Ceremonial Ayahuasca in Amazonian Retreats-Mental Health and Epigenetic Outcomes From a Six-Month Naturalistic Study

This was a prospective, observational naturalistic study conducted at the Ayahuasca Foundation, a retreat and research centre near Iquitos, Peru. Participants were self-selected individuals who had already enrolled on commercial retreats; the research team informed prospective attendees before and on arrival and obtained informed consent. The institutional ethics committee approved the protocol and a medically qualified doctor was present during retreats. Inclusion/exclusion screening was carried out by the retreat centre prior to acceptance; individuals with known psychotic disorders, schizophrenia, bipolar affective disorder and personality disorders were excluded according to the centre's criteria. Participants observed a 2-week "washout" period prior to arrival with guidance to avoid substances and certain foods to reduce tyramine-related risks from ayahuasca's MAOI components. Retreats varied in duration (8, 14, 21 or 28 days) and included multiple evening ceremonies (typically four to 11 offered depending on retreat length). Attendance at every ceremony was optional and the researchers recorded the number of ceremonies each person attended. Dosing was traditional and non-standardised; on average participants consumed about 150 ml of brew per ceremony, administered by the curandero in a Shipibo-style maloka with facilitators present. Psychometric assessments were completed at three time points: pre-retreat (the night before travel into the jungle), post-retreat (the morning after the last ceremony) and at 6-month follow-up via electronic survey. Primary instruments were the Beck Depression Inventory-II (BDI-II), the Trait subscale of the State-Trait Anxiety Inventory (STAI-T), the Self-Compassion Scale (SCS) and the Clinical Outcomes in Routine Evaluation-Outcome Measure (CORE-OM). Secondary measures included the Childhood Trauma Questionnaire (CTQ; pre-retreat only), the Mystical Experience Questionnaire (MEQ; post-retreat only), and a sentence completion task for autobiographical memory (SCEPT) coded for specificity and valence. Inter-rater reliability for SCEPT coding was checked on 25% of responses. For epigenetic analyses, saliva was collected pre- and post-retreat (2 × 4 ml). DNA extraction and bisulfite conversion were performed; 55 paired samples were obtained initially but 48 paired samples were used for pyrosequencing owing to laboratory constraints. Pyrosequencing assays targeted five CpG sites in the SIGMAR1 promoter and a re-designed FKBP5 assay; the BDNF assay failed. Statistical analyses used SPSS and R, with repeated-measures ANOVA (Greenhouse-Geisser correction) and Bonferroni-corrected post-hoc pairwise comparisons for longitudinal psychometric data. Correlations (Pearson's r) examined relationships between number of ceremonies, retreat length, prior ayahuasca use, CTQ/MEQ scores and change in outcomes; alpha adjustment for multiple comparisons was applied where stated.

The sample comprised 63 participants; the extracted text reports 35 males and 25 females (these counts sum to 60, so the gender breakdown in the extraction is unclear). Ages ranged from 19 to 63 years (mean 37.0, SD 9.7). Most participants were White (reported 79.4%). Retreat lengths were distributed across 8-day (n = 18), 14-day (n = 12), 21-day (n = 12) and 28-day (n = 19) programmes. Prior ayahuasca experience varied (37 participants were naïve; others reported 1–80 prior uses, mean 5.9). Psychiatric history included 15 reporting depression and 15 anxiety disorders; 27 reported prior problematic substance use. Psychometric outcomes showed statistically significant improvements from pre- to post-retreat. Reported group-level changes (from extracted text) included reductions in BDI-II (example figures reported in the abstract: 13.9 to 6.1), STAI (example figures: 44.4 to 34.3), and CORE-OM (example figures: 37.3 to 22.3), all with p < 0.001 at post-retreat. Self-compassion (SCS) increased (example figures: 3.1 to 3.6, p < 0.001). Improvements were largely sustained at 6 months: follow-up scores remained significantly improved relative to pre-retreat (p < 0.001) but did not differ significantly from immediate post-retreat scores (e.g., BDI-II p = 0.153; STAI-T p = 1.0; CORE-OM p = 1.0), consistent with maintenance of gains. Focusing on the subsample meeting BDI-II depression cut-offs at baseline (n = 31: 11 mild, 11 moderate, nine severe), the mean pre-retreat BDI-II was 24.2 (median 23.0). At post-retreat 24 of these 31 (77.4%) no longer met the depression threshold (post-retreat mean 8.7, median 5.0). At 6 months the majority remained below threshold (follow-up mean 5.2, median 4.0; four missing). Change in BDI-II in this depressed subsample was highly significant across time, F(2,25) = 55.5, p < 0.001. Measures of autobiographical memory specificity (SCEPT) did not show changes in specificity scores. However, analyses of memory valence indicated reductions in negatively valenced memories over time: the mean SCEPT negative-valence scores differed significantly across time points, F(2,110) = 5.68, p < 0.005, with a significant reduction from pre-retreat to 6-month follow-up (p = 0.004). Analyses exploring contextual or dosing variables found no significant correlations between change in outcomes and number of ceremonies, retreat length or prior ayahuasca frequency (alpha set at 0.01 due to multiple comparisons). Correlational analyses identified relationships between childhood trauma and outcome change: higher CTQ total scores correlated with greater BDI-II change post-retreat in the overall sample (r = 0.318, p = 0.011) and in the depressed subsample (r = 0.393, p = 0.029). In the depressed subgroup, greater mystical experience (MEQ mystical subscale) was associated with larger immediate post-retreat BDI-II improvements (r = -0.357, p = 0.049); this correlation was not sustained at 6 months and was not present in the full sample. Epigenetic results: BDNF assays failed and were excluded. SIGMAR1 methylation increased significantly across the five assayed CpG sites (paired t-test: t = 2.58, df = 38, p = 0.01); the reported mean change was modest (authors note a 2.1% increase). FKBP5 methylation showed no statistically significant change (p = 0.13). Changes in SIGMAR1 methylation correlated with CTQ total scores (r = 0.387, p = 0.015), indicating larger methylation changes in those reporting higher childhood trauma, but SIGMAR1 methylation change did not correlate significantly with BDI-II change.

Ruffell and colleagues interpret their findings as evidence that participation in ceremonial ayahuasca retreats was associated with reduced depressive and anxiety symptoms and lower global distress immediately after retreat, with improvements generally maintained at 6 months. The investigators highlight that most participants who met BDI-II criteria for depression prior to retreat no longer met that threshold after the retreat, although they note many in that subsample presented with mild depression. The absence of changes in autobiographical memory specificity, coupled with reductions in negatively valenced memories, is discussed as a possible cognitive correlate of improved mood rather than a change in overgenerality per se. The authors also report an association between higher reported mystical experience and greater immediate improvement in depressive symptoms within the depressed subgroup, consistent with previous psychedelic research linking subjective mystical-type effects to therapeutic gains. On epigenetics, the team observed a small but statistically significant increase in SIGMAR1 promoter methylation post-retreat and a positive correlation between childhood trauma scores and SIGMAR1 methylation change. They caution that the mean methylation increase (reported as about 2.1%) is modest and its biological significance is unclear; common methylation rules typically link hypermethylation to gene silencing, so the functional consequences remain speculative. Nevertheless, the authors consider SIGMAR1 a candidate mechanism warranting further study given prior preclinical links to stress response, neuroplasticity and PTSD-related biology. Several limitations acknowledged by the investigators temper causal interpretation. The study lacked a control group and was subject to self-selection and expectancy bias because participants had chosen and paid for retreats; contextual factors inherent to the retreat environment (group setting, removal from daily life, ritual elements, music, facilitator support) could have contributed substantially to outcomes. Dosing was non-standardised, medical histories could not be independently verified, and some participants were not naïve to ayahuasca. For the epigenetic work, saliva-derived peripheral DNA may not reflect central nervous system changes, the analysis was candidate-gene in design (raising bias concerns) and BDNF data were lost. The authors therefore call for replication using randomised controlled designs, larger and clinical samples, broader biological marker panels and neuroimaging, as well as qualitative work to characterise the nature of transformative experiences. Finally, the researchers note ethical and safety issues around "ayahuasca tourism," including lack of regulation, risks from inadequately trained facilitators, and cultural appropriation concerns. They recommend cautious scientific and public engagement given these considerations.

The study reports that ceremonial ayahuasca use in a traditional Amazonian retreat context was associated with statistically significant improvements in depression, trait anxiety, self-compassion and global distress that were sustained at 6-month follow-up. Epigenetic analysis revealed a small increase in SIGMAR1 promoter methylation post-retreat, with greater changes correlated with higher childhood trauma scores, though the functional significance is uncertain. The authors conclude these findings support further investigation of ayahuasca's potential therapeutic effects and of SIGMAR1 as a candidate mechanistic target, ideally in randomised controlled trials with broader biological and neuroimaging assessments.