Naturalistic psilocybin use is associated with persisting improvements in mental health and wellbeing: results from a prospective, longitudinal survey

Psilocybin is a naturally occurring serotonergic psychedelic found in many Psilocybe species and has a long history of ritual use. In recent decades, small clinical trials and pilot studies have suggested potential therapeutic benefits for mood, existential distress, and substance use disorders, and research in healthy volunteers has indicated possible improvements in psychological wellbeing and spiritual experiences. Concurrent cultural and policy shifts have led to increased naturalistic (non-research) use of psilocybin, but evidence on the public health impact of that use outside controlled settings remains limited. Nayak and colleagues conducted a prospective, longitudinal online survey study to add empirical data on contemporary naturalistic psilocybin use. The study aimed to characterise users and their dosing and setting, to assess prospective changes in mental health, personality, wellbeing, and functioning from before to after a planned psilocybin session, and to examine how pre-session mindset (e.g., absorption, State of Surrender, adverse childhood experiences) and elements of setting (e.g., presence of a sitter) related to acute subjective effects and persisting outcomes.

This was a prospective longitudinal survey of English-speaking adults (≥18 years) who planned to take psilocybin outside clinical research. Recruitment occurred online between July 22, 2020 and July 14, 2022, and participants completed six automated web surveys: initial consent/demographics, ~2 weeks pre-session (baseline battery), 1 day pre-session, 1–3 days post-session, 2–4 weeks post-session, and 2–3 months post-session. The study received Institutional Review Board approval at Johns Hopkins University School of Medicine. Baseline-only measures included lifetime drug-use history, the Tellegen Absorption Scale (TAS, 0–34), and the Adverse Childhood Experience scale (ACE, 0–14). Repeated longitudinal measures administered at baseline and both follow-ups comprised a modified Beck Depression Inventory-II (BDI-II, 0–60; suicidality item removed), the Alcohol Use Disorders Identification Test (AUDIT, 0–40), the Short State-Trait Anxiety Inventory (STAI, state and trait, each 10–40), the Emotion Regulation Questionnaire (ERQ; cognitive reappraisal and expressive suppression), the Cognitive Flexibility Scale (CFS), the PROMIS-GH physical health subscale, the FACIT-Sp spiritual wellbeing scale, the Copenhagen Burnout Inventory (CBI), and the Big Five Inventory (BFI, at baseline and 2–3 months). The 1-day-pre and 1–3-day-post surveys collected planned and actual dose/form, concurrent substances, setting details (alone, with friends, with a sitter/guide), intentions, the State of Surrender (SOS) scale pre-session, and acute subjective measures post-session: the Mystical Experience Questionnaire (MEQ30) and the Challenging Experience Questionnaire (CEQ). Data analysis used linear mixed-effects models (lme4 in R) for eleven longitudinal outcomes, including baseline score and covariates (timepoint, age, sex, race, education, presence of a sitter, psilocybin dose in dried grams, prior psychedelic use, ACE, MEQ30, CEQ, SOS, TAS) and random intercepts for participants. MEQ30, CEQ, SOS, and TAS were allowed to interact with time. Five simple linear models were used for the five BFI dimensions. Correlations between ACE/TAS/SOS and MEQ30/CEQ were calculated. To control for multiple testing across 22 primary tests, a Šidák-corrected alpha of 0.0023 was applied. Two exploratory t-tests compared acute subjective effects between those reporting concurrent antidepressant use and those not; these were not included in the Šidák correction.

Survey retention declined across timepoints: N = 8,006 at consent, N = 2,833 at ~2 weeks pre-session, N = 1,551 at 1–3 days post-session, N = 1,182 at 2–4 weeks, and N = 657 at 2–3 months. The sample was majority White (81–87%), male (54–59%), US-resident (73–83%), college-educated (54–66% with bachelor’s or higher), mean age ~40 years, and highly experienced with psychedelics (mean ~16–17 prior uses). Most participants reported intentions of self-exploration (81%), mental health (71%), or therapeutic aims (48%); about 74% set a specific intention. Sessions typically occurred at home (≈70%), usually alone (43%) or with friends also using (26%); 16% had a sober sitter. The predominant reported form was dried mushrooms (≈42% whole, 19% ground), average initial dose among those reporting grams (excluding outliers) was 3.1 g (SD 2.3). Common co-used substances were cannabis (31%), caffeine (15%), and alcohol (12%); 4.6% reported concurrent antidepressant use. Acute subjective effects: mean MEQ30 total score was 0.50 (SD 0.25), with 21.6% meeting criteria for a "complete mystical experience". Mean CEQ was 0.14 (SD 0.14). MEQ30 scores predicted greater reductions in depression, personal and work burnout, and state anxiety, and greater increases in cognitive flexibility and spiritual wellbeing. CEQ did not predict linear changes in longitudinal outcomes. TAS correlated weakly with MEQ30 (r = 0.32, p < 0.001) but not CEQ; SOS correlated weakly with MEQ30 (r = 0.24, p < 0.001) and very weakly with CEQ (r = -0.10, p < 0.001). ACE scores were not significantly correlated with MEQ30 or CEQ. Adverse events and persisting negative effects were uncommon: 0.4% sought medical care and 3.4% sought psychological care within 1–3 days after the session. Persisting negative effects were reported by 7–11% across follow-ups, most commonly mood fluctuations (4.7% at 2–4 weeks, 3.0% at 2–3 months) and depressive thoughts (3.1% at 2–4 weeks, 1.7% at 2–3 months). Primary longitudinal outcomes showed statistically significant improvements from baseline to follow-ups after covariate adjustment. Modified BDI-II mean scores fell from 15.3 (SD 11.2) at baseline to 6.4 (8.3) at 2–4 weeks and 7.5 (8.9) at 2–3 months. Covariate-adjusted standardised mean differences (SMD; interpretable as adjusted Cohen’s d) for BDI were -0.71 [95% CI -0.79, -0.63] at 2–4 weeks and -0.58 [-0.66, -0.50] at 2–3 months (both p < 0.001), indicating moderately large reductions in depressive symptoms. AUDIT scores decreased slightly (SMD -0.06 [-0.10, -0.02] at 2–4 weeks and -0.09 [-0.13, -0.05] at 2–3 months; p < 0.001), reflecting small reductions in alcohol misuse. State and trait anxiety showed reductions (details partially fragmented in extraction but reported as decreased). Cognitive reappraisal (ERQ) increased with SMD 0.19 [0.11, 0.27] at 2–4 weeks and 0.24 [0.16, 0.32] at 2–3 months (p < 0.001), while expressive suppression showed no consistent change. Cognitive flexibility (CFS) increased with SMD 0.23 [0.15, 0.31] at 2–4 weeks and 0.22 [0.15, 0.30] at 2–3 months (p < 0.001). Personal burnout decreased (CBI personal SMD -0.33 [-0.37, -0.29] at 2–4 weeks and -0.36 [-0.42, -0.31] at 2–3 months; p < 0.001); work-related burnout decreased significantly only at 2–3 months (SMD -0.19 [-0.25, -0.12], p < 0.001). No significant change was observed in self-reported physical health (PROMIS-GH). Personality measures showed small but significant increases in extraversion (from 3.12 to 3.17, p < 0.001) and decreases in neuroticism (from 2.88 to 2.81, p < 0.001); openness, agreeableness, and conscientiousness did not change significantly. Set and setting variables: prior psychedelic experience and presence of a sitter/guide did not significantly predict longitudinal outcomes. Higher MEQ30 scores were associated with greater improvements on several outcomes. In exploratory analyses, concurrent antidepressant use (n = 72) was associated with lower acute MEQ30 (mean 0.40 vs 0.51, p = 0.0007) and CEQ (0.10 vs 0.15, p = 0.001) scores, indicating attenuated acute subjective effects among this subgroup. At follow-ups, 29–36% rated the session among the 10 most personally meaningful or spiritually significant experiences of their lives, and most respondents (84–89%) reported positive changes in wellbeing and life satisfaction attributable to the experience. Approximately half reported improved relationships and substantial proportions reported other positive behavioural changes (increased exercise, improved work life, better diet).

Nayak and colleagues interpret these results as supportive of persisting improvements in mental health, wellbeing, and psychological functioning following naturalistic psilocybin use in this convenience sample. The longitudinal reductions in depressive symptoms, anxiety, and alcohol misuse, together with increases in cognitive flexibility, cognitive reappraisal, spiritual wellbeing, and reductions in burnout, are presented as broadly consistent with prior clinical and epidemiological findings. The authors note the moderately large adjusted effect sizes for depression (SMD ≈ 0.58–0.71) and emphasise that mystical-type acute experiences (MEQ30) were associated with greater persisting improvements across several domains. The discussion contrasts acute subjective effects observed here with those from laboratory studies, noting lower proportions of "complete mystical experiences" and lower mean MEQ30 scores in this naturalistic sample, which likely relates to lower average doses and heterogeneous contexts in non-research settings. The authors highlight that challenging acute experiences were low on average and did not predict longitudinal outcomes, suggesting that while such challenges may pose acute risk, they did not on average predict therapeutic benefit in this dataset. The exploratory finding of attenuated acute subjective effects among participants concurrently taking antidepressants is noted and compared with mixed evidence from small experimental studies; the authors call for targeted research on interactions between psilocybin and psychiatric medications. Several limitations acknowledged by the investigators temper interpretation. These include a self-selected, predominantly White and well-educated sample that limits generalisability; anonymous, self-report data that cannot be externally verified; attrition across follow-ups that may bias results; possible confounding from concurrent substance use; and the fact that many participants took additional psilocybin doses after the reference experience (26% at 2–4 weeks and 50% at 2–3 months), complicating attribution of persisting effects to a single session. The authors also note the small absolute rates of acute medically significant reactions but caution that individuals with severe adverse events may have been less likely to complete follow-ups. Finally, the investigators suggest set and setting variables still play an important role in acute and persisting effects and that therapeutic contexts may leverage these factors to boost outcomes beyond pharmacological effects alone. Overall, the authors conclude that their large prospective dataset aligns with a growing body of evidence indicating potential therapeutic and wellbeing-related benefits of psilocybin, while underscoring the need for further targeted, controlled, and more demographically diverse research to clarify causal mechanisms, optimal dosing regimens, and safety in the context of concurrent medications and varied real-world settings.

In spite of the study’s limitations, the authors conclude that findings from this large prospective naturalistic cohort are consistent with prior clinical and epidemiological work, indicating that psilocybin use in non-research settings was associated with lasting improvements in several domains of mental health and wellbeing. They present these results as supportive of psilocybin’s broad therapeutic potential while noting the need for further rigorous research to address remaining questions about safety, optimal use conditions, medication interactions, and generalisability.