Nolli et al. evaluated voluntary alcohol consumption and cFos expression across five groups (including naltrexone and three ayahuasca doses); naltrexone reduced alcohol intake (p < 0.05) but none of the ayahuasca doses reduced alcohol consumption. Alcohol increased cFos in several cortical and accumbal regions; ayahuasca treatment altered cFos expression with complex dose-dependent patterns, and a low ayahuasca dose (Aya0.5) yielded lower cFos in medial orbital cortex versus control (p < 0.01). Reis et al. studied methylphenidate CPP in mice and found both methylphenidate (10 mg/kg) and ayahuasca (100 mg/kg) induced CPP, with a larger effect for methylphenidate; ayahuasca treatment reduced methylphenidate-induced reinstatement of CPP (p = 0.01) and normalised drug-induced Fos expression in several limbic and cortical regions (Cg1, PrL, OFC-LO, NAc shell/core; multiple p-values reported). The preclinical studies averaged about 39% on quality assessment (low to moderate) and the most frequent methodological shortcomings were lack of randomisation, absence of blinding, and inadequate allocation procedures. Adverse effects and heterogeneity: Human studies reported few adverse events in the extracted text, though Lawn et al. recorded stronger “negative effects while high” for ayahuasca compared with LSD and psilocybin without specifying which effects. The authors highlighted substantial heterogeneity across studies in designs, doses and frequency of ayahuasca use, and the inability of observational designs to infer causality.

Rodrigues and colleagues report largely concordant signals across preclinical and observational human studies: animal experiments showed reductions in amphetamine self-administration, attenuation or blockade of alcohol- and psychostimulant-related conditioned place preference, modulation of drug-induced cFos/Fos expression, and reduced anxiety-like behaviours; observational human studies indicated reductions in recent substance use, lower depression and anxiety symptoms, and improved quality of life and well-being among ritual users and patients treated in ayahuasca-centred programmes. Nevertheless, the authors emphasise important caveats. Preclinical evidence is limited by low-to-moderate methodological quality, variable findings (for example, ayahuasca induced CPP at some doses and failed to reduce alcohol consumption in one study), and uncertain translational validity from animals to humans. Observational human studies, although judged medium-to-high quality on checklist scores, are subject to selection bias, retrospective self-reporting, confounding by religious or therapeutic contexts, lack of control groups in some cohorts, and small samples or short follow-up periods; causality therefore cannot be established. Mechanistically, the authors discuss plausible biological and psychosocial pathways that could mediate therapeutic effects. Biologically, 5-HT2A receptor agonism and related neuroplastic changes in limbic and prefrontal circuits may underlie emotional regulation and cognitive flexibility improvements; harmine and other β-carbolines may affect dopaminergic transmission in nucleus accumbens, which could explain reductions in drug reinforcement and the observed modulation of cFos/Fos markers. Psychologically, intensity of subjective/mystical experiences, enhanced spirituality, changed priorities and improved emotional regulation have been associated with sustained reductions in substance use after serotonergic psychedelic sessions in prior studies. Social and cultural factors are also highlighted: participation in syncretic ayahuasca religions and communitarian support, prolonged therapeutic programmes that combine ayahuasca with other interventions, and increased social bonds may act as important mediators. The authors identify several limitations of the evidence base and of their review. Only nine studies met inclusion criteria; four were preclinical with frequent biases (selection bias, lack of randomisation, absence of blinding) and limited generalisability. The five human studies were observational and often embedded in complex therapeutic or religious settings, making it difficult to disentangle the specific contribution of ayahuasca. Additional limitations include small samples, variable dosing and frequency of ayahuasca, limited follow-up duration, and reliance on self-selected samples in surveys. Given these constraints, the authors conclude that more rigorous research is needed, specifically randomised, placebo-controlled clinical trials, to replicate and clarify the therapeutic potential and mechanisms of ayahuasca and its alkaloids for SUDs.

Preclinical and observational human studies published between 2016 and 2020 provide preliminary and mostly positive evidence that ayahuasca and its alkaloids may have therapeutic effects relevant to SUDs, including reductions in recent substance use, improvements in mood and anxiety symptoms, enhanced quality of life, and modulation of neural markers associated with reward. However, the evidence is limited by methodological weaknesses: low-to-moderate quality animal studies, observational human designs that cannot establish causality, small samples, heterogeneity in dosing and contexts of use, and potential bias from studies conducted within religious or multi‑modal therapeutic settings. No randomized clinical trials were identified in this update. Rodrigues and colleagues therefore recommend greater methodological rigour in future research, particularly randomized, placebo-controlled trials, to more definitively evaluate efficacy, optimal dosing, safety, and mechanisms of action.