Therapeutic Alliance and Rapport Modulate Responses to Psilocybin Assisted Therapy for Depression

This trial (n=59) assessed the relationships between therapeutic alliance and rapport, the quality of the acute psychedelic experience and treatment outcomes in depressed patients. Following psilocybin administration at a second session, the therapeutic alliance had a direct impact on final depression scores, not mediated by the acute experience, with a weaker alliance ahead of the second psilocybin session predicting higher absolute depression scores at the endpoint.


Background: Across psychotherapeutic frameworks, the strength of the therapeutic alliance has been found to correlate with treatment outcomes; however, its role has never been formally assessed in a trial of psychedelic-assisted therapy. We aimed to investigate the relationships between therapeutic alliance and rapport, the quality of the acute psychedelic experience and treatment outcomes.

Methods: This 2-arm double-blind randomized controlled trial compared escitalopram with psychedelic-assisted therapy for moderate-severe depressive disorder (N = 59). This analysis focused on the psilocybin condition (n = 30), who received two oral doses of 25 mg psilocybin, 3-weeks apart, with psychological preparation, in-session support, and integration therapy. A new psychedelic therapy model, called “Accept-Connect-Embody” (ACE), was developed in this trial. The primary outcome was depression severity 6 weeks post-treatment (Quick Inventory of Depressive Symptomatology, QIDS-SR-16). Path analyses tested the hypothesis that therapeutic alliance (Scale To Assess the Therapeutic Relationship Patient Version, STAR-P) would predict depression outcomes via its influence on the acute psychedelic experience, specifically emotional-breakthrough (EBI) and mystical-type experiences (MEQ). The same analysis was performed on the escitalopram arm to test specificity.

Results: The strength of therapeutic alliance predicted pre-session rapport, greater emotional-breakthrough and mystical-type experience (maximum EBI and MEQ scores across the two psilocybin sessions) and final QIDS scores (β = −0.22, R2 = 0.42 for EBIMax; β = −0.19, R2 = 0.32 for MEQMax). Exploratory path models revealed that final depression outcomes were more strongly affected by emotional breakthrough during the first, and mystical experience during the second session. Emotional breakthrough, but not a mystical experience, during the first session, had a positive effect on therapeutic alliance ahead of the second session (β = 0.79, p < 0.0001). Therapeutic alliance ahead of the second session had a direct impact on final depression scores, not mediated by the acute experience, with a weaker alliance ahead of the second psilocybin session predicting higher absolute depression scores at endpoint (β = −0.49, p < 0.001)

Discussion: Future research could consider therapist training and characteristics; specific participant factors, e.g., attachment style or interpersonal trauma, which may underlie the quality of the therapeutic relationship, the psychedelic experience and clinical outcomes; and consider how therapeutic approaches might adapt in cases of a weaker therapeutic alliance.”

Authors: Roberta Murphy, Hannes Kettner, Rick Zeifman, Bruna Giribaldi, Laura Kartner, Johnny Martell, Tim Read, Ashleigh Murphy-Beiner, Michelle Baker-Jones, David J. Nutt, David Erritzoe, Rosalind Watts & Robin L. Carhart-Harris



Psychedelic-assisted therapy is a mental health intervention that involves the administration of a psychedelic substance in combination with therapy or psychological support. It was widely researched in the 1950s and 60s, but was shut-down due to restrictive governmental regulations. In a double-blind randomised controlled trial, psilocybin-assisted therapy was found to be as effective as escitalopram at improving mental health, but more rapid and wide-ranging improvements were found in the psilocybin arm.

Research suggests that the quality of the acute psychedelic experience is predictive of the therapeutic effects of psychedelic-assisted therapy.

Several studies have provided evidence that the mystical-type experience is a predictor of positive treatment outcomes following psychedelic-assisted therapy. Recent prospective research and published accounts of trial participant’s psychedelic therapy experiences have highlighted the importance of emotional breakthrough as a therapeutic mechanism, but this role has not yet been evaluated within a clinical sample or in the context of a controlled psychedelic-assisted therapy clinical trial.

Empirical research on psychedelic-assisted therapy has focused on the acute psychedelic experience as a mediator of treatment outcomes, with less focus on other important components of psychedelic-assisted therapy.

The relationship between the individual receiving psychedelic-therapy and the treatment provider(s) is a key contextual component in this regard. The terms “participant” and “patient” are often used interchangeably depending on the specific context in psychedelic literature. Psychedelic-assisted therapy involves a supportive, open and non-directive approach from a therapist or guide, who is ideally with personal and/or professional experience with the effects of psychedelic compounds or other non-ordinary states.

The therapeutic relationship is recognised to be of fundamental importance across different psychotherapeutic models or approaches, and is influenced consciously and unconsciously by both therapist and patient. The therapeutic alliance is also recognised to be a dynamic phenomenon that is sensitive to change over the course of therapy.

The therapeutic relationship is the emotional bond between the therapist and the patient, which is influenced by pre-treatment patient characteristics and therapist characteristics and qualities.

Despite strong theoretical assumptions and strong evidence for the importance of the therapeutic relationship within psychotherapeutic interventions, little empirical research has been done to examine its role in shaping psychedelic experiences and psychedelic-assisted therapy treatment outcomes.

Although many qualitative reports have suggested that psychedelic-assisted therapy may intensify the therapeutic relationship, there has been no published quantitative research into whether the quality of the therapeutic relationship is predictive of the quality of the acute psychedelic experience.

The present article uses data from a recently published DB-RCT to examine the role of therapeutic alliance and rapport in mediating the quality of the psychedelic experience and subsequent key clinical outcomes.

Examine whether the quality of the acute experience (mystical-type experiences and emotional breakthrough) mediates the relationship between the therapeutic relationship and changes in depression.

The relationship between the therapeutic relationship, the acute experience and depression score was explored on a session-by-session basis.

Trial Oversight

This study received approval from several regulatory agencies and was licensed by the Home Office to prescribe psilocybin. It took place at the National Institute for Health Research-funded Imperial Clinical Research Facility.


This study recruited men and women aged 18 – 65 years old via the NIHR Clinical Research Network and online. They were required to have moderate-to-severe MDD.

Participants had to build a therapeutic alliance and rapport relatively quickly with their therapists, as this would be necessary for them to be able to speak openly about their difficulties, be emotionally vulnerable, and understand their own relational patterns.

Design and Procedures

In a phase II, double-blind, two-armed, randomised controlled trial, participants with moderate-severe MDD were randomised either to an ‘escitalopram arm’ or to a ‘psilocybin arm’.

Each participant was assigned two guides, one of whom was the main guide. The other guide answered questions and discussed the study with the participant 1 week prior to the first psilocybin dosing session.

In visit 1, participants were given 5 hours of preparation by their guides for the next day’s psilocybin session, in visit 2, they were given 25 mg of psilocybin, and in visit 3, they were given a 3-week course of placebo capsules and 10 mg of escitalopram. Three weeks after the first psilocybin session, patients had a second 25 mg or 1 mg psilocybin dosing day, followed by a next day integration session.

Three weeks after their second integration session, patients returned for a final follow-up visit with their guides. They discussed their experience of the trial and their current mood state.

Therapy Protocol

A psychedelic-specific therapy model, informed, in part, by Acceptance and Commitment Therapy (ACT), emerged during the trial, and is referred to as The ACE model. It emphasises building trust between participants and their guides, engaging with and accepting difficult emotions, connecting to personal meaning, values, self and others, and giving mindful awareness to embodied or somatic experiences.

The preparation session focused on building trust and rapport, setting intentions, and sharing information about psychedelic experiences. Discussions were had around physical contact, and visualisation exercises were used to support processes of attuning to their bodies.

Participants wore eye masks and earphones and listened to a pre-selected playlist during the psilocybin sessions. The sessions were safe and contained and the guide provided compassionate support and recognition of any psychological insights or changes.


The 16 item self-rated Quick Inventory of Depressive Symptomatology (QIDS) was used to assess changes in depression severity.

The therapeutic alliance was assessed using the STAR-P and a self-constructed single-item measure of rapport was used to assess the extent of rapport in the hours immediately before the psilocybin session.

Participants completed a questionnaire after each dose of psychedelics to assess mystical-type peak experiences and cathartic release of difficult emotions.

Statistical Analyses

The primary hypothesis was tested by longitudinal path modelling, which found that therapeutic alliance affected treatment outcomes by modulating the intensity of acute subjective effects.

To simplify the analysis across two separate psilocybin sessions, the highest MEQ and EBI score across the two sessions was used to predict depression outcomes.

To investigate session-specific mechanisms of therapeutic alliance and the acute psychedelic experience, additional near-saturated sequential mediation analyses were conducted for the first and second psilocybin session. These analyses included therapeutic alliance, rapport, MEQ and EBI as covariates, and depression severity as an outcome.

All path analyses used maximum likelihood estimation, and multiple indicators of model fit were reported, including model Chi-Square, Comparative Fit Index (CFI) and Root Mean Square Error of Approximation (RMSEA). Thresholds for acceptable and good fit were taken from the extant literature.

Demographic Information

Approximately 1,000 individuals were screened, 103 were invited to an in-person screening visit, 59 were enrolled onto the trial, and 30 were randomised into the psilocybin arm and 29 into the escitalopram arm. One participant was excluded from the analysis because they did not comply with required restrictions on drug-taking.

Primary Confirmatory Analysis: Therapeutic Alliance Predicts Outcomes via Acute Subjective Effects

Both the EBI and MEQ significantly predicted depression severity at the key endpoint, but the EBI had a slightly larger effect than the MEQ, suggesting that emotional breakthrough was a more reliable predictor of depression improvements in this trial.

Secondary Exploratory Analyses: Psilocybin Session-specific Effects

The effect of therapeutic alliance on depression severity was significantly mediated by the EBI, but not by the MEQ. The MEQ-based model explained 14% of variance in STAR-P scores, which was approximately equal to the amount of variance accounted for by baseline STAR-P scores.

Results show that only MEQ scores significantly affected final depression outcomes, while therapeutic alliance and pre-session rapport had a positive effect. The correlation between depression severity at the intermediate and the final endpoints was very high, thus accounting for 0.742 = 55% of variance.

The study found that depression severity at the intermediate timepoint was associated with worse therapeutic alliance and rapport ahead of the second session.

Supplementary Result

The supplementary material includes path models for the escitalopram arm, histograms comparing acute scores between the escitalopram and psilocybin arms, demographics, and correlations between all measures.


The importance of the therapeutic relationship in psychedelic-assisted therapy has long been treated as conventional wisdom, but has never been empirically tested. Here, we demonstrate how a strong therapeutic alliance predicts pre-session rapport, which predicts greater emotional breakthrough and “mystical type” experience towards improved clinical outcomes.

As the field of psychedelic-assisted therapy develops, it is important to test assumptions and consider how the nature and quantity of therapy might need to adapt to more complex clinical populations, such as individuals with emotionally unstable or borderline personality disorder.

Our stage-by-stage analysis suggests that emotional breakthrough in the first psilocybin session is the most potent predictor of subsequent improvements in depressive symptoms, and this relationship is mediated by pre-session therapeutic alliance and rapport.

The model results indicate that the therapeutic alliance changes somewhat ahead of the second psilocybin session, and that a weak therapeutic alliance at this key treatment midpoint translates into poor depressive outcomes at the end of the trial.

The Therapeutic Relationship

In the present study, the therapeutic alliance was either stronger or similar to that observed in previous clinical samples. This may be due to the promissory culture surrounding psychedelic-assisted therapy at present, or to the double-blind and randomised design of the study.

Preparation work, specific patient factors, and early relational experiences might all influence the experience of psilocybin therapy. Attachment style and internalised early relational experiences might also influence the experience of psilocybin therapy.

A strong therapeutic alliance and rapport prior to the first psilocybin session resulted in more emotional breakthrough, better alliance and rapport before the second session, and better eventual clinical outcomes.

After a first psychedelic-assisted therapy session, the therapeutic relationship has a greater impact on the final outcome of the therapy, independent of the nature of the acute experience within the second psilocybin therapy session.

A strong therapeutic alliance appears to support participants towards more powerful emotional breakthroughs and mystical experiences, but the same was not true for mystical-type experiences linked to session one.

Psychedelic-assisted therapy can involve both explicit verbal processes and non-verbal experiential processes. These processes may be particularly relevant to psychedelic-assisted therapy, with the neuroplastic brain state associated with psychedelics serving to catalyse deep processes of change.

In a study where 40% of participants failed to achieve remission, participants who reported weaker therapeutic alliance and rapport experienced less therapeutic breakthrough during sessions. This suggests that early warning signs of poor response, such as poor initial therapeutic alliance, might signal a postponement of subsequent sessions.

Here we have shown that a weaker therapeutic alliance leads to a less emotional breakthrough and mystical experience during psilocybin sessions.

Participants who experience more challenging therapeutic processes may need a different therapeutic approach, which may entail a longer preparation period, greater number of psilocybin-therapy sessions and/or more integration therapy. However, this may be more challenging in the context of tightly controlled and constrained short-term clinical trials.

The Acute Experience

Our results demonstrate that depression severity scores were significantly reduced in relation to the intensity of both emotional breakthrough and mystical type experiences, and that this effect was greater for maximum EBI.

Emotional breakthrough may be a more intuitive and familiar construct as a therapeutic mechanism, but the secularisation of psychedelic treatment spaces fails to address the complex role of therapists’ ontological assumptions in shaping their patients’ experiences.

The Grofian model, devised by psychiatrist Stanislav Grof, suggests that people first “work through” personal, biographical material prior to moving to the so-called transpersonal realm. Higher baseline depression was associated with less mystical experience but had no impact on emotional breakthrough.


The path models for the escitalopram arm differed from those of the psilocybin arm in most respects, supporting the specificity (to psychedelic assisted-therapy) of the main models presented above.

Strengths and Limitations

This study provides quantitative evidence that the therapeutic relationship plays an important role in shaping the acute psychedelic experience and subsequent treatment outcomes.

The current study was able to inform on causal relations in the treatment process by implementing sequential mediation analyses on variables that were measured across successive time points. However, the small sample size available for the present analyses is an important limitation.

This trial was conducted with a specific population of individuals with current major depressive disorder. It is important to note that the demographic in this trial was one typical of clinical research. Lack of diversity and minority exclusion in psychedelic research is a recurrent issue that needs to be addressed. PPI has been used to address this issue, and future research should consider diversifying clinical teams, specific culturally-sensitive and targeted recruitment strategies, and appropriate training.

Although the correlation between therapeutic alliance prior to the first psilocybin session and the acute experience of that session was low, it is worth considering that pre-session rapport may be a more informative predictor of the nature of session one (and thus, longer-term outcomes) than an earlier measure of therapeutic alliance.

The primary endpoint of this trial was at 6 weeks, so the results are only relevant to this relatively brief window of time. Longer term follow-up assessment results may change the findings.

The present work did not assess countertransference or clinician rated therapeutic alliance, but future research might consider including relevant assessments to compile a more complete, bidirectional picture of the therapeutic relationship.

Clinical Implications and Future Directions

These findings highlight the importance of the therapeutic relationship in the psychedelic-assisted therapy treatment process. The experience of the guide and the flexibility of the clinical protocol should be carefully considered when working with participants who may find it more challenging to build a strong therapeutic alliance and rapport.

In the United Kingdom, traditional psychotherapies are delivered both privately and in the NHS. Psychedelic-assisted therapy may be used as an occasional catalyst to resolve therapeutic impasses, and further research is needed to explore how much preparation, how many dosing sessions, how far apart, and with how much integration are important.

It has been suggested that the field of psychotherapy should move its focus away from outcome to process-outcome research, and that qualitative data and case reports may do well to focus on non-responders in order to better understand how to improve treatment for these individuals going forward.

It is worth considering that standardised measures and fixed time frames do not capture the complexity of psychological change and the therapeutic process, and that more flexible treatment protocols with longer-term follow-up and support are needed.


In this paper, we have found evidence that the quality of the psychedelic experience was associated with the quality of the therapeutic alliance. A better therapeutic alliance predicted lower eventual depression scores.


The studies involving human participants were reviewed and approved by several regulatory agencies, and the patients provided their written informed consent to participate.


RM directed the project, HK designed the statistical analyses, BG oversaw the trial, RM wrote the manuscript, RCH supervised the writing of the paper, BG, JM, MB-J, AM-B, RW and RM worked as the clinical team, TR provided group therapy supervision, and DE, DN, and RC-H provided feedback on the manuscript.


This paper is based on the work of many people including volunteers, guides, medical professionals, students, and others involved in the field of psychedelic therapy.


Authors associated with this publication with profiles on Blossom

David Nutt
David John Nutt is a great advocate for looking at drugs and their harm objectively and scientifically. This got him dismissed as ACMD (Advisory Council on the Misuse of Drugs) chairman.

Rosalind Watts
Rosalind Watts is a clinical psychologist and clinical lead at the Psychedelics Research Group at Imperial College London. She is also known for developing the 'Accept, Connect, Embody' psychedelic therapy model.

Robin Carhart-Harris
Dr. Robin Carhart-Harris is the Founding Director of the Neuroscape Psychedelics Division at UCSF. Previously he led the Psychedelic group at Imperial College London.

David Erritzoe
David Erritzoe is the clinical director of the Centre for Psychedelic Research at Imperial College London. His work focuses on brain imaging (PET/(f)MRI).

Michelle Baker Jones
Michelle Baker Jones is an integrative psychotherapeutic counselor and lead guide in the ongoing Imperial College London psychedelic research.

Tim Read
Tim is a psychiatrist and psychotherapist based in London. He is the Lead Supervisor at Synthesis and has had training in holotropic breathwork from Stan Grof.

Linked Research Papers

Notable research papers that build on or are influenced by this paper

Trial of Psilocybin versus Escitalopram for Depression
This double-blind placebo-controlled study (n=59) compared psilocybin (2x25mg; 3 weeks apart) to escitalopram (SSRI) over a six-week period and found large improvements in depression scores for those suffering from depression (MDD) in both groups. On the main measure of depression, the QIDS-SR-16, there was no significant difference between both groups. The study did find significant differences, favoring psilocybin, on the HAM-D-17, MADRS, avoidance, flourishing, wellbeing, and suicidality.

PDF of Therapeutic Alliance and Rapport Modulate Responses to Psilocybin Assisted Therapy for Depression