Examining the Therapeutic Effect of Ceremonial Ayahuasca on Narcissistic Personality and Antagonistic Externalizing in Adults

This was a prospective, observational study of adults attending ceremonies at three ayahuasca retreat centres in Peru and Costa Rica between December 2016 and June 2021. Invitations were sent to 1,301 reservation-holders and 355 participants enrolled. No study-specific exclusion criteria were applied; however, retreat intake staff required discontinuation of antidepressants within about 2 weeks and excluded candidates with a personal or first- or second-degree family history of psychotic disorders. Participants were aged 18 or older; standing on narcissistic traits was not an inclusion criterion. Final analysed samples varied by measure because of invalid responding and missing data (primary explicit-measures samples: NPI n = 257, PES n = 255; FFM sample n = 314; informant-report FFM n = 110). Retention across relevant timepoints was approximately 84%. Assessments were administered at baseline (pre-retreat), shortly after the retreat (post) and at 3-month follow-up. Explicit narcissism measures (NPI-13 adapted to a 5-point Likert scale, and the nine-item Psychological Entitlement Scale, PES) were collected at baseline and 3 months. FFM personality (IPIP-NEO-120 for self-report; IPIP-NEO-60 for informants) was collected at baseline, post and 3 months and used to derive a composite proxy for NPD by averaging facet scores empirically and clinically linked to NPD. Informant reports were obtained from close others identified by participants. Acute-experience measures included the Ego Dissolution Inventory (EDI) and the Revised Mystical Experience Questionnaire (MEQ); validity variables included measures of suggestibility (MISS), expectancy items regarding expected trait change, and items designed to flag overreporting or improbable experiences. Participants received modest incentives and were entered into a retreat raffle; informed consent and institutional review board approval were obtained. Ceremony contexts involved Shipibo-lineage shamans, with all participants taking part in ayahuasca ceremonies and sharing circles; some participants additionally had access to bufo (5-MeO-DMT), huachuma (mescaline), nunu, flower baths and kambo depending on centre and timing. The frequency and dose of these adjunctive substances varied and were recorded. Long-term adverse cognitive effects were queried at 3 months with an open response item. Analyses used linear mixed-effects models (LME) to test time effects (repeated measures nested within participants). Primary outcomes had a significance threshold of p < .05; secondary analyses were judged using p < .001 to balance Type I and Type II errors. Moderation tests added individual moderators separately as fixed covariates to base models. Effect sizes reported included Cohen's d and within-subject dz; unstandardised B coefficients indicated mean differences. Sensitivity power analyses indicated the self-report sample had 80% power to detect time effects exceeding about 0.12 standard deviations for primary explicit measures and 0.17 for the NPD composite; the informant sample was powered to detect larger effects (~0.27).

Sample characteristics: after exclusions for invalid responding and missing data, 257 participants provided NPI data, 255 provided PES data at baseline and 3 months, and 314 provided FFM data at baseline and at least one subsequent timepoint. Informant-report FFM data were available for 110 participants. The cohort was predominantly White (85%), with about 49% reporting income > $60K; about 40% reported a lifetime mental-health diagnosis (no participant reported a lifetime diagnosis of NPD). On average, baseline standing on explicit antagonism measures was not markedly elevated compared with some normative samples. Primary explicit outcomes: mixed results were observed. NPI Entitlement-Exploitativeness decreased modestly from baseline to 3 months (B = -0.10 on a 5-point scale, 95% CI [-0.18, -0.03], p = .020; dz = -0.15, ds = -0.08). NPI Leadership-Authority increased modestly (B = 0.11, 95% CI [0.03, 0.18], p = .004; dz = 0.18, ds = 0.09). No significant time effects were found for NPI Grandiose-Exhibitionism (p = .075) or the PES entitlement scale (p = .126). FFM-derived NPD composite: self-report FFM composites showed a statistically significant main effect of time, with lower NPD composite scores both within a week post-retreat (B = -0.07, 99.9% CI [-0.10, -0.04], p < .0001; dz = -0.39, ds = -0.19) and at 3 months (B = -0.05, 99.9% CI [-0.08, -0.02], p < .001; dz = -0.27, ds = -0.15). By contrast, informant-report NPD composites showed no significant change at 3 months (B = -0.01, 99.9% CI [-0.08, 0.06], p = .649). Self- and informant-reported NPD composites correlated moderately at both timepoints (r ≈ .39). Moderation and acute-experience analyses: no moderator (including suggestibility, expectancy, age, sex, education, use of additional psychedelics, dose, number of ceremonies, ego dissolution or mystical experience) reliably accounted for change in primary outcomes after correction thresholds. Some trend-level interactions appeared (for example, associations between use of bufo and PES change), but the authors did not interpret these given the low proportion of significant tests relative to Type I error. Mean acute-experience scores indicated generally moderate-to-strong mystical experiences (MEQ mean ~4.67 on a 6-point scale) and moderate ego dissolution (EDI mean ~3.51 on a 5-point scale); these did not moderate changes in antagonism or NPD. Adverse effects: among a subset of 188 participants asked about cognitive side effects at 3 months, 11 participants (6%) reported negative events they attributed to ceremony participation. Reported issues included difficulties relating to people, hypersensitivity, intrusive recollections/flashbacks, distressing dreams, auditory hallucinations, exacerbation of perceptual disturbances, speech impairment, brain fog, and concentration difficulties.

Weiss and colleagues interpret the findings as providing modest, qualified evidence that ceremonial ayahuasca use is associated with small reductions in some indicators of narcissistic antagonism and an FFM-based proxy of NPD, alongside small increases in agentic extraversion-related leadership authority. Changes were generally small in magnitude, inconsistent across measures, and not corroborated by informant reports; thus the investigators caution that the results do not demonstrate meaningful therapeutic change in narcissism. The study team situates these outcomes within prior literature showing psychedelic-associated reductions in antagonism/agreeableness deficits and increases in extraversion-related traits, noting that observed increases in Leadership-Authority are empirically linked to adaptive functioning but may also reflect ego inflation or spiritual superiority in some cases. Importantly, acute mystical or ego-dissolution experiences did not appear to account for changes in antagonism or the NPD proxy, suggesting that unitive-type states may not be the primary mechanism for change in the Antagonistic Externalizing spectrum according to these data. The authors discuss the discrepancy between self- and informant-report findings. Potential explanations include placebo or expectancy effects in self-report, limited agreement between participants and informants on the facets making up the NPD composite (observed correlations were r < .40), and possible limitations in informant observability for certain traits. They recommend greater use of informant-report and controlled designs in future work. Strengths highlighted include the prospective design, measurement of validity variables (suggestibility, expectancy), follow-up, informant data for some measures, and a relatively large sample with acceptable attrition. Principal limitations acknowledged are the absence of a placebo or control group, reliance on an observational retreat sample that was not clinically characterised for high antagonism or NPD, the use of an FFM-derived proxy rather than structured clinical NPD assessments, incomplete informant coverage for explicit narcissism measures, demographic homogeneity (primarily White, middle/upper-middle socioeconomic status), and heterogeneity in administered substances (some participants also received 5-MeO-DMT or mescaline). The authors stress the need for randomised controlled trials, assessment in clinical samples with elevated antagonistic externalizing, and therapy protocols tailored to antagonism/NPD, while weighing the small incidence of reported long-term adverse effects against any therapeutic benefit. Overall, the investigators consider their results sufficiently suggestive to warrant further controlled research but insufficient to claim substantial therapeutic effects of ceremonial ayahuasca on narcissistic pathology.

This first prospective investigation of psychedelic-associated change in narcissistic traits found small, mixed evidence of reduced narcissistic antagonism and lowered scores on an FFM-based NPD proxy, alongside modest increases in agentic extraversion. Effects were modest, inconsistently observed across measures, and not supported by informant reports. The authors conclude that these preliminary findings justify further investigation—particularly randomised controlled trials with participants who manifest clinical levels of antagonistic externalizing or NPD and antagonist-focused therapeutic approaches—before clinical utility can be established.