Prospective examination of the therapeutic role of psychological flexibility and cognitive reappraisal in the ceremonial use of ayahuasca

Abstract

Background:

Evidence suggests that psychedelic-assisted therapy carries transdiagnostic efficacy in the treatment of mental health conditions characterized by low mood and the use of avoidance coping strategies.

Aims:

While preliminary evidence suggests that psychological flexibility and emotion regulation processes play an important role within psychedelic therapy, this prospective study addressed methodological gaps in the literature and examined the ability of ayahuasca to stimulate acute states of cognitive reappraisal and long-term changes in psychological flexibility and mood. The study also explored whether moderating factors predisposed participants to experience therapeutic changes.

Methods:

Participants (N = 261) were recruited from three Shipibo ayahuasca retreat centers in Central and South America and completed assessments on mood, psychological flexibility, and acute ceremonial factors. Expectancy, demand characteristics, and invalid responding were controlled for with several validity scales.

Results/Outcomes:

Participants reported significant reductions in negative mood after three months, as well as increases in positive mood and psychological flexibility. Acute experiences of reappraisal during the ayahuasca ceremony exerted the strongest moderating effects on increases in positive mood and psychological flexibility. Increases in psychological flexibility statistically mediated the effects of acute psychological factors, including reappraisal, on changes in positive mood.

Conclusions/Interpretation:

These results highlight the role of acute psychological processes, such as reappraisal, and post-acute increases in psychological flexibility as putative mechanisms underlying positive outcomes associated with psychedelics. These results also provide support for the integration of third-wave and mindfulness-based therapy approaches with psychedelic-assisted interventions.

Introduction

Mood disorders exert a substantial and devastating disease burden globally (Simon, 2003). Approximately 10% of the global population meets diagnostic criteria for a mood disorder at some point in their lifetime, and major depression comprises the leading cause of years lived with disability (Friedrich, 2017). Evidence suggests that individuals with mood disorders such as major depression present with key psychological characteristics, including a tendency to react negatively to their emotions and to use maladaptive emotion regulation strategies. Over time, these coping strategies appear to amplify psychological distress and increase the frequency and intensity of negative emotions (Aldao et al., 2010; Morris and Mansell, 2018). The frequent use of emotion regulation strategies characterized by emotional avoidance (e.g. suppression or rumination) is considered an important etiological and maintenance factor across mood and related psychiatric conditions (e.g. anxiety and substance use disorders; Bradley et al., 2011; Vujanovic et al., 2021).

Several empirically supported interventions have been developed to target these maladaptive emotion regulation strategies including “third-wave” contextual and mindfulness-based therapies (e.g. Acceptance and Commitment Therapy; ACT) as well as non-directive therapies involving classic psychedelics (e.g. psilocybin and NN-dimethyltryptamine (DMT)–containing ayahuasca). These interventions have shown particular promise in enhancing outcomes with moderate-to-large effect sizes compared to control conditions (Luoma et al., 2020; Öst, 2008). It has been proposed that they exert their effects through the stimulation of psychological flexibility processes that decrease emotional avoidance and increase approach-oriented emotional coping (Davis et al., 2020; Kashdan et al., 2020; Zeifman et al., 2020), which has rendered psychological flexibility a key target for study. Psychological flexibility is defined as the ability to flexibly alternate between and manage challenging thoughts, emotions, or mental sets in service of carrying out valued actions (Hayes et al., 2011).

Given their putative overlapping mechanisms, there has been recent interest in the integration of classic psychedelics with contextual-based therapies, focused on psychological flexibility processes, to treat disorders characterized by emotional avoidance and maladaptive emotion regulation (Sloshower et al., 2020; Watts and Luoma, 2020; Wolff et al., 2020). A few recent studies, including three prospective studies, have suggested that psychological flexibility is a key component in psychedelic-induced outcomes (Close et al., 2020; Davis et al., 2020; Zeifman et al., 2020). In particular, across three prospective samples, psychological flexibility showed signs of increasing and covarying over time with clinical outcomes including depression and suicidality (Close et al., 2020; Zeifman et al., 2020). In a recent cross-sectional study, increases in psychological flexibility were found to statistically mediate the relation between acute insight, mystical experience, and decreases in depression and anxiety (Davis et al., 2020). Another study found that emotion regulation ability cross-sectionally mediated the relation between psychedelic use and lower intimate partner violence (Thiessen et al., 2018). It is furthermore notable that following psychedelic use, acute experiences of emotional engagement and breakthrough, which may themselves accompany states of psychological flexibility, have shown signs of promoting well-being and decreasing depression (Close et al., 2020; Roseman et al., 2019; Thiessen et al., 2018). These findings provide meaningful preliminary support for the notion that psychological flexibility processes may be involved in psychedelic-induced changes and suggest the potential for this component as a transdiagnostic therapeutic target. Notwithstanding the importance of these contributions, a number of methodological limitations necessitate further empirical development in this area. First, two of the relevant studies used cross-sectional designs that inherently preclude analysis of process-level psychedelic changes and may be impacted by retrospective bias (Davis et al., 2020; Thiessen et al., 2018). Second, although two of these studies employed prospective designs, their use of naturalistic, volunteer samples with high rates of attrition (> 50%) may have contributed self-selection and sample biases that limit generalizability (Close et al., 2020; Zeifman et al., 2020). Finally, the studies included relatively brief follow-up periods (e.g., 1 month).

In addition to psychological flexibility, reappraisal is a related cognitive process within contemporary psychotherapies that has shown some evidence of occurring during psychedelic experience. Cognitive reappraisal is generally considered an adaptive strategy for decreasing emotional avoidance and flexibly regulating emotions. It represents an approach-oriented emotion regulation strategy that involves engaging with an emotion-eliciting circumstance and reinterpreting it in a way that modifies its emotional impact, response, and meaning (Gross and John, 2003; Lazarus and Folkman, 1984). Cognitive reappraisal is considered more effective than inhibitory strategies that involve suppression or avoidance (Aldao et al., 2010; Webb et al., 2012) and has been proposed as an underlying mechanism of cognitive-behavioral (CBT) and mindfulness-based therapies (Buhle et al., 2014; Sheppes et al., 2014; Teasdale et al., 2002). The habitual use of cognitive reappraisal has been associated with reduced symptoms of depression, increased positive emotions, and psychological well-being (Gross and John, 2003). Important to the present examination, the use of cognitive reappraisal strategies is thought to itself depend on psychological flexibility. The ability to flexibly attend to, modulate, and process emotional stimuli subserves the implementation and maintenance of new reappraisals and meanings (McRae et al., 2012; Malooly et al., 2013; Schmeichel et al., 2008). As such, reappraisal may represent a meaningful index of psychological flexibility processes. Despite increased attention in the literature on the psychological processes involved in psychedelic interventions, few studies have applied an emotion regulation perspective such as reappraisal toward an understanding of the dynamics of psychological processes catalyzed by a psychedelic. One example is the work of Davis et al. (2021) who, using a preliminary cross-sectional design, found relations between an original measure of insight (that appears theoretically comparable to cognitive reappraisal) and adaptive changes in psychological flexibility, well-being, and life satisfaction following use of a psychedelic. However, most studies that have used an emotion regulation perspective have focused more heavily on acute emotional engagement and breakthrough rather than on the ability to evaluate and challenge the validity of automatic cognitions and core beliefs during a psychedelic session (Close et al., 2020; Roseman et al., 2019; Thiessen et al., 2018).

Perhaps most germane to the present study, there is some evidence that cognitive reappraisal and acceptance-based emotion regulation in psychedelic experience are important features of the ceremonial use of NN-DMT-containing ayahuasca. Ayahuasca is a potent psychoactive plant brew indigenous to the South American Amazon basin that combines the Banisteriopsis caapi vine with the leaves of Psychotria viridis plant (Schultes and Farnsworth, 1980), and is traditionally consumed by Amazonian indigenous communities in group ceremonial contexts involving chanting of prayers (e.g. icaros), sacred tobacco, and interactions with a shaman (e.g. an ayahuascero/a). Ayahuasca stimulates emotional dream-like experiences involving potent cognitive-affective changes and spiritual insights (Dobkin de Rios, 1971) that have been associated with substantial and enduring antidepressant effects (Palhano-Fontes et al., 2019; Weiss et al., 2021). Among these cognitive-affective changes, recent evidence suggests that acute and post-acute experiences of cognitive reappraisal and psychological flexibility support introspection and the reframing of difficult life experiences and negative core beliefs (Weiss et al., 2021). These processes appear to be associated with adaptive changes in neuroticism, extraversion, agreeableness, and conscientiousness (Weiss et al., 2021). Specifically, an acute psychological factor involving strong themes of reappraisal and psychological flexibility emerged from qualitative work on the ceremonial ayahuasca experience (Ayahuasca Experience Inventory (AEI) Reappraisal subscale; Weiss et al., 2020), suggesting that states involving flexibility, reappraisal, and insight may be organic to psychedelic-induced experiences in a shamanic setting. In addition, several studies have found positive associations between ayahuasca use and emotional acceptance processes, which appears to support the acute processing of previously avoided emotions and use of healthy emotion regulation strategies (Domínguez-Clavé et al., 2019; Franquesa et al., 2018; Sampedro et al., 2017; Soler et al., 2016, 2018).

This prospective study had three goals. First, the study examined change in psychological flexibility and negative and positive affect following the ceremonial Shipibo use of ayahuasca. Next, the study examined whether affect and psychological flexibility covaried over time. The study also explored whether moderating factors predisposed participants to experience therapeutic changes, with specific attention to the acute psychological factors of psychological flexibility/cognitive reappraisal processes (referred heretofore as “flexibility/reappraisal”). Few previous studies have rigorously examined potential moderators (e.g. sample and design characteristics), which may provide necessary conditions for change. Finally, to address a substantial limitation present in previous literature (Aday et al., 2021; Muthukumaraswamy et al., 2021), expectancy, demand characteristics, and invalid responding were controlled for using several validity scales. It was hypothesized that (1) acute factors (e.g. flexibility/reappraisal) during ayahuasca ceremony would be associated with persisting improvements in positive and negative affect and increases in psychological flexibility and (2) increases in psychological flexibility would have indirect effects on the relationship between acute factors and improvements in positive and negative affect (i.e. increases in positive affect and decreases in negative affect).

Methods

Participants and procedure

A total 330 participants were recruited from three ayahuasca retreat centers across Central and South America: Arkana Spiritual Center (Requena, Loreto, Peru), Soltara Healing Center (Gulf of Nicoya, Costa Rica), and La Medicina (Cordilliera Escalera mountain range, Peru). The research team provided prospective participants an informational sheet about the optional study and then received informed consent to participate in the study before each retreat. The research team was not involved in the retreat recruitment or in the ayahuasca administrations. Participants were entered into a raffle for a week-long retreat at Arkana Spiritual Center (valued at US$1580.00), and further incentives were used to promote compliance and reduce attrition (see the study by Weiss et al., 2021 for details). Assessments were performed at two timepoints (i.e. 2 weeks before the retreat (baseline) and 3 months after the retreat). All procedures were approved by the University of Georgia Institutional Review Board. Full study procedures are described by Weiss et al. (2021). Participants were required to complete online medical screening questionnaires on the retreat centers’ websites prior to enrollment in the retreats.

Eligibility criteria for participation in the retreats were determined by the participating retreat centers. General psychiatric exclusion criteria included known diagnoses of schizophrenia, bipolar I and II disorders, personality disorders, and active substance use disorders. Medical criteria included uncontrolled hypertension or cardiovascular conditions, epilepsy, active tuberculosis, and kidney or liver failure. In addition, retreat centers required that individuals taper off beta-blockers as well as selective serotonin reuptake and serotonin–norepinephrine reuptake inhibitors under the guidance of their personal medical providers up to 6 weeks prior to the retreats. Inclusion criteria for participation in this study included being at least 18 years old, able to read and write in fluent English, and participation in at least one ayahuasca retreat at a participating retreat center.

Retreat center experience.

Arkana Spiritual Center, Soltara Healing Center, and La Medicina regularly host shamans from the Shipibo lineage (originating geographically in the west Amazon basin). Among the ceremonies offered to retreat clients, the ayahuasca ceremony (offered approximately four times each week) represents the most time-intensive and immersive practice. Participants in the study were asked to provide information regarding characteristics of their retreat experience including the number of ceremonies in which they consumed ayahuasca, retreat length, average dosage of ayahuasca, and experiences with other plant medicines (e.g. huachuma, sapo). Dosage was computed as the approximate ayahuasca quantity consumed across ceremonies in terms of glasses (e.g. one and one-half glasses, approximately 120 to 150 mL in volume of ayahuasca brew). Of note, glass size was not standardized and varied within and across retreat centers. Complete data for dosage and frequency of ceremonies were available for 240 participants. Complete data concerning other plant medicines used were available for 220 participants.

Measures

Evaluation of validity (expectancy and suggestibility).

To examine whether change in outcomes was present while controlling for validity variables, the study examined self-reported change in primary variables among participants who endorsed no expectation of favorable change and, in a separate group, among those who exhibited low trait suggestibility at baseline.

Expectancies.

An original scale was developed, consisting of six dichotomous (yes/no) items measuring expected change in personality domains (“I will become more spiritual”), and depression symptoms (e.g. “I will experience a significant reduction of depressive thinking”).

Suggestibility.

The Multidisciplinary Iowa Suggestibility Scale-Short (MISS; Kotov et al., 2004) was used to measure participants’ susceptibility to internalize external influences. The MISS is a 21-item self-report scale that uses a 5-point Likert-type scale ranging from “Strongly disagree” to “Strongly agree.” Low suggestibility was indexed by scores below the 50th percentile (equivalent to “Somewhat Disagree” on response scale).

Overreporting of mystical-type experiences.

An original three-item scale was used to measure overreporting of acute mystical-type experiences. Participants were asked the degree to which they experienced the following low base-rate phenomena: “Experience of a distant childhood friend you have not seen or thought of in a long time,” “Rapidly fluctuating pattern of feelings alternating from joy to sadness and back again,” and “Experience of bodily fragmentation, such that parts of your body are separated from one another.” The cut-off for invalid responding was defined by the study team as “Strong” or “Extreme” endorsement of all three validity items (on a 6-point Likert-type scale). Based on this cut-off, data for measures indexing acute experience were excluded for 11 participants.

Invalid responding.

Two eight-item validity scales from the Elemental Psychopathy Assessment (Lynam et al., 2011) were used to detect invalid responding on measures of personality. These scales were the Infrequency scale and the Unlikely Virtue scale. Participants who endorsed more than three Infrequency scale items and more than two Unlikely Virtue scale items were removed from analyses.

Outcome measures

Psychological flexibility.

The Acceptance and Action Questionnaire (AAQ; Hayes et al., 2004) was administered to assess psychological flexibility, and measures difficulty managing negative emotions (“When I feel depressed or anxious, I am unable to take care of my responsibilities”) and avoidance of thoughts and feelings (“I’m (not) afraid of my feelings”). The measure was administered at baseline and the 3-month follow-up. Due to an administrative error, the last two items were accidentally excised from the measure administered to participants, and the final version consisted of seven total items using a 7-point Likert-type scale. Cronbach’s alphas at baseline and 3-month follow-up, respectively, were 0.75 and 0.64; these values are similar to the low-to-moderate internal consistency values demonstrated in previous studies with this measure (Hayes et al., 2004). Complete data were available for 261 participants.

Positive and negative affect.

The Positive and Negative Affect Schedule (PANAS; Watson et al., 1988) was used to measure dispositional experiences of negative and positive affect. The PANAS consists of 20 items with 10 positive (PANAS-P) and 10 negative (PANAS-N) affective descriptors, and yields two factors. The PANAS uses a 5-point Likert-type scale. Participants responded according to how they felt during the past week. The PANAS has demonstrated good psychometric properties (Watson et al., 1988). Cronbach’s alphas in our study at baseline and 3-month follow-up, respectively, were 0.90 and 0.91 for PANAS-P and 0.88 and 0.85 for PANAS-N. Complete data were available for 260 participants.

Acute experience factors

Mystical experience.

The Mystical Experience Questionnaire (MEQ-30) a 30-item scale used to assess mystical aspects of participants’ experiences during their ayahuasca ceremonies. It includes four subscales that were assessed: Mystical (15-item; e.g. “Experience of the fusion of your personal self into a larger whole”), Positive mood (6-item; e.g. “Sense of awe or awesomeness”), Transcendence of time and space (6-item; e.g. “Loss of your usual sense of space”), and Ineffability (3-item; e.g. “Sense that the experience cannot be described adequately in words”). Items asked participants to consider the degree to which they had experienced the preceding phenomena at any time during the ceremonies. Items used a 6-point Likert-type scale. One of the items from the Mystical subscale (item 30) was excluded due to administrative error. Cronbach’s alphas ranged from 0.87 (Ineffability) to 0.96 (Mystical). Complete data were available for 236 participants.

Ego dissolution.

The Ego Dissolution Inventory (EDI; Nour et al., 2016) was used to measure dissolution of ego during the acute effects of ayahuasca. The EDI consists of eight items (e.g. “I experienced a disintegration of my ‘self’ or ego”) using a 5-point Likert-type scale. Internal consistency was good (α = 0.89). Complete data on this measure were available for 236 participants.

Ayahuasca experience inventory.

Using qualitative methods, the AEI (Weiss et al., 2021) was developed to measure thoughts, feelings, behaviors, and attitudes that arise within ayahuasca ceremonies. Participant scores, using Thurstone regression-based weighting, were generated from factor analyses (see model fit results and loadings in Weiss et al., 2021). The AEI consists of three factors and does not consist of a total score factor: Clarity (32 items) captures psychological clarity, peace, self-connection, and self-esteem (e.g. “I was able to perceive my thoughts/feelings with exceeding clarity” and “I saw a roadmap for living in alignment with my authentic self”). Reappraisal (30 items) captures cognitive reappraisal through indexing the processing of negative beliefs about self/others (“I felt that I was forced to confront negative perceptions I’ve had of myself”) and coming to adaptive conclusions (“I was able to see new positive meaning in a past trauma,” “I overcame/resolved my fears,” “I realized how courageous I am”). This factor is also thought to index psychological flexibility in reflecting productive engagement with challenging psychological content (“I identified aspects of myself that cause me pain”) and the initiative to enact valued actions (e.g. “I felt capable of living courageously”). Discomfort (15 items) captures unpleasant feelings of torment, discomfort, and isolation (e.g. “I felt lost” and “I felt overwhelmed by the experience”). Items asked participants to consider the degree to which they had experienced phenomena at any time during ceremony. Items were measured using a 6-point Likert-type scale. AEI factors exhibited excellent internal consistency, with Cronbach’s alphas ranging from 0.94 (discomfort) to 0.97 (clarity). Convergent validity was assessed by examining correlations between AEI subscales and long-standing acute factors with demonstrated links to favorable outcomes (e.g. MEQ, EDI). Criterion validity (or predictive validity) was assessed by examining correlations between AEI subscales and change scores in study outcomes (i.e. psychological flexibility, trait affect). Complete data on this measure were available for 187 participants. The full AEI scale can be found in Supplementary Materials.

Analytic plan.

Four main sets of analyses were undertaken. To reduce Type I error, across analyses, a statistical significance threshold was set at p < 0.01 for all analyses. Prior to conducting the primary analyses, independent samples t-tests were conducted to test for the presence of sample bias driven by differences in personality characteristics between validly responding participants who missed one of the two timepoints (n = 36), and validly responding participants with no missing data (n = 261) on primary outcome measures. All assumptions were met.

The first set of exploratory analyses assessed the degree to which psychological flexibility, positive affect, and negative affect changed following the ayahuasca ceremonies. To determine the persisting effects of the ayahuasca ceremonies, each outcome was compared between baseline and the 3-month follow-up using linear mixed models (similar to one-way repeated measures analysis of variance (ANOVA)) in which outcome scores were separately regressed onto time. The following indices of effect size were used: unstandardized (B) coefficients indicate mean differences between timepoints. d_z indicates effect size change in personality scores in terms of the standard deviation of within-subject change scores (e.g. time point 2 − time point 1; Lakens, 2013). Marginal R² values were used to indicate the degree to which the fixed effect of time accounted for variance in outcomes.

The second set of confirmatory analyses examined the degree to which psychological flexibility covaried over time with negative affect and positive affect. Change scores were calculated by subtracting participants’ 3-month follow-up scores from baseline scores, and Pearson’s correlations were used to calculate correlations between change scores.

In the third set of exploratory analyses, to investigate the degree to which changes between timepoints were affected by predisposing and experiential factors, linear mixed models were conducted in which moderators were separately added to the base models (outcome ~ time) as fixed covariates. Four sets of moderator variables were examined including validity variables (including expectancies and suggestibility), participant characteristics (including baseline psychological flexibility, demographic variables, lifetime use of psychedelics), acute experiential factors, and ceremony characteristics. For continuous moderators, unstandardized (B) coefficients were used to indicate the added effect of the moderator to the effect of time at one standard deviation above the mean of the moderator. For dichotomous moderators, unstandardized (B) coefficients were used to indicate the added effect of the moderator to the effect of time at one level of the moderator versus the other. Marginal R² values were used to indicate the degree to which fixed effect variables (i.e. time, moderator) accounted for variance in outcomes. Correlations were additionally calculated between acute factors and outcome change scores (3-month follow-up subtracted from baseline). (Note: These analyses are conceptually identical to moderation-based analyses presented above but are presented to facilitate reader interpretation and comparison across other findings in the literature.)

In the fourth set of confirmatory analyses, the indirect effects of psychological flexibility (i.e. statistical mediation) were assessed in the relation between acute experiential factors and change in mood using the PROCESS macro (Hayes and Preacher, 2013) of SPSS version 25. In this approach, effects are assessed with bias corrected bootstrap confidence intervals (CIs) that are significant when the upper and lower bound of the bias-corrected 95% CIs do not contain zero. The following a priori hypothesis was tested: Changes in psychological flexibility (i.e. change score relative to baseline) would bear an indirect effect on the relationships between acute factors (MEQ-30 positive mood, AEI reappraisal, and AEI clarity) on changes in mood (increases in positive affect). Due to the high correlation between psychological flexibility (indexed with the AAQ) and negative affect in previous studies (Rochefort et al., 2018; Tyndall et al., 2018; Wolgast, 2014), a competing model was also tested to determine whether change in negative affect also exerted significant indirect effects in the relation between acute experiential factors and change in positive mood. Each experiential factor variable was tested separately.

Results

Participant characteristics

Of the 330 participants who were recruited, 3% (n = 10) met criteria for invalid responding based on the Elemental Psychopathy Assessment (Lynam et al., 2011) validity scales, leaving a sample of 320 validly responding participants possessing data from at least one timepoint. Data were furthermore removed in pairwise manner for seven participants because of inadequate time commitment and untimely completion (i.e. the total survey time duration fell below one standard deviation or there was substantial deviation from their survey timepoints). All data met assumptions of normality according to guidelines indicated by Hair et al. (2010) and Byrne (2010).

The final sample consisted of 261 participants after 69 were removed because of not providing 3-month follow-up data (18% attrition rate; 166 self-identified males and 94 self-identified females; mean age = 34.8 (SD = 10.0); 91% non-Hispanic White, 2% Black, 6% Asian American/Native Hawaiian or Other Pacific Islander, 9% Latinx, and 3% American Indian/Native American). Approximately 24% of participants reported an annual income of less than US$30,000, 30% reported annual income greater than US$30,000 but less than US$60,000, 15% reported greater than US$60,000 but less than US$90,000, and 31% greater than US$90,000. Approximately 26% of participants reported completing some high school, 42% reported graduating high school, 16% reported completing some college, and 15% reported obtaining at least a bachelor’s degree. Approximately 13% of participants had self-reported diagnoses of major depression, 9% with post-traumatic stress disorder, 6% with anxiety disorders, and 4% with a traumatic brain injury. At baseline, participants with missing data differed from participants with full data only based on lower conscientiousness scores (d = 0.35, p < 0.01).

Examining change in psychological flexibility and positive and negative affect

Main effects of timepoint were observed, indicating increases in psychological flexibility (B = 0.50, SE = 0.05, p < 0.001, mR² = 0.08, Cohen’s d_z = 0.60) and positive affect (B = 0.35, SE = 0.05, p < 0.0001, mR² = 0.05, d_z = 0.41) from baseline to 3-month follow up, as well as decreases in negative affect (B = −0.35, SE = 0.05, p < 0.001, mR² = 0.06, d_z = −0.46). These results are displayed in Figures 1 to 3.

Figure 1.

Box and Violin plot illustrating change in psychological flexibility between baseline and 3-month follow-up.

Figure 3.

Box and Violin plot illustrating change in positive affect between baseline and follow-up.

Covariation between outcomes over time

Results indicated that, relative to baseline, increases in psychological flexibility were associated with (1) decreases in negative affect (r = −0.53; p < 0.001) and (2) increases in positive affect (r = 0.45; p < 0.001) over time. Increases in positive affect also associated with decreases in negative affect (r = −0.43; p < 0.001) over time. Correlations are presented in Table 1.

Table 1.

Correlations among acute factors and primary variables.

Measure	1	2	3	4	5	6	7	8	9	10
1. MEQ-30 mystical
2. MEQ-30 positive mood	0.80*
3. MEQ-30 transcend time/space	0.65*	0.54*
4. MEQ-30 ineffable	0.50*	0.43*	0.50*
5. Ego dissolution	0.65*	0.56*	0.57*	0.39*
6. AEI clarity	0.85*	0.84*	0.55*	0.45*	0.64*
7. AEI reappraisal	0.63*	0.61*	0.46*	0.39*	0.53*	0.74*
8. AEI discomfort	0.05	0.02	0.27*	0.25*	0.10	0.04	0.35*
9. ΔPsychological flexibility	0.17*	0.22*	0.14	0.11	0.10	0.25*	0.26*	0.16	.
10. ΔPositive affect	0.16	0.18*	0.11	0.11	0.08	0.19*	0.19*	0.01	0.45*
11. ΔNegative affect	0.04	0.10	0.09	0.06	0.04	0.06	0.14	0.16	−0.53*	−0.43*

MEQ-30: Mystical Experience Questionnaire-30; AEI: Ayahuasca Experience Inventory.

^*p < 0.01.

Moderation of change in psychological flexibility and positive and negative affect

A summary of significant results from omnibus tests and change in marginal R² after adding interaction terms is provided in Table 2. See Supplementary Table 1 for full results.

Table 2.

Degree to which moderators influence the effect of time on outcomes—interaction coefficients.

Moderator	N	PF	NA	PA	Moderator	N	PF	NA	PA
Expectancies					Acute experience factors
Expected increase in spirituality	261	0.05	0.00	−0.02	Ayahuasca-naïve	235	0.07	−0.11	0.04
Expected increase in connectedness	261	0.08	0.02	−0.02	MEQ-30 mystical	236	0.14	−0.03	0.14
Expected decrease in N	261	0.20	−0.16	0.14	MEQ-30 positive mood	236	0.17	−0.08	0.16
Expected increase in E	261	0.10	−0.10	0.11	MEQ-30 transcend time/space	236	0.11	−0.07	0.10
Expected decrease in anxiety	261	0.11	−0.16	0.09	MEQ-30 ineffable	236	0.09	−0.05	0.10
Expected decrease in depression	261	0.22	−0.16	0.09	Ego dissolution inventory	236	0.09	−0.03	0.07
Suggestibility	261	0.20	−0.08	0.14	AEI clarity	187	0.20	−0.04	0.16
Baseline PF	261		−0.29	0.27	AEI reappraisal	187	0.21	−0.09	0.16
Sex (male)	261	0.00	−0.01	−0.02	AEI discomfort	187	0.13	−0.11	−0.01
Age	261	−0.11	0.04	0.01	Number of ceremonies	240	0.04	−0.01	−0.02
Education level	261	0.03	−0.01	0.01	Average consumed	240	−0.01	0.04	0.02
Parent income	261	−0.04	0.06	−0.01	Additional psychedelic	220	−0.06	−0.04	0.02
Psychedelic-naïve	235	0.08	−0.11	0.14	Retreat length	261	0.00	0.01	−0.03

PF: psychological flexibility; NA: negative affect; PA: positive affect; N: neuroticism; E: extraversion; MEQ-30: Mystical Experience Questionnaire-30; AEI: Ayahuasca Experience Inventory.

Incremental variance explained by moderators in unstandardized (B) coefficients. Interaction term coefficients meeting p < 0.01 threshold are in bold.

Expectancy and suggestibility.

In separate models, expectancies of favorable changes in personality and mental health benefits (i.e. expected reductions in anxiety, depression, and neuroticism, and increases in spirituality, connectedness, and extraversion) were analyzed to see if they moderated changes in psychological flexibility, positive affect, and negative affect (see Table 2). Results showed that expectancies of favorable changes (on dichotomous expectancy scale) in neuroticism and depression were associated with greater increases in psychological flexibility (ExpectancyΔNeurot: B = 0.20, SE = 0.05, p < 0.001; ExpectancyΔDep: B = 0.22, SE = 0.05, p < 0.001) and reductions in negative affect (ExpectancyΔNeurot: B = −0.16, SE = 0.05, p < 0.001; ExpectancyΔDep: B = −0.16, SE = 0.05, p < 0.001). Expectancy of favorable change in anxiety was also associated with greater reductions in negative affect (ExpectancyΔAnx: B = −0.16, SE = 0.05, p < 0.001).

Notably, however, among the participants who denied (on dichotomous expectancy scale) expectancies of favorable changes in neuroticism, depression, and anxiety, statistically significant changes were still detected in psychological flexibility and negative affect (e.g. ExpectancyΔNeurot: d_zPF = 0.36; d_zNA = −0.26; N = 109, p < 0.001). With regard to trait suggestibility, higher suggestibility scores at baseline were associated with greater increases in psychological flexibility (B = 0.20, SE = 0.05, p < 0.001) and positive affect (B = 0.14, SE = 0.05, p < 0.001) but not negative affect. In addition, participants who exhibited low trait suggestibility (indexed by scores below the 50th percentile) also showed significant changes in psychological flexibility and positive affect (d_zPF = 0.40; d_zPA = 0.29; N = 142, p < 0.001).

Participant and retreat characteristics.

With respect to participant characteristics, baseline psychological flexibility moderated changes in negative affect (B = −0.29; SE = 0.04, p < 0.001) and positive affect (B = 0.27; SE = 0.05, p < 0.001), suggesting that participants with higher psychological flexibility at baseline showed larger improvements in negative affect and positive affect. No significant moderation effects were detected for demographic variables or retreat characteristics (see Table 2).

Acute experience factors.

With respect to acute factors, MEQ-30 mystical (B = 0.14, SE = 0.05, p < 0.001, MEQ-30 positive mood (B = 0.17, SE = 0.05, p < 0.001), AEI clarity (B = 0.20, SE = 0.06, p < 0.001), and AEI reappraisal (B = 0.21, SE = 0.06, p < 0.001) significantly moderated increases in psychological flexibility (see Table 2) and positive affect. Furthermore, MEQ-30 positive mood (B = 0.16, SE = 0.06, p < 0.001, AEI clarity (B = 0.16, SE = 0.06, p < 0.001, and AEI Reappraisal (B = 0.16, SE = 0.06, p < 0.001) significantly moderated increases in positive affect. In addition, AEI reappraisal was unique among moderators in bearing significant associations with initial levels of psychological flexibility (i.e. lower baseline psychological flexibility predicted higher AEI reappraisal; r = −0.28, p < 0.001), and showing the largest moderating effect on change in psychological flexibility (B = 0.21, SE = 0.06, p = 0.008) and positive affect (B = 0.16, SE = 0.06, p < 0.001). Number of ceremonies, retreat length, average amount consumed, and taking an additional psychedelic were not significantly associated with any outcomes (all p > 0.01; see Table 1). In general, higher levels of mystical experience, clarity, and reappraisal during the ceremony were associated with steeper increases in psychological flexibility and positive affect over time.

Table 1 illustrates the significant correlations found between all acute factor measures (all four MEQ-30 factors, AEI, and EDI) except for the AEI discomfort factor, which demonstrated significant associations only with AEI Reappraisal (r = 0.35, p < 0.001), MEQ-30 transcendence of time/space (r = 0.27, p < 0.001), and MEQ-30 ineffability (r = 0.25, p = 0.001). With regard to acute factors predicting outcome change scores, MEQ-30 positive mood, AEI reappraisal, and AEI clarity were significantly correlated with increases in psychological flexibility (r = 0.22, p < 0.001; r = 0.26, p < 0.001; r = 0.25, p < 0.001, respectively) and positive affect (r = 0.18, p = 0.005; r = 0.19, p = 0.009; r = 0.19, p = 0.009, respectively). MEQ-30 mystical was also significantly correlated with increases in psychological flexibility (r = 0.26, p = 0.008).

Indirect effect of psychological flexibility on changes in affect

Change in psychological flexibility showed significant, moderate, indirect effects in the relationships between acute factors and positive affect: MEQ-30 positive mood (B = 0.07; SE = 0.02; 95% bias-corrected bootstrap CI = 0.12, 0.03), AEI clarity (B = 0.08; SE = 0.03; CI = 0.13, 0.03), and AEI reappraisal (B = 0.08; SE = 0.03; CI = 0.14, 0.04). In a competing model, change in negative affect did not reach significance when testing for its indirect effect in the relationship between acute factors and positive affect: MEQ-30 positive mood (B = 0.03; SE = 0.02l; CI = −0.07, 0.01), AEI clarity (B = 0.02; SE = 0.06; CI = −0.07, 0.03), and AEI reappraisal (B = 0.04; SE = 0.02; CI = −0.09, 0.01). Results are presented in Figure 4.

Figure 4.

Diagram showing that indirect effects of psychological flexibility on positive mood outcomes.

Diagram showing that increases in psychological flexibility statistically mediate the relation between acute experience (acute reappraisal, mystical-type positive mood, and acute clarity) and increases in positive affect while controlling for changes in negative affect (N = 235). *p < 0.01.

Discussion

This study reports on psychological outcomes from a well-sized prospective study of Shipibo ceremonial ayahuasca use occurring at three retreat centers across Central and South America. The study primarily examined (1) whether ayahuasca experiences were associated with enduring changes in psychological flexibility and positive and negative mood; and (2) whether predisposing and acute factors served as moderators and statistical mediators (i.e. indirect effects) of adaptive changes.

In this study, psychological flexibility and mood (increased positive and reduced negative mood) improved following the retreat, results that parallel those found in previous studies involving use of psychedelics in therapeutic or retreat contexts (Close et al., 2020; Davis et al., 2020; Zeifman et al., 2020). It is important to note that expectancy and placebo effects predicted outcomes in this study, with favorable changes in mental health benefits (e.g. expected reductions in anxiety and depression) accruing more heavily for those with positive expectancies and trait suggestibility. This raises the serious challenge that positive effects may not be directly attributable to ayahuasca but rather to the special mind-set of ayahuasca retreat-goers amid the perceived significance of the ceremony experience. Nevertheless, statistically significant changes in psychological flexibility and mood persisted among the subsample who endorsed no positive expectancies and low suggestibility (i.e. those who were less optimistic and more skeptical about the benefits of ayahuasca). This provides greater confidence in the validity of the findings. The contribution of expectancy and placebo effects may not undermine efficacy as some critics suggest (Muthukumaraswamy et al., 2021), as these effects may work synergistically with genuine psychedelic-induced therapeutic mechanisms to potentiate greater change. For instance, individuals who expect ayahuasca will serve as a genuine source of healing may be more willing and able to engage in reappraisal and psychological flexibility processes during ceremony.

Psychological flexibility was associated with changes in mood and other variables as demonstrated in a variety of models in this study, which suggest common processes underlying these changes. Improvements in psychological flexibility from baseline to follow-up were significantly associated with improvements in mood (increased positive and reduced negative mood) over the same interval. Furthermore, baseline psychological flexibility moderated changes in negative affect; participants with higher psychological flexibility at baseline showed larger improvements in negative affect and positive affect over time. Furthermore, psychological flexibility mediated the relationship between acute factors during ceremony (e.g. AEI reappraisal) and improvements in positive mood at the 3-month follow-up. This finding suggests that adaptive change in dispositional psychological flexibility may account for some of the effect of psychological and spiritual factors during the ceremony on long-term increases in positive mood. Furthermore, processes of acute psychological clarity, introspective awareness, and reappraisal of difficult life experiences during ayahuasca may upregulate a dispositional capacity for psychological flexibility, and increased psychological flexibility may, in turn, functionally drive positive mood. It is also important to note that psychological flexibility did not mediate the effects of acute factors on negative affect. One possibility is that this may reflect some method effects in that the acute factors studied likely relate to more positive mood states versus the negative mood where mediation effects were not identified. Another possibility is that the acute effects may result in alterations in aspects of psychological flexibility more likely tied to positive affect, such as increased contact with life values or increases in value-based behavior, rather than aspects of psychological flexibility that might be more linked with negative affect such as acceptance. However, additional research is needed to clarify this speculative hypothesis. Finally, the present findings may be suggestive of a mediation pathway between acute factors, psychological flexibility, and positive affect; however, only studies with a control group can establish evidence of causation and mechanisms (Tryon, 2018). Future studies are also needed to clarify the impact of acute and post-acute mediators (e.g. therapeutic changes occurring 1 week after a psychedelic session) of effects on final outcomes (Kangaslampi, 2020).

This study utilized a novel measure to index flexibility/reappraisal, clarity, and spiritual-type experiences during ceremony (AEI: Weiss et al., 2020, 2021; planned validation in a forthcoming article) as well as a well-established measure of mystical experience (MEQ-30; Barrett et al., 2015). Among acute factors, higher levels of reappraisal during the ceremony (AEI reappraisal) showed the strongest associations with increases in positive mood and psychological flexibility from baseline to 3-month follow-up. Furthermore, the intensity of these acute factors (with the exception of MEQ-30 mystical) predicted increases in psychological flexibility and positive mood at the 3-month follow up. These results are aligned with previous studies associating mystical experience with improved outcomes (Griffiths et al., 2016; Ross et al., 2016) and add support to the novel idea that cognitive reappraisal may have a role in therapeutic psychedelic-induced changes. These results build upon previous cross-sectional work suggesting links between experiences of insight during a psychedelic and enhanced well-being and life satisfaction post-session (Davis et al., 2021). Notwithstanding these important acute experiences of reappraisal and other factors, it is important to acknowledge potential differences between how appraisal is measured in this study and how it is defined in CBT, ACT, and the broader literature. The AEI Reappraisal scale specifically indexes (1) access to challenging psychological content and (2) the emergence of resolved perspectives (Weiss et al., 2020, 2021). As such, future research is needed to investigate the degree to which psychedelic versus non-psychedelic processes of reappraisal converge. Furthermore, the therapeutic mechanism underlying cognitive therapy is believed to involve the capacity to hold rational cognitions about self, others, and world. Although this mechanism may be operant within psychedelic reappraisal, it is important to recognize alternative mechanisms unique to psychedelic experience that may lead to radically different perspectives (e.g. forgiveness). These may include the relative depth of revelatory cognitions about the self, others, and world; the possibility of direct knowledge (e.g. noetic quality); as well as a process of undergoing personal transformation such that new perspectives are felt to be more reasonable and adequate (Forstmann et al., 2020).

These results converge with previous studies that have shown that ayahuasca and other psychedelics, when used in therapeutic or healing contexts, often result in acute increases in positively valenced emotional states (e.g. captured in this study with MEQ-30 positive mood). Furthermore, these results are consistent with the idea that acute increases in positive mood states, perhaps those related to heightened self-transcendent positive emotions (e.g. gratitude, love, joy, and awe; Agin-Liebes et al., 2021; Belser et al., 2017; Swift et al., 2017; Watts et al., 2017), may create a context for enhanced flexibility or reappraisal to occur by rendering engagement with difficult psychological material more palatable during retreat sessions and for some time thereafter. Indeed, in this study, there was a significant positive correlation between reappraisal and mystical-type positive mood (r = 0.61), and a smaller but significant correlation between reappraisal and discomfort (r = 0.35) during ceremony, which suggests the occurrence of a nuanced psychological process. Some data suggest that psychedelics may decrease threat sensitivity and attenuate reactivity in brain regions associated with emotional processing during the acute phase of a psychedelic session (Kraehenmann et al., 2015, 2016). Other functional magnetic resonance imaging (fMRI) data suggest that psilocybin may increase amygdala responsiveness to both fearful and joyful faces while reducing the top-down inhibitory effects of prefrontal control regions on limbic regions 24 h after the session (Mertens et al., 2020; Roseman et al., 2018). Ayahuasca and other classic psychedelics may lead to broadened metacognitive awareness accompanying psychological flexibility and reappraisal that contribute to organizing and resolving psychologically difficult content through acute engagement with intense emotions and memories (i.e. “emotional breakthroughs”; Roseman et al., 2019). The processing of challenging emotions may then yield profound states of insight and release and, in turn, to long-term eudaimonic positive mood. This notion is supported by Carhart-Harris and Friston’s (2019) relaxed beliefs under psychedelics model, which proposes that psychedelics precipitate a freer and less constrained states of consciousness. These conditions may increase one’s sensitivity to bottom-up information (e.g. emotions/sensations and contextual evidence undermining maladaptive beliefs), allowing for the long-term revision of cognitive and emotional schemas.

Another potentially convergent model, Garland and Fredrickson’s (2019) mindfulness-to-meaning theory, posits that the metacognitive effects of mindfulness are integral to psychedelic experience and function to disrupt default habits of self-referential processing, stimulate attentional broadening, and elicit positive emotions. These processes, together, are thought to enable integration of positive contextual information into one’s personal narrative (e.g. gratitude for support of loved ones in spite of personal struggles). By integrating a broadened scope of contextual features into awareness, one may perceive a wider array of possibilities and forge novel situational appraisals (Garland and Fredrickson, 2019). Notably, the attentional broadening facet of this process has been longitudinally linked to increases in “decentering” (Garland et al., 2017), the ability to observe one’s thoughts and feelings as mental events, rather than representations of reality, in a way that permits new perspective-taking and enhanced reality testing (Fresco et al., 2007; Kerr et al., 2011). Decentering has been found to be dispositionally higher in ayahuasca users (Franquesa et al., 2018) and shows evidence of being elevated 24 h post-ayahuasca (Murphy-Beiner and Soar, 2020; Soler et al., 2016). The detachment component of decentering may support psychological flexibility by reducing reactivity to the putative claims of one’s negative thoughts/beliefs and assist cognitive reappraisal by expanding one’s band-width for evaluating recurrent high-level beliefs and reducing information bias. The capacity for observing one’s internal states may be instrumental in expanding awareness of the self’s cognitive, emotional, behavioral, and relational patterns, thereby facilitating deliberation about possibilities that serve versus encumber one’s values. Psychedelics may furthermore enhance the capacity for flexibility in adopting new perspectives, reappraisals, and conclusions based on this contextual information.

It is also important to consider how acute experiences of heightened flexibility/reappraisal, as well as changes in dispositional psychological flexibility and mood, stem from the unique context of ayahuasca shamanism. It is widely accepted that the therapeutic effects of psychedelic experiences reflect synergies between pharmacological actions, mind-set, setting, and ontological frameworks (i.e. philosophies about the nature of being and existence; Hartogsohn, 2016). First, facilitators and shamans are instrumental in cultivating a group atmosphere in which emotional exploration of challenging content is valued. While acute effects of ayahuasca may naturally promote this exploration, the retreat atmosphere heavily reinforces it, such that individuals may approach challenging emotional material with greater curiosity and engagement (Uthaug et al., 2021). Second, the ayahuasca ceremony appears to present individuals with states of psychological difficulty (both psychological and physical), and the group format baldly displays this process to other group members. Examples include audible displays of grief, moaning, and physical purging that command others’ awareness. This experience of shared vulnerability and emotional processing may set in motion processes congruent with psychological flexibility such as heightened feelings of compassion toward others/oneself that help one transform their relationship to their own and others’ emotional pain. Third, the icaro, in which the shaman directs a prayer to an individual group member, may encourage group members’ productive engagement with challenging emotional content. Also relevant to this discussion is the increasing recognition of the importance of interpersonal and relational processes in models of mental health and psychosocial functioning (Maercker and Horn, 2012). Findings from recent research on group-based psychedelic experiences suggest that feelings of group belonging and connectedness predict psychological well-being and positive mood (Forstmann et al., 2020; Kettner et al., 2021). Future group-based research would benefit from the inclusion of validated relational measures such as social connectedness (e.g. Lee and Robins, 1995) and cohesion (e.g. “communitas”; Kettner et al., 2021), and examinations of how these experiences may heighten psychological flexibility and reappraisal. An additional area of inquiry is the broader philosophical or spiritual changes that may accompany more flexible perspectives on self and world in these contexts (e.g. monistic idealism). Finally, given the unique ontological aspects of the Shipibo ceremony, it is important that Western researchers not abstract these practices from their historical and social contexts. Indeed, for thousands of years, the shaman has been working effectively in ways that do not map directly onto modern Western psychology (Adams et al., 2015). It would be foolish not to learn from them and instead foreclose on a limited view of these plant medicines.

These results must be considered against the backdrop of several limitations. The most notable limitations are the absence of a control group and the lack of strict eligibility criteria for participation in the ayahuasca retreats, which the naturalistic design of the study precluded. Therefore, it cannot be ascertained whether therapeutic improvements are directly attributable to participants’ ayahuasca ceremonies or other uncontrolled variables. In addition, while each of the participating retreat centers conducted medical and psychiatric screening of individuals prior to enrollment, these procedures were not standardized across sites. While this heterogeneity increases the study’s external validity, it limits its internal validity. Another important limitation concerns the unstandardized volume of the ayahuasca brew consumed by retreat participants, which the study team could not verify in this naturalistic study context. The concentrations of key active compounds found in ayahuasca (i.e. DMT, tetrahydroharmine, harmine, and harmaline) are notoriously difficult to quantify given the substantial heterogeneity of ayahuasca preparation methods, which renders rigorous standardization of future medicinal use of ayahuasca a key challenge (Kaasik et al., 2021). In addition, acute retreat factors were assessed at the same time point as the follow-up, thereby introducing the possibility of recall biases. The study employed a series of methodological safeguards, however, to decrease placebo, expectancy, and demand threats.

It also bears noting the psychometric shortcomings of our measures. The AEI Reappraisal scale is a novel measure that was piloted in the primary study (Weiss et al., 2021) and has not yet undergone comprehensive psychometric validation. In addition, due to an administrative error, two items were accidentally excised from the final version of the AAQ administered to participants. The AAQ was updated in large part due to concerns over its psychometric limitations, including suboptimal construct validity (Cherry et al., 2021). The revised version of the AAQ-II (Bond et al., 2011) improves upon the AAQ in its psychometric consistency and has been widely used in the literature. However, several researchers have pointed out that both the AAQ and AAQ-II may not discriminate from established measures of psychological distress (Rochefort et al., 2018; Tyndall et al., 2018; Wolgast, 2014). It is notable, nevertheless, that change in psychological flexibility did not statistically mediate relations between acute factors and negative affect, which would be expected if the constructs merely overlap. Future investigators interested in exploring psychological flexibility processes in relation to psychedelic experiences should employ more psychometrically rigorous measures such as the Personalized Psychological Flexibility Index (Kashdan et al., 2020) or Brief Experiential Avoidance Questionnaire (Gámez et al., 2014) to operationalize this construct. Next, it bears noting that our measures of affect were not global assessments; rather they were assessments of self-reported affect over the preceding week. Finally, while the study sample was relatively diverse with regard to educational backgrounds, participants were primarily White and middle-to-upper middle class. Ayahuasca retreats, which can cost up to several thousands of dollars, tend to remain financially inaccessible to those who cannot afford the costs of the retreats and associated travel (Adams et al., 2015). In psychedelic research studies, Black, Indigenous, and People of Color are vastly underrepresented even as the multigenerational effects of centuries of racialized policies burden them with high rates of trauma and other mental health sequelae. This presents stakeholders with an ethical imperative to prioritize providing opportunities to individuals from historically underrepresented communities to ensure generalizability and that those who could stand to benefit most are not excluded (George et al., 2020; Michaels et al., 2018).

This study contributes important insights to the growing literature on emotion regulation and psychological flexibility processes in psychedelic interventions. The results are suggestive that ayahuasca ceremonial experience produces therapeutic outcomes, in part, through promoting cognitive reappraisal, mystical experience, positive mood, and psychological clarity acutely, and psychological flexibility dispositionally. Our results provide further support for the assertion that classic psychedelics increase the malleability of self-related processing and reduce, in a potentially enduring way, the negative psychological impact of ruminative cognitive-affective repertoires. Additional time-ordered studies are required to establish whether increased psychological flexibility leads to improved outcomes. These insights support the recent burgeoning efforts to integrate mindfulness and third-wave approaches (e.g. ACT) with psychedelic therapy paradigms to enable a fruitful synergy between these approaches.

Figure 2.

Box and Violin plot illustrating change in negative affect between baseline and follow-up.