Effects of psychedelics in older adults: A prospective cohort study

Emotional functioning and social connectedness change across the adult life span, with older adults (OA) often prioritising emotional regulation and reporting more positive affect than younger people. Nevertheless, affective symptoms—such as anxiety, low mood, loneliness and apathy—remain common and have important consequences for cognition and mental health in later life, including increased risk for late-life depression and dementia. Conventional antidepressants have shown reduced efficacy and higher rates of some adverse effects in older populations, motivating interest in alternative treatments. Serotonergic psychedelics (for example psilocybin, LSD, DMT/ayahuasca and mescaline) have emerged in recent clinical research as candidate interventions for affective disorders; their acute subjective effects (e.g. ego-dissolution, mystical-type experiences, emotional breakthrough) have often been proposed as mediators of therapeutic benefit. However, older adults are markedly underrepresented in psychedelic trials, leaving safety, efficacy and mechanisms of action in this age group largely unknown. L. and colleagues set out to examine whether guided psychedelic group sessions are associated with changes in mental well-being among older adults, and whether the psychological mechanisms that predict any changes differ from those observed in younger adults (YA). Using a prospective observational cohort design in people attending naturalistic guided group psychedelic ceremonies or retreats, the study aimed to (1) measure longitudinal changes in well-being up to six months post-session in OA, (2) compare those changes with a matched sample of younger adults, and (3) investigate baseline, acute and post-acute predictors of well-being change, with particular attention to relational experiences in group settings.

The study used a prospective cohort design recruiting a convenience sample of people planning to attend organised psychedelic retreats or group ceremonies. Recruitment was conducted online (social media, forums, email newsletters) and via retreat facilitators. Eligible participants were aged ≥18, able to read English, and intending to use a listed psychedelic (psilocybin/magic mushrooms/truffles, LSD, ayahuasca/DMT/5-MeO-DMT, mescaline, or iboga/ibogaine). Participants provided informed consent and completed web-based surveys at multiple time points: within two weeks before the session (baseline), 24 hours after the session (acute effects), the day after leaving the retreat (post-retreat measures), and at two weeks, four weeks and six months post-session (outcomes). The study was approved by Imperial College London research ethics bodies. Older adults were defined as participants reporting age ≥60. From the pool of participants who completed baseline and the 24-hour survey, 62 OA were identified and a matched sample of 62 younger adults was selected using nearest-neighbour matching (MatchIt in R). Matching variables included gender, education, psychiatric history, previous psychedelic use, baseline well-being and reported drug dose, with matching intended to reduce confounding in subsequent parametric analyses. Primary outcome was mental well-being measured by the Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS), assessed at baseline and at two weeks, four weeks and six months. Acute subjective measures collected 24 hours post-session included the Ego-Dissolution Inventory (EDI), the Mystical Experience Questionnaire (MEQ), the Challenging Experience Questionnaire (CEQ), the Emotional Breakthrough Inventory (EBI), and the Communitas Scale (COMS, assessing acute relational togetherness). Post-acute mediators assessed the day after leaving the retreat included the Psychological Insight Scale (PIS) and a post-retreat version of Communitas (COMS-PR) evaluating relational experiences across the retreat. For analysis, the authors used a series of mixed linear effects models with random intercepts to assess WEMWBS changes from baseline to each endpoint: a time-only model in OA, a model adding baseline demographic predictors and their interactions with time, and a model comparing OA with matched YA via a time-by-age-group interaction. Additional analyses included paired t-tests for within-group changes, MANOVA/ANOVA to compare acute subjective effects between age groups, and a fourth mixed model testing three-way interactions between age group, time and acute predictors. Variance inflation factors were computed to check multicollinearity; the MEQ was removed from the three-way interaction model due to VIF >2.5. Pearson correlations were used to illustrate associations between predictor scores and WEMWBS change scores. Statistical significance was set at p < 0.05, with Bonferroni correction reported where applicable.

Sample and retention: Of 882 sign-ups, 819 provided baseline data. Among 106 people reporting age ≥60, 62 completed baseline and the 24-hour post-session survey and formed the OA analytic sample. Of these 62 OA, 53 completed the post-retreat questionnaire, 44 completed the two-week endpoint, 61 completed the four-week endpoint, and 23 completed the six-month endpoint. The OA sample had mean age 65.1 years (SD = 4.02; range 60–75), 50% male (31/62), 68.5% with a master's degree or higher (43/62), 74.2% reporting no diagnosed mental illness (46/62), and 56.5% reporting no prior psychedelic experience (35/62). Sixteen OA reported a psychiatric diagnosis; the most common were major depressive disorder (10/16) and anxiety disorder (9/16). Psychedelic use was predominantly psilocybin mushrooms/truffles (59 OA; 57 sessions at retreat centres in the Netherlands or Jamaica) and three reported ayahuasca. A matched YA sample of 62 participants (mean age 46.5 years, SD = 10; range 24–59) was selected for comparisons. Primary outcomes — well-being changes: Mixed effects regression in OA showed statistically significant increases in WEMWBS at two weeks (B = 4.09, 95% CI [1.87, 6.31], p < 0.001) and at four weeks (B = 3.05, 95% CI [1.04, 5.06], p = 0.004), interpreted as meaningful improvements in well-being. At six months, WEMWBS was nominally elevated by 1.72 units but this change was not significant (B = 1.72, 95% CI [-1.00, 4.44], p = 0.22). Paired t-tests confirmed significant within-group increases at two weeks (p = 0.006, Cohen's d = 0.48) and four weeks (p = 0.004, Cohen's d = 0.44), but not at six months (p = 0.63, Cohen's d = 0.10). Baseline predictors: In a mixed effects model including demographic predictors and their interactions with time, only history of mental illness interacted with time, such that OA reporting a lifetime psychiatric diagnosis showed larger well-being increases (interaction B = 6.62, 95% CI [1.53, 11.71], p = 0.019 at two weeks; B = 5.69, 95% CI [1.14, 9.94], p = 0.016 at four weeks). This effect remained after controlling for expectations. Comparisons with younger adults: A mixed effects model including OA and matched YA found no significant age-group by time interactions, indicating that the magnitude of well-being improvements at two and four weeks did not differ statistically between OA and YA (two-week interaction B = -2.36, 95% CI [-5.48, 0.76], p = 0.14; four-week interaction B = -2.55, 95% CI [-5.53, 0.42], p = 0.097). Acute subjective effects: A MANOVA comparing acute measures (EBI, MEQ, EDI, CEQ, COMS) across OA and YA was significant (Pillai’s Trace = 0.14, F(5,110) = 3.72, p = 0.004), indicating group differences in acute subjective responses. Follow-up analyses reported that OA had lower acute subjective effect scores on most metrics compared with YA, with the exception of challenging experiences (CEQ), where OA did not show lower scores. Mechanisms predicting well-being change: In a mixed effects model testing three-way interactions (age group × time × acute predictor), the MEQ was removed due to multicollinearity (VIF = 3.7). The model showed a significant negative three-way interaction for Emotional Breakthrough Inventory (EBI) × older age × endpoint at two weeks (b = -0.78, 95% CI [-1.43, -0.13], p = 0.02) and four weeks (b = -0.80, 95% CI [-1.46, -0.15], p = 0.02), indicating that emotional breakthrough predicted well-being gains less in OA than in YA. Conversely, a large but marginally significant positive three-way interaction was observed for post-retreat Communitas (COMS-PR) × OA × four-week endpoint (b = 2.04, 95% CI [-0.21, 4.28], p = 0.08), suggesting relational experiences across the retreat may have been more predictive of benefit in OA. Correlational results illustrating these effects showed that EBI correlated with four-week well-being change in YA (r[35] = 0.43, p = 0.008) but not in OA (r[46] = -0.04, p = 0.79). By contrast, COMS-PR correlated with four-week well-being change in OA (r[27] = 0.37, p = 0.03) but not in YA (r[34] = 0.01, p = 0.94). The authors report that other acute and post-acute predictors that correlated with change in YA did not significantly correlate with change in OA.

L. and colleagues interpret their findings as evidence that guided psychedelic group sessions are associated with clinically meaningful improvements in mental well-being among older adults at two and four weeks post-session, consistent with previous naturalistic studies in younger populations. They highlight that these improvements occurred despite OA reporting attenuated acute subjective psychedelic effects (e.g. lower scores for ego-dissolution, mystical and emotional breakthrough measures), suggesting that the mechanisms linking psychedelic experiences to well-being may differ by age. Specifically, the authors propose that psychosocial or relational processes inherent to group and retreat settings (measured here as Communitas across the retreat) may play a more prominent role for OA than intrapersonal acute effects such as emotional breakthrough. They note that OA who reported a lifetime psychiatric diagnosis showed larger well-being increases, a finding the authors align with the transdiagnostic potential of psychedelic treatments and argue is not explained by expectations in this sample. The authors discuss possible explanations for attenuated acute effects in OA, including sample differences from prior controlled studies and potential age-related reductions in cortical serotonin 2A receptor density, which could reduce subjective intensity. They also emphasise that attenuated acute intensity may be reassuring for the psychological safety profile of psychedelics in older people, though this observation contrasts with prior work implicating the acute experience quality as a key mediator of benefit. Key limitations acknowledged by the authors include the naturalistic, retreat-based context limiting generalisability to other settings or to controlled therapeutic protocols; a sample biased towards white, highly educated participants; imprecise measurement of dose and inability to control for co-use of other substances; and potential attrition bias given remote survey administration. The authors call for replication in larger, more representative samples and for controlled laboratory studies in older adults to clarify the roles of acute drug effects versus group psychosocial processes. They also suggest future work should characterise the specific psychotherapeutic and group activities occurring at retreats and validate instruments assessing experiences in older populations.

The authors conclude that, in this observational cohort, psychedelic group sessions were associated with rapid and sustained improvements in well-being among older adults, including those with prior psychiatric diagnoses. Compared with younger participants, older adults reported overall attenuated acute subjective psychedelic effects, while psychosocial experiences of togetherness across the retreat appeared relatively more important in predicting well-being gains. Given the underrepresentation of older adults in contemporary psychedelic trials, the authors recommend further research to investigate clinical efficacy, mechanisms and long-term utility of psychedelics for mental-health issues prevalent in later life.