Ayahuasca-assisted meaning reconstruction therapy as an early resource for bereavement: a non-randomized clinical trial

Soto-Angona and colleagues frame bereavement as a normal but variably prolonged response to loss, noting that a subset of bereaved people develop persistent, functionally impairing grief within the first year after the death. Earlier research has suggested that ayahuasca can induce phenomenological experiences that correlate with reductions in depressive symptoms and may support emotional processing, meaning-making, and the redefinition of attachment to the deceased. The authors note that existing early interventions have not reliably prevented Prolonged Grief Disorder (PGD) and that antidepressant medication lacks evidence as a standalone treatment for PGD, highlighting a gap for effective early supportive therapies that might reduce both psychological and early physical risks after bereavement. This study set out to evaluate whether an ayahuasca-assisted therapy integrated with a meaning reconstruction framework (A-MR) reduces grief severity and PGD symptoms more than the meaning reconstruction therapy alone (MR) or a no-treatment control (NT) in people who had lost a first-degree relative within the previous 12 months. The investigators also examined secondary outcomes of post-traumatic growth and quality of life, and assessed safety and tolerability. They hypothesised that A-MR would produce larger reductions in grief and PGD symptoms and greater improvements in post-traumatic growth and quality of life that would persist to a three-month follow-up.

This was a three-arm, sequentially allocated, open-label clinical trial conducted in Spain. Eligible participants had experienced the death of a first-degree relative within the prior 12 months and met study screening criteria; further exclusion criteria are only partially visible in the extracted text but included psychiatric and substance-related exclusions and concurrent grief treatment. Participants were assigned in a predetermined sequence to one of three conditions until group targets were met: Meaning Reconstruction Therapy (MR), No Treatment (NT), and Ayahuasca-assisted Meaning Reconstruction Therapy (A-MR). No procedures to achieve baseline equivalence or randomisation were undertaken. Assessments occurred at baseline (T0), shortly after intervention (T3), and at 3-month follow-up (T4); subjective effects of ayahuasca in the A-MR group were additionally assessed the week after each administration (T1 and T2). The psychotherapeutic component, common to A-MR and MR, comprised nine weekly online 50-minute sessions delivered by four clinical psychologists, structured in three phases addressing the event story, the relationship back story, and the participant's personal story to support meaning reconstruction and identity reorganisation. A‑MR participants additionally received up to two group ayahuasca sessions conducted onsite after Phases 1 and 2, with individual online integration sessions in the subsequent week. Ayahuasca was prepared from Banisteriopsis caapi and Psychotria viridis and analysed by LC‑MS; the brew concentration and a medium dosing regimen of 0.75 mg DMT/kg administered in two intakes (0.30 mg DMT/kg then 0.45 mg DMT/kg after 30 minutes) are reported. Sessions used a standardised music playlist and participants remained under medical vigilance for 24 hours; blood pressure and heart rate were monitored pre‑ and 6 hours post‑administration. Primary outcomes were grief severity measured with the Texas Revised Inventory of Grief (TRIG) and PGD symptoms measured with the Traumatic Grief Inventory-Self Report (TGI-SR). Secondary outcomes included post-traumatic growth (PTGI‑SF) and the Psychological Health and Social Relationships domains of the WHOQOL‑BREF. Adverse events were screened at 24 hours using the PANSS‑EC and by participant self-report questions at post‑assessment; no standardised instrument for adverse events was used. Statistical analysis followed clinical-trial best practices for this exploratory study: a completer analysis was reported, missing data were tested for Missing Completely at Random and imputed by Multivariate Imputation by Chained Equations, and repeated measures ANCOVAs (RM‑ANCOVA) were used to test Treatment x Time interactions with months since death included as a covariate. Effect sizes (partial eta squared and Cohen's d) and Bonferroni-adjusted p values were reported.

Eighty-four participants were sequentially allocated equally to the three groups (n = 28 per group). Overall attrition was low (2.4% total); one participant in the A‑MR group dropped out after the first ayahuasca session and one in the NT group declined further participation before post-intervention assessments. Dropouts had higher baseline grief scores than completers (reported TRIG and TGI‑SR values and significant t-tests), and 55 participants completed the nine-session psychotherapy component (A‑MR: n = 27; MR: n = 28). In the A‑MR arm, 27 participants completed the first ayahuasca session and 24 completed the second; no differences in primary or secondary outcomes were found between participants who attended one versus two ayahuasca sessions. Missing data were few (n = 15 values, 0.40%) and Little's MCAR test supported data being missing completely at random (p = .328). Safety: Ayahuasca was generally well tolerated with no serious adverse events reported. Common acute physical effects during sessions included nausea, vomiting or diarrhoea (n = 16), headache (n = 4), fainting (n = 2), and anguish (n = 2). Eight participants reported experiencing a negative effect during the ayahuasca session and three were uncertain. Only three participants reported adverse events in the week following a session (headache, distraction, minor sickness). Average PANSS‑EC score at 24 hours post‑administration was 5.07 (SE 0.267). Two participants required ibuprofen 8 hours post‑administration for persistent headache; no participants required psychiatric medication in the 24‑hour period following administration. Primary outcomes: Repeated measures analyses showed a significant Intervention x Time interaction for both TRIG (F(3.57,139.13) = 22.70; p < .0001; partial eta squared = 0.37) and TGI‑SR (F(3.57,139.37) = 24.88; p < .0001; partial eta squared = 0.39). Months since death did not interact significantly with treatment effects. There were no baseline differences among groups. At post‑intervention and at three‑month follow‑up, the A‑MR group showed greater reductions in normal grief severity than both NT and MR: versus NT at T3 p < .0001 (Cohen's d = 1.50) and at T4 p < .0001 (d = 1.25); versus MR at T3 p = .0001 (d = 1.20) and at T4 p = .0008 (d = 1.05). No significant differences were found between MR and NT at post‑intervention or follow‑up. Secondary outcomes: For post‑traumatic growth (PTGI‑SF) there was a Treatment x Time interaction (F(3.55,138.26) = 9.51; p < .0001; partial eta squared = 0.20). A‑MR participants had higher PTGI‑SF scores than MR at post‑intervention (T3: p = .023; d = 0.80), and higher scores than NT at post‑intervention (T3: p = .029; d = 0.67) and follow‑up (T4: p = .014; d = 0.77). Within‑group comparisons showed large pre‑to‑post increases in PTGI‑SF for A‑MR (T0 vs T3: p < .0001; d = 1.30; T0 vs T4: p < .0001; d = 1.78). The MR group showed a smaller increase at post‑intervention only (T0 vs T3: p = .016; d = 0.57), while the NT group showed no significant improvement. Quality of life: The Psychological Health and Social Relationships domains of the WHOQOL‑BREF showed significant Intervention x Time interactions (Psychological Health: F(4,156) = 7.01; p < .0001; ηp2 = 0.15; Social Relationships: F(3.47,135.2) = 9.92; p < .0001; ηp2 = 0.20). No baseline differences were observed. The A‑MR group differed from the MR group on the Psychological Health domain at post‑intervention (T3: p = .027; d = 0.66). Between‑group comparisons for the Social Relationships domain at post‑treatment did not show significant differences (T3 and T4: p ≥ .794). Overall, improvements in psychological quality of life were larger in A‑MR than in control groups, and MR showed some improvements over time but generally smaller in magnitude. The covariate months since death did not significantly moderate outcomes.

Soto-Angona and colleagues interpret their findings as preliminary support for the safety and potential efficacy of integrating ayahuasca sessions within a meaning reconstruction therapeutic protocol for people with severe grief within the first year after losing a first‑degree relative. They report that the A‑MR arm showed larger reductions in grief severity and PGD symptomatology and greater increases in post‑traumatic growth than MR or NT, with effects sustained to three months. The authors suggest that ayahuasca may act as a catalyst for meaning‑making and the emotionally salient retrieval of autobiographical memories, which can then be consolidated through integration and restorative narrative retelling, producing a synergistic effect with structured psychotherapy. The investigators situate their results relative to prior observational and small intervention studies, noting that the magnitude of effect observed for grief severity in A‑MR exceeds that reported for other early interventions for PGD. They nonetheless acknowledge important limitations that constrain internal and external validity. Chief among these is the lack of randomisation and the sequential allocation driven by COVID‑19 restrictions, which raises the risk of selection bias, temporal confounding, and procedural differences across groups. The open‑label design admits expectancy and reporting biases given reliance on self‑report measures. External validity is limited by participant self‑selection and prior psychedelic experience—about three‑quarters of the sample had previously used psychedelics—which may have amplified expectancy effects and limits generalisability. Additional limitations highlighted include the absence of standardised instruments to capture adverse effects comprehensively and the lack of an independent rater for attributing adverse events, preventing precise characterisation of risk. The study design did not permit isolation of pharmacological effects from non‑pharmacological components because ayahuasca was always delivered with preparatory and integration psychotherapy and in a group, natural‑setting context. Finally, the short three‑month follow‑up limits conclusions about long‑term durability and relapse risk. The authors recommend replication in larger, randomised trials, use of standardised adverse‑event assessments, and ethical engagement with Indigenous and ayahuasca‑using communities to ensure responsible application in Western settings.

In their concluding remarks, the authors state that ayahuasca‑assisted meaning reconstruction therapy may catalyse psychotherapeutic work on meaning reconstruction and improve adaptation to loss, producing substantial reductions in grief and PGD symptoms and increases in post‑traumatic growth that persisted to three months. They emphasise the need for further research—particularly larger randomised trials—and stress the importance of consensual cooperation with Indigenous peoples and ayahuasca religious communities to ensure ethical and responsible use.