Reports of self-compassion and affect regulation in psilocybin-assisted therapy for alcohol use disorder: An interpretive phenomenological analysis

Alcohol use disorder (AUD) is highly prevalent and disabling, yet most people with AUD never access treatment and existing pharmacological and behavioural interventions often fail to produce durable recovery. Renewed interest in classic psychedelics as adjuncts to psychotherapy has been driven by historical trials with LSD and recent uncontrolled pilot work with psilocybin showing large within-group reductions in substance use; observational studies also report decreases in alcohol and drug use after naturalistic psychedelic experiences. Despite emerging clinical findings, psychological theories explaining how psychedelic-assisted therapy produces change remain underdeveloped, particularly regarding self-related processes such as self-compassion, mindfulness capacities, and emotion regulation that are plausibly relevant to addiction recovery. Agin-Liebes and colleagues set out to characterise psychological mechanisms of change attributed to a structured psilocybin-facilitated treatment for AUD by analysing semistructured interview narratives from a subset of participants enrolled in the first double-blind randomized controlled trial of psilocybin-assisted psychotherapy for AUD. The primary aim of this qualitative study was to illuminate participants' lived experiences before, during, and after psilocybin treatment, with special attention to affect regulation, self-compassion, and social connectedness as potential therapeutic processes supporting reductions in drinking.

This qualitative study drew from participants in a double-blind controlled parent trial (NCT02061293) in which participants received two 8-hour medication sessions one month apart, randomised to psilocybin (25 mg/70 kg, with the second dose titrated up to 40 mg/70 kg as indicated) or an active placebo (diphenhydramine 50 mg, titratable to 100 mg). After the 36-week double-blind phase, eligible participants were offered an open-label psilocybin session; the behavioural programme accompanying medication included motivational enhancement therapy and cognitive behavioural therapy with standardised preparation and integration around medication sessions. Eligibility for the parent trial required age ≥25, DSM-IV-TR criteria for alcohol dependence with a desire to stop or reduce drinking, and exclusion of medical or psychiatric conditions that would jeopardise safety (for example, seizure disorder, significant liver impairment, schizophrenia, bipolar disorder, current major depressive episode, current PTSD, current suicidality or recent medically serious suicide attempt, and recent other substance dependence). For the qualitative substudy, the investigators selected the 14 most recently completed participants (no more than 18 months since their final drug administration); 13 agreed to participate. All 13 had completed three medication sessions overall: six had received psilocybin in both double-blind sessions and again in the open-label session, while seven had received diphenhydramine in the double-blind phase and psilocybin in the open-label session. Interviews took place between two and 18 months after the third medication session. Data collection used a semistructured interview guide probing participants' drinking histories, coping with strong emotions, experiences of preparation, medication sessions, debriefing, therapist relationships, and hypothesised mechanisms; questions about stigma were collected but excluded from the present analysis. Two study members conducted all interviews. Transcripts were coded in Dedoose and analysed using interpretive phenomenological analysis (IPA), a method that emphasises fine-grained accounts of lived experience and is recommended for small samples (generally ≤15). Coding proceeded after pooled interrater agreement (reported as pooled κ) exceeded 80, and each transcript was reviewed by at least two team members. The analysis team comprised licensed clinicians who disclosed social constructivist and psychodynamic/attachment-oriented positionalities and had training in psychedelic facilitation; the team noted these perspectives as influencing interpretation.

Thirteen participants (age range 28–63, mean 48.0 years, SD 11.9) comprised the qualitative sample. Most identified as non-Hispanic White (92%), there were approximately equal proportions of cisgender men and women, and the median household income reported for the subsample was higher than in the parent study ($145,000 v. $100,000). At screening participants reported drinking on about 81% of days in the prior month; by the open-label Week 38 time point average drinking indices had fallen markedly at the group level (for example, percent heavy drinking days declined from roughly 55.8%–57.9% at screening to about 4.5%–7.6% at Week 38; drinking days declined from about 77%–84.4% to 18.6%–19%; drinks per day declined from 4.9–5.9 to 0.5–0.9). Five participants reported having read Michael Pollan's How to Change Your Mind prior to participation. Thematic analysis organised findings into temporal categories: before, during, and after psilocybin treatment. Prior to the intervention, participants described entrenched patterns of emotional avoidance and ‘‘autopilot’’ coping in which alcohol served to numb distress, social anxiety, or other difficult feelings. Many reported limited emotional vocabulary and reliance on substances for short-term relief. Twelve participants described pervasive shame, self-blame, and internalised self-criticism; several linked these inner critical voices to early caregiver messages. A majority (8) described alcohol as a surrogate for connection, noting disconnection from self and others and modelling of parental alcohol misuse. During psilocybin sessions, most participants (11) reported intense affective experiences characterised by spacious awareness, emotional catharsis, and the capacity to perceive personal history and behaviour with new clarity. Many described the sessions as lowering psychological barriers and permitting processing of painful material related to drinking. Twelve participants reported accessing feelings of security, reprieve from shame, and a ‘‘self-compassion template’’—states in which they could separate harsh internal voices from more compassionate self-views and sometimes reframe relationships with loved ones. Participants gave concrete metaphors for self-compassion, including calming imagined cravings (for example, picturing a small ‘‘alcohol weasel’’ and soothing it). Eight participants described spiritual or existential insights and a profound sense of interconnectedness that influenced how they understood their life narrative and the role of alcohol within it. After the sessions, participants frequently reported greater mindfulness and improved affect regulation. Several described an ability to observe and tolerate negative emotions without immediately resorting to drinking, using the acute session experiences as a template for coping. Eight participants reported notable reductions in cravings and alcohol use, though some described relapse or difficulty maintaining gains when faced with stressors; one participant explicitly stated the treatment was not a ‘‘magic pill.’' Interpersonal changes were also described: seven participants reported improved relationships, increased openness to support, and greater patience and compassion toward others. Participants emphasised the importance of the therapeutic setting and therapist support in creating safety during sessions and reinforcing integration afterwards. Where available, numerical data from the parent study time points were reported in interview contexts to indicate substantial group-level reductions in drinking by Week 38, and participants’ narrative reports generally aligned with these quantitative trends. The sample heterogeneity in psilocybin exposure (three psilocybin sessions for six participants v. one psilocybin session for seven participants) was reported as a possible source of variability in outcomes.

Agin-Liebes and colleagues interpret these qualitative narratives as supporting the proposition that psilocybin-assisted therapy can increase the malleability of self-related processing, reduce shame-based and self-critical thought patterns, and foster self-compassion, connectedness, and improved affect regulation among people with AUD. Participants characterised acute experiences of spacious awareness and emotional catharsis as providing a ‘‘template’’ for later, more adaptive responses to stress, cravings, and negative affect. The authors situate these findings alongside preliminary quantitative work showing increased self-compassion after psychedelic use and with theoretical models such as self‑entropic broadening and frameworks proposing that psychedelics loosen rigid protective patterns, enabling revision of entrenched cognitive–affective schemas. The investigators also emphasise that psilocybin-induced neuroplasticity coupled with enhanced feelings of connectedness and self-understanding may have supported the learning of more accepting, self-compassionate responses to distress, thereby reducing reliance on alcohol. They highlight the reinforcing role of preparatory and integration psychotherapy, the therapeutic alliance, and community or peer supports in consolidating gains. The authors caution that the intervention was not uniformly curative: some participants experienced lingering cravings or relapse when exposed to stress, and several reported difficulties integrating the experience. They therefore argue that structured aftercare and psychosocial supports are important for sustaining change. Key limitations acknowledged by the authors include variable timing of interviews (2–18 months post-session), heterogeneity in psilocybin exposure across participants due to the parent trial design, likely self-selection bias (including influence from popular accounts of psychedelics), and the analytic team's own positionalities potentially shaping interpretation. The sample was also demographically homogeneous and relatively high in socioeconomic status, limiting generalisability—particularly to groups historically underrepresented in psychedelic research. The authors recommend further controlled studies to test whether increases in self-compassion and affect regulation mediate improved outcomes in AUD, suggest assessing adult attachment measures pre/post intervention to examine effects on attachment security, and propose integrating self-compassion training into psychedelic-assisted therapies as a future avenue for research and clinical practice.