This cross-sectional survey (n=1,088) found that age significantly moderates the relationship between lifetime psychedelic use and mental health, with classic psychedelics such as psilocybin being associated with lower depression and anxiety in younger adults, but these benefits diminish with age and even reverse for anxiety in older participants.
Depression and anxiety affect one in five adults, with age affecting prevalence. While clinical trials suggest classic psychedelics (e.g., psilocybin, LSD) and non-classic psychedelics (e.g., MDMA, ketamine) may alleviate these symptoms, it remains unclear how these relationships function in naturalistic settings or how they vary across the lifespan. We conducted a cross-sectional survey of 1,088 adults (18–55 + years) to assess how lifetime psychedelic use - categorized as classic, non-classic, or mixed - relates to mental health. Using structural equation modeling, we found that age significantly moderates the relationship between psychedelic use and mental health outcomes. Specifically, classic psychedelic use was linked to lower depression and anxiety among younger adults, but these effects diminished with age - even reversing for anxiety in older participants. These age-related effects persisted independently of drug-use parameters - including dosage, frequency, and recency of use - and were moderated by mystical experiences for depression, but not for anxiety. Our findings suggest that age may be a meaningful moderator of mental health outcomes from psychedelic use. This underscores the potential value of age-stratified research to optimize the efficacy and safety of psychedelic-assisted interventions, including in aging populations.
Papers cited by this study that are also in Blossom
Olson, J. A. · Neuroscience Insights (2018)
Carhart-Harris, R. L., Bolstridge, M., Rucker, J. et al. · Lancet Psychiatry (2016)
Robison, R., Barrow, R., Conant, C. et al. · JAMA (2025)
Ross, S., Bossis, A. P., Guss, J. et al. · Journal of Psychopharmacology (2016)
The paper situates its inquiry in the context of the persistent global burden of depression and anxiety across adulthood and growing interest in psychedelics as potential interventions. The authors distinguish between "classic psychedelics" (primarily 5-HT2A agonists such as psilocybin, LSD and DMT) and "non-classic psychedelics" (e.g., ketamine, MDMA, ibogaine) that differ in neuropharmacology despite some experiential overlap. They note that clinical research has promising but limited evidence, often based on small or narrowly selected samples, while naturalistic (recreational) use is increasing and produces mixed findings for mental health outcomes. Age emerges as a plausible moderator because younger people reportedly experience more intense mystical or ego-dissolution phenomena and may be more vulnerable to acute adverse effects, while ageing is associated with reduced 5-HT2A receptor density and other neurobiological changes that could alter psychedelic responsiveness. Di Gregorio and colleagues therefore set out to test two primary questions in a cross-sectional, naturalistic sample: (1) whether lifetime use of classic and non-classic psychedelics is associated with differences in anxiety, depression, wellbeing, and life satisfaction compared with non-users; and (2) whether these associations vary across the adult lifespan. To address gaps created by exclusion of older adults in clinical trials and the predominance of younger participants in observational studies, the study deliberately recruited an age-diverse sample and classified participants into mutually exclusive groups (Classic-only, Non-classic-only, Classic+ [mixed use], and non-users) to enable theoretically informed comparisons while accounting for polydrug patterns.
The researchers conducted an anonymous, cross-sectional international survey delivered via Qualtrics. Inclusion criteria were being at least 18 years old and fluent in English. Recruitment occurred between November 2024 and August 2025 through multiple channels (Prolific, a university participant pool, social media, email, snowball sampling and other online platforms) to capture a broad, age-diverse and ecologically valid sample. The extracted text reports that 1,270 individuals enrolled and 182 did not complete the survey, yielding a final sample of 1,088; elsewhere the extraction refers to 1,008 adults as the full sample, and the text does not resolve this discrepancy. Participants supplied demographics and detailed lifetime psychedelic use information (which substances, age at first use, frequency within life-stage bands, typical dose category, motives, presence/absence of a guide, concurrent substance use and whether they had ever had a mystical experience). A checklist adapted from the MEQ-30 assessed ten mystical-type items (e.g., unity, transcendence of time/space, ego dissolution, encounters with the divine, feelings of love/compassion). Mental health measures were standard self-report scales: GAD-7 for anxiety (0–21), PHQ-9 for depression (0–27), WHO-5 for wellbeing (converted to 0–100), and the Satisfaction With Life Scale (SWLS; 5–35). Internal consistency was high (Cronbach's α: GAD-7 = .92; PHQ-9 = .89; WHO-5 = .90; SWLS = .91). If a participant had one missing item on a scale it was imputed by their mean on the remaining items; participants with two or more missing items were excluded from that scale's analyses. Substances were classified pharmacologically: classic psychedelics were defined as primary 5-HT2A agonists (e.g., psilocybin, LSD, DMT, mescaline, ayahuasca, 4-AcO-DMT), while non-classic psychedelics included dissociatives (ketamine, PCP, MXE, DXM/DXO, nitrous oxide), entactogens (MDMA, MDA, 6-APB) and oneirogenics (ibogaine). Atypical serotonergic compounds reported in free text (e.g., DOx, 2C-x, NBOMe, 5-MeO) were not classed as classic for analysis. For primary comparisons participants were grouped into three mutually exclusive categories: Non-Classic Only, Classic Only, and Classic+ (classic plus non-classic/atypical serotonergic substances). Lifetime cumulative use was estimated by converting frequency categories within six age bands into numeric midpoints and summing; years since last use were estimated by subtracting a midpoint for the most recent age-band of use from current age. For analysis the researchers fitted structural equation models (SEMs) with latent variables for anxiety, depression, wellbeing and life satisfaction. Each latent outcome was regressed on orthogonal contrast codes representing psychedelic-use category, age, and their interaction. Orthogonal contrasts allowed targeted hypothesis testing while keeping contrasts statistically independent. Models used robust standard errors with a Satorra–Bentler correction (MLM in Lavaan), and gender was included as a covariate to account for group imbalances. Follow-up analyses restricted to users examined additional use-history variables (cumulative lifetime use, age at first use, typical dose, recency). Significant interactions were probed with simple-slopes at mean ± 1 SD of age and, when near significance, with Floodlight (Johnson–Neyman) procedures. Logistic regression models assessed whether age and psychedelic-use contrasts predicted likelihood of reporting a mystical experience.
Participant grouping and use characteristics: The extracted text reports that, across the analysed sample, 562 participants reported no lifetime psychedelic use, 90 reported exclusive non-classic use (Non-classic-only), 126 reported exclusive classic use (Classic-only), and 310 reported mixed use (Classic+). Most psychedelic users initiated use between ages 18 and 25. Recreational motives dominated (reported by 77–93% across groups), therapeutic motives were reported by 6–20%, reported dosages were broadly similar across groups, and 83–86% of users reported taking psychedelics without a guide. Lifetime mystical experiences were reported by 60% of Classic-only users, 73% of Classic+ users, and 35% of Non-classic-only users. Primary SEM findings — anxiety and depression: In full-sample SEMs age was negatively associated with anxiety (β = -0.151, p < 0.001) and depression (β = -0.222, p < 0.001). There was a main effect in the anxiety model indicating lower anxiety among classic-psychedelic users (Classic-only and Classic+) compared with Non-users and Non-classic-only users (β = 0.093, p = 0.014, as reported). Crucially, a significant Age × Classic-vs-Nonclassic/Non-user interaction emerged for anxiety (β = -0.147, p < 0.001). Simple-slopes probing this interaction showed that classic-psychedelic use was associated with lower anxiety in younger adults (approximately 23 years: β = 0.227, p < 0.001), the association attenuated at the sample mean age (≈36 years: β = 0.083, p = 0.023), and was absent in older adults (≈50 years: β = -0.06, p = 0.201). For depression there was no clear main effect of psychedelic-use category (the Classic vs Non-classic/Non-user contrast approached significance: β = 0.073, p = 0.052), but there was a significant Age × Classic contrast interaction (β = -0.089, p = 0.017). Simple-slopes indicated lower depressive symptoms among classic users only in younger adults (β = 0.148, p = 0.009), with no significant associations at mean or older ages. Wellbeing and life satisfaction: WHO-5 wellbeing scores showed no significant main effects or age interactions with psychedelic-use category. Life satisfaction (SWLS) produced a different pattern: Non-classic-only users reported higher life satisfaction than Non-users (β = -0.084, p = 0.019). Life satisfaction also showed an Age × Classic contrast interaction (β = 0.087, p = 0.039). Simple-slopes suggested no differences at younger or average ages, but a trend toward lower life satisfaction for Classic-only/Classic+ users in older participants (β = 0.114, p = 0.057). A Johnson–Neyman analysis identified a significant lower life satisfaction for Classic users versus Non-classic/non-users at age 53 (p = 0.047). Robustness to use-pattern controls: Analyses restricted to psychedelic users (n = 562, per extracted text) replicated the age-moderated association between classic use and anxiety/depression for the Classic-only/Classic+ vs Non-classic-only contrast. In user-only simple slopes younger Classic users had lower anxiety than Non-classic-only users (β = 0.158, p = 0.017), whereas older classic users showed higher anxiety than non-classic users (β = -0.116, p = 0.038). Depression showed a similar trend though within-age-group comparisons did not reach significance. Importantly, controlling for years since last use, typical dose, and cumulative lifetime use did not eliminate the Age × Classic-use interactions: for anxiety the interaction remained significant when adjusting for years since last use (β = -0.138, p = 0.002), dose (β = -0.148, p = 0.001), and cumulative use (β = -0.144, p = 0.001). For depression similar interaction coefficients remained significant (years since last use β = -0.097, p = 0.024; dose β = -0.105, p = 0.025; cumulative β = -0.096, p = 0.031). The authors therefore report that recency, dosage and cumulative exposure do not account for the age-dependent associations observed. Mystical experience analyses: Logistic regression showed that participants who had used classic psychedelics (Classic-only and Classic+ combined) were more likely to report mystical experiences than non-classic-only users (OR = 1.560, p < 0.001). Higher reported dose was associated with increased odds of mystical experience (OR = 1.70, p < 0.001). The authors report that Classic+ users were more likely than Classic-only users to report mystical experiences (reported OR = 0.746, p = 0.009, noting this is how it appears in the extracted text). Significant interactions with age were observed (Age × Classic contrast: OR = 0.984, p = 0.024; Age × Classic+ vs Classic-only: OR = 1.973, p = 0.002). Simple-slopes indicated that classic users at younger and mean ages were significantly more likely to report mystical experiences than non-classic-only users, but this effect was not significant at older ages. When mystical experience was introduced into the mental-health models it attenuated the age-dependent association between classic use and depression (Age × Classic contrast became non-significant: β = -0.084, p = 0.064), whereas the age-dependent association with anxiety remained significant (β = -0.124, p = 0.006). The authors therefore report that age-related differences in mystical experience explain the age-dependent association between classic-psychedelic use and depression but not the equivalent pattern for anxiety.
Di Gregorio and colleagues interpret their results as identifying chronological age as a central moderator of the relationship between lifetime psychedelic use and mental-health measures in naturalistic settings. They emphasise that classic psychedelic use was associated with lower anxiety and depression only among younger adults in their sample; in later adulthood the pattern for anxiety reversed, with non-classic psychedelics associated with more favourable anxiety outcomes than classic compounds. Life satisfaction followed a related, age-dependent pattern: classic use was linked to higher levels among younger participants but to lower life satisfaction among older users (with a Johnson–Neyman cutoff observed at about 53 years). These age-dependent patterns were not explained by recency of use, cumulative lifetime exposure, dose, or age at first use, according to their adjusted models, suggesting that chronological age itself is the key moderator identified in this dataset. The authors propose converging biological, neural and psychological mechanisms that could underlie the observed age moderation. At a molecular level they point to age-related declines in 5-HT2A receptor density that could reduce the intensity of classic-psychedelic experiences. At the neural and cognitive levels they cite reduced plasticity and cognitive flexibility and greater stability/segregation of brain networks with age, which may limit reorganisation following salient experiences. Psychologically, more mature ego structures in older adults may make them less susceptible to ego-dissolution states. Together, these factors are presented as plausible contributors to attenuated psychedelic responsiveness with age. The distinction between classic and non-classic psychedelics was important to the authors' interpretation. They argue that including a Non-classic-only group allowed a stricter control for openness to altered states and revealed potentially clinically relevant differences: non-classic compounds may retain or show greater effectiveness for anxiety in older adults because their primary molecular targets decline less with age than 5-HT2A, and they may pose different medical risk profiles in older populations. The authors also note that Classic+ users were more likely to report mystical experiences in older age than Classic-only users, suggesting that co-use patterns could shape subjective experience and potentially outcomes, though they cannot disentangle pharmacological from baseline trait differences in this study. Regarding mechanisms linking experience to outcomes, the authors report that mystical-type experiences appear more important for depression than for anxiety: controlling for mystical experience removed the age-dependent association between classic use and depression but not the age-dependent association for anxiety, implying symptom-specific psychological mediators. The authors also note domain-specific findings: psychedelics did not affect general wellbeing (WHO-5) in this sample, perhaps due to ceiling effects or the scale's limitations, while non-classic use correlated with higher life satisfaction, which may relate to contextual or social factors associated with use. The authors explicitly acknowledge methodological limitations: the study's cross-sectional design precludes causal inferences and is vulnerable to selection and self-selection biases (pre-existing psychological traits may differ between user groups). All outcomes were self-reported and socioeconomic confounders were not accounted for. They note that use of a Non-classic-only comparison group partially mitigates selection bias by including people with similar openness to altered states. Finally, they highlight strengths: a relatively large, age-diverse sample and analyses that model latent variables and orthogonal contrasts to test theoretically motivated hypotheses, as well as ecological validity from capturing real-world polydrug patterns. The authors argue these results motivate future longitudinal and clinical studies that explicitly test age as a moderator of psychedelic treatment effects and examine compound-specific and context-specific mechanisms.
The present observational study used an online platform (Qualtrics) to deliver a cross-sectional, international survey designed to assess participants' mental health and naturalistic psychedelic use. Inclusion criteria for participants were: (1) at least 18 years old and (2) uent in English language.
Participants were recruited between November 2024 and August 2025. The study was conducted on a fully anonymous basis, with no personally identifying information collected (including location or IP address). Ethical approval was obtained from the Department of Psychology Ethics Committee at Durham University (study #2570). Recruitment was conducted through multiple channels, including the online research platform Proli c (www.proli c.com), the university's participant pool, online platforms, social media, word-of-mouth, email, and snowball sampling (e.g., referrals from participants to others in their social networks). Multi-channel recruitment was used to ensure a balanced proportion of psychedelic users, enhance demographic and psychological diversity, and capture complex, heterogeneous patterns of polydrug use. A total of 1270 individuals enrolled in the study. Of these, 182 did not complete the full survey, resulting in a nal sample of 1088 participants.
Participants rst provided demographic information, including age and gender. Next, participants were then asked whether they had ever used psychedelics at any point in their lives. Those who reported no psychedelic use were directed to complete subsequent mental health questionnaires (see below). Participants who reported lifetime use were asked to indicate which substances they had taken from a list including (LSD, psilocybin, DMT, ayahuasca, mescaline, MDMA, ketamine, or other-with option to include free text). Psychedelic users then provided additional details on their use patterns, including age at rst use, frequency of use across life stages ("1-2 times," "3-5 times," "6-10 times," "10 times +"), typical dosage (microdose, mild, moderate, high, or varies), and reasons for use (recreational, therapeutic, spiritual, or other-with option to include free text). They also reported whether use was for mental health purposes (self-medication), occurred with or without a guide, involved concurrent substances use, and whether they ever had a mystical experience during psychedelic use.
Participants who reported mystical experiences completed a checklist (yes/no) adapted from the Revised Mystical Experience Questionnaire (MEQ-30, covering the following items: (1) sense of unity or openness, (2) transcendence of time and space, (3) ego dissolution, (4) encounter with the divine or a higher power, (5) profound emotional insight or healing, (6) sacred geometry or visions, (7) feelings of love and compassion, (8) mystical or religious revelations, (9) awe and wonder, and (10) experiences of rebirth or renewal.
Anxiety symptoms were assessed with the 7-item Generalized Anxiety Disorder Scale (GAD-7; Spitzer et al., 2006), which evaluates symptom frequency over the past two weeks. It assesses seven core symptoms: nervousness, inability to control worrying, excessive worry, restlessness, di culty relaxing, irritability, and fear that something awful might happen. Each item is rated on a four-point scale from "not at all" to "nearly every day," producing a total score ranging from 0 to 21. Higher values re ect greater levels of anxiety. Depression symptoms were assessed with the 9-item Patient Health Questionnaire (PHQ-9; Kroenke et al., 2001), which evaluates symptom frequency over the past two weeks. Each item is rated on a fourpoint scale from "not at all" to "nearly every day," yielding a total score between 0 and 27. Higher values re ect greater levels of depressive symptoms. General wellbeing was assessed with the World Health Organization 5-item Well-Being Index (WHO-5; World Health Organization, 2020), which measures positive mood, vitality, and general interest over the past two weeks. Items are rated on a six-point scale from "at no time" (0) to "all of the time" (5), generating a raw score between 0 and 25. This score is multiplied by four to produce a percentage score (0-100), with higher values re ecting greater wellbeing. Life satisfaction was measured with the Satisfaction with Life Scale (SWLS; Diener et al., 1985), which consists of ve items assessing global cognitive judgments of life satisfaction. Each item is rated on a seven-point scale from "strongly disagree" (1) to "strongly agree" (7), yielding a total score ranging from 5 to 35. Higher scores indicate greater satisfaction with life. Internal consistency was high across all scales. Cronbach's α was .92 for the GAD-7, .89 for the PHQ-9, .90 for the WHO-5, and .91 for the SWLS, indicating excellent reliability for all measures. For all four scales, if only one item was missing, its value was imputed using the participant's mean score on the remaining items. Participants with two or more missing items were excluded from analyses for that scale.
Participants were classi ed according to the pharmacological pro les of the substances they reported using. Classic psychedelics, de ned by primary 5-HT 2A receptor agonism, included psilocybin/psilocin (magic mushrooms), lysergic acid diethylamide (LSD), N,N-dimethyltryptamine (DMT), mescaline (peyote, San Pedro), ayahuasca, 4-AcO-DMT, and LSD analogues. Non-classic psychedelics comprised dissociatives (ketamine, phencyclidine [PCP], methoxetamine [MXE], dextromethorphan/dextrorphan [DXM/DXO], nitrous oxide), entactogens (3,4-methylenedioxymethamphetamine [MDMA], 3,4methylenedioxyamphetamine [MDA], 6-APB), and oneirogenics (ibogaine, noribogaine). Atypical serotonergic psychedelics reported as free text (DOx series, 2C-x series, NBOMe derivatives, 5-MeO compounds, and LSA-containing seeds) were not considered classic psychedelics because, despite their 5-HT2A agonism, they exhibit broader serotonergic activity and less established safety or therapeutic pro les. Substances reported in free-text responses but excluded from analysis were those lacking a clear psychedelic, neuroplastic, or therapeutic pro le, acting primarily as stimulants, or serving only as facilitators in psychedelic mixtures; these included salvia divinorum, mephedrone, cannabis, and harmala alkaloids. For statistical comparisons, participants were grouped into three mutually exclusive categories: (1) Non-Classic Only (exclusive use of non-classic psychedelics), (2) Classic Only (exclusive use of classic psychedelics), and (3) Classic+ (use of classic psychedelics in combination with nonclassic and/or atypical serotonergic psychedelics).
To estimate cumulative lifetime use, for each participant, categorical responses within each of six age bands were converted into numeric midpoints ("1-2 times" = 1.5, "3-5 times" = 4, "6-10 times" = 8, "10 times+" = 10) and summed to create a cumulative estimate of lifetime psychedelic use. To estimate years since last use, we identi ed the most recent age band in which a participant reported use, assigned a representative midpoint to each age band ("Under 18" = 16, "18-25" = 21, "26-34" = 30, "35-44" = 40, "45-54" = 50, "55+" = 60), and subtracted that midpoint from each participant's current age to yield an approximate number of years since last psychedelic use.
We tested a series of structural equation models (SEMs) using robust standard errors and a Satorra-Bentler correction ("MLM" in Lavaan 0.6. 19; Rosseel, 2012). In these models, each latent variable of interest (anxiety [GAD-7], depression [PHQ-9], life satisfaction [SWLS], and wellbeing [WHO-5]) was regressed onto orthogonal contrast codes representing psychedelic drug type, age, and their interaction. Using orthogonal contrasts allowed us to test theoretically motivated comparisons while ensuring that each contrast captured unique variance, facilitating clearer interpretation compared with other tests that do not target speci c hypotheses. In addition, modelling latent variables in SEM has the advantage of reducing measurement error and individual-level variability, providing more reliable estimates than analyses based on raw observed scores. For initial full-sample models (including both psychedelic users and non-users), the orthogonal contrasts tested were: (A) the difference between individuals who had ever used classic psychedelics versus those who had never used classic psychedelics (Classic-only = - 1, Classic+ = -1, Non-classic-only = + 1, Nonusers = + 1); (B) the difference between individuals who had used only classics versus those who had used both classics and non-classics (Classic-only = - 1, Classic+ = +1, Non-classic-only = 0, Non-users = 0); (C) the difference between users of only non-classics versus non-users (Classic-only = 0, Classic + = 0, Non-classic-only = - 1, Non-users = + 1). Follow-up analyses were conducted to evaluate whether additional variables characterizing psychedelicuse history (e.g., cumulative lifetime use, age at rst use, typical dosage, and recency of use) altered associations. Because these variables were not available for non-users, models included only users and the contrasts tested in this subsample were: (A) the difference between individuals who had ever used classic psychedelics versus those who had only used non-classic psychedelics (Classic-only = - 1, Classic+ = -1, Non-classic-only = + 2); (B) the difference between individuals who had used only classics versus those who had used both classics and non-classics (Classic-only = + 1, Classic+ = -1). Signi cant interactions were probed using simple-slopes analyses at the mean and ± 1 SD of age. When interactions approached signi cance (p > 0.05 and < 0.10) in simple-slope analyses, Floodlight (Johnson-Neyman) procedures were applied to determine the speci c age ranges at which group differences reached statistical signi cance. Finally, age-related variation in mystical experience was examined using logistic regression. The likelihood of reporting a mystical experience was regressed onto age, psychedelic-use contrasts, and their interactions. Simple slopes were probed at the mean and ± 1 SD of age to determine whether age moderated group differences in the probability of experiencing mystical effects.
Tableshows demographic characteristics of the 1008 adults who completed the full survey. Of these, 562 reported no lifetime use of psychedelic substances, 90 reported exclusive use of non-classic psychedelics ("Non-classic-only"), 126 reported exclusive use of classic psychedelics ("Classic-only"), and 310 reported use of classic psychedelics in combination with non-classic and/or atypical serotonergic psychedelics ("Classic+"; see Methods for full criteria on classi cation). Psychedelic use characteristics across drug groups are presented in Table. Most psychedelic users reported initiating use between the ages of 18 and 25 years. Recreational motives were predominant across all groups (77-93%), followed by therapeutic intentions (6-20%). Reported dosages were broadly comparable across groups. Most participants (83-86%) indicated using psychedelics without the presence of a guide. Lifetime mystical experiences were reported by 60% of Classic-only users and 73% of Classic+ users, whereas substantially fewer Non-classic-only users (35%) reported such experiences.
To disentangle theoretically distinct comparisons between psychedelic-use groups while avoiding redundant tests, we tted structural equation models incorporating orthogonal contrasts (see Methods). These contrasts tested: (1) a main effect of classic psychedelic use (Classic [only/+] vs Non-users/Nonclassic); (2) whether use of classic psychedelics alone differs from use in combination with non-classic or atypical serotonergic substances (Classic-only vs Classic+); and (3) a main effect of non-classic psychedelic use (Non-classic-only vs Non-users). All models showed good ts (Table) and included gender as a covariate to account for group imbalances (Table). We rst examined the association between psychedelic-use group and anxiety (GAD-7). In the rst-order model, age showed a negative association with anxiety (β = -0.151, p < 0.001; Table). There was also a signi cant main effect of using Classic psychedelics, with lower anxiety among classic-psychedelic users (Classic-only and Classic+) compared with Non-users and Non-classic-only users (β = 0.093, p = 0.014) (Fig.). The other contrasts were not signi cant (Classic-only vs Classic+: β = -0.001, p = 0.975; Non-classic-only vs Non-users: β = 0.038, p = 0.291). In the second-order model, a signi cant age × Classic vs Non-classic/Non-users interaction (β = -0.147, p < 0.001) indicated that the association between classic psychedelic use and anxiety varied with age. To characterize this moderation, we probed simple slopes at three representative age groups: younger adults (- 1 SD, 23 years), mean age (36 years) and older adults (+ 1 SD, 50 years). These analyses showed that classic psychedelic use was associated with lower anxiety in younger adults (β = 0.227, p < 0.001; Fig.), that this association attenuated at the sample mean age (β = 0.083, p = 0.023), and that it was absent in older participants (β = -0.06, p = 0.201). Next, we examined the association between psychedelic use group and depression (PHQ-9). Age showed a negative association with depressive symptoms (β = -0.222, p < 0.001; Table). There was no signi cant main effect of psychedelic-use group, although the contrast comparing classic-psychedelic users (Classic-only and Classic+) with participants who had never used classic psychedelics (Non-classic-only and Non-users) was close to signi cance (β = 0.073, p = 0.052, Fig.). Depression outcomes nonetheless showed a signi cant interaction between age and whether participants used or not classic psychedelics (Classic-only/Classic + vs Non-classic-only/Non-users, β = -0.089, p = 0.017). Simple-slopes analyses indicated that Classic-only/Classic+ were associated with lower depression only among younger adults (β = 0.148, p = 0.009; Fig.), with no signi cant association at the mean age or at older ages. For wellbeing (WHO-5), we found no signi cant main effect of psychedelic use category or interactions with age (Table, Fig.). In contrast, life satisfaction (SWLS) showed both a main effect and an agemoderated association, but with a different pattern from anxiety and depression. Non-classic-only users reported higher life satisfaction than Non-users (β = -0.084, p = 0.019; Table, Fig.), whereas other contrasts between psychedelic-use categories were not signi cant. Life satisfaction also showed a signi cant interaction between age and classic-psychedelic use (Classic-only/Classic + vs Non-classiconly/Non-users; β = 0.087, p = 0.039), while the Non-classic-only vs Non-user contrast did not interact with age (p = 0.328). Simple-slopes analyses (Fig.) indicated no signi cant differences between Classic-only/Classic+ users and Non-classic-only/Non-users at younger or average ages, but a trend toward lower life satisfaction for Classic-only/Classic+ users emerged among older participants (β = 0.114, p = 0.057). To more precisely identify potential ages at which classic-psychedelic use was associated with lower life satisfaction, we conducted a oodlight (Johnson-Neyman) analysis. This analysis revealed that Classic-only/Classic+ users reported signi cantly lower life satisfaction compared with Non-classic-only/Non-users at 53 years of age (p = 0.047). Overall, these results indicate that age moderates associations between type of psychedelic use and mental health outcomes: classic psychedelic use was linked to lower anxiety and depression in younger adults, use of non-classic psychedelics was associated with higher life satisfaction overall, and classicpsychedelic use was associated with lower life satisfaction in older adults.
We next asked whether the age-dependent associations could re ect obvious confounds in drug-use patterns. For example, older participants may have used psychedelics longer ago (so effects could have faded), accumulated more lifetime exposure, or younger and older users may differ in typical dosages. To account for effects of psychedelic use patterns, we generated a fresh set of structural models with orthogonal contrasts including psychedelic users only (n = 562). Con rming the full-sample ndings, age signi cantly moderated the association between psychedelicuse type and both anxiety and depression for the Classic-only/Classic + vs Non-classic-only contrast (Table). Simple-slopes analyses indicated that younger Classic-only/Classic+ exhibited lower anxiety scores than Non-classic-only (β = 0.158, p = 0.017; Fig.), whereas no differences were observed at the mean age. In older users, the pattern was reversed, with classic users showing higher anxiety scores than non-classic users (β = -0.116, p = 0.038). Depression displayed a qualitatively similar age-related trend for the Classic-only/Classic + vs Non-classic-only contrast, although none of the comparisons reached statistical signi cance within any of the three age groups (Fig.). With the substantive full-sample pattern replicated in the user-only analyses, we then tested whether speci c aspects of psychedelic use could account for these age-related effects. Controlling for years since last use (Table) preserved the signi cant Age × Classic-only/Classic + vs Non-classic-only interaction for both anxiety (β = -0.138, p = 0.002) and depression (β = -0.097, p = 0.024). Adjusting for dosage produced the same result (anxiety: β = -0.148, p = 0.001; depression: β = -0.105, p = 0.025). Likewise, accounting for cumulative lifetime use left the interaction intact for both anxiety (β = -0.144, p = 0.001) and depression (β = -0.096, p = 0.031). This indicates that the age-related pattern linking classic-psychedelic use to anxiety and depression is not explained by how recently, how often, or at what dosage participants used these substances.
Given that differences in drug-use patterns did not account for age-related associations between classicpsychedelic use with depression and anxiety, we next asked whether results could be explained by factors known to vary with age. Mystical experience was a prime candidate, as prior research shows younger adults are more likely than older adults to report these experiences (Kettner et al., 2024). Using logistic regression, we modelled the likelihood of reporting a prior mystical experience based on age and psychedelic drug type. (Fig., Table). Consistent with prior evidence, participants who had used classic psychedelics (Classic Only and Classic+ combined) were more likely to report mystical experiences than those who had used only non-classic psychedelics (OR = 1.560, p < 0.001, Fig.). As expected, higher doses were associated with higher likelihood of mystical experience (OR = 1.70, p < 0.001), aligning with established evidence that high doses are needed to induce these experiences. Interestingly, participants who had used combinations of classic and non-classic or atypical serotonergic psychedelics (Classic+) were more likely to report mystical experiences than those who had used only classic psychedelics (OR = 0.746, p = 0.009). Consistent with prior work (Kettner et al., 2024), both the Age × Classic (only/+) vs Non-classic-only contrast and the Age × Classic + vs Classic-only contrast showed signi cant interactions with age (OR = 0.984, p = 0.024; OR = 1.973, p = 0.002, respectively). Dose did not account for these effects (Table), indicating that age-related differences in mystical experience could not be explained by dosage. Simple-slopes analyses for the Classic (only/+) vs Non-classic-only contrast showed that younger adults (23 years) and participants at the mean age (36 years) who had used classic psychedelics (Classic Only or Classic+) were signi cantly more likely to report mystical experiences than Non-classic-only users (Younger: OR = 0.523, p < 0.001; Mean: OR = 0.649, p < 0.001; Fig.). By contrast, at older ages (50 years), the effect was nonsigni cant (OR = 0.807, p = 0.122), indicating a decline in the ability of classic psychedelics to induce mystical experiences with age. For the Classic-only vs Classic+ contrast, we found similar likelihood of mystical experiences in younger participants (OR = 0.861, p = 0.405), a nonsigni cant trend toward higher likelihood in Classic+ users at mean age (OR = 1.260, p = 0.054), and signi cantly higher likelihood in older Classic+ users (OR = 1.830, p < 0.001; Fig.). Given the age-related decline in mystical experiences associated with Classic psychedelics, we tested whether controlling for the mystical experience eliminated the age-dependent associations between Classic-psychedelic use and mental health. Controlling for mystical experience rendered the Age × Classic (only/+) vs Non-classic interaction non-signi cant for depression (β = -0.084, p = 0.064), consistent with prior evidence linking mystical experience to antidepressant outcomes. In contrast, the interaction for anxiety remained signi cant (β = -0.124, p = 0.006; Fig.). Overall, these results suggest that age-related differences in mystical experience mediate the association between classic-psychedelic use and lower depression in younger adults, but do not explain age-dependent effects on anxiety, leaving the latter largely unexplained by any of the control variables tested.
Our study identi es age as a central moderator of mental-health associations with psychedelic use. Classic psychedelics were linked to lower anxiety and depression only among younger adults. Interestingly, the pattern for anxiety reversed in later adulthood, with non-classic psychedelics showing more favorable anxiety outcomes than classic compounds. These age-related effects were not explained by factors such as recency, typical dose, age at rst use, or cumulative exposure, and instead appear to re ect chronological age itself. This same age-related pattern emerged regarding life satisfaction, where classic psychedelic use was associated with higher levels in younger participants but lower levels in older users. Furthermore, the likelihood of reporting mystical-type experiences declined with age and explained the age-dependent association between classic psychedelic use and depression but not anxiety, suggesting that the therapeutic importance of the mystical experience may be symptomspeci c.
The observed association between classic psychedelic use and positive mental health outcomes aligns with prospective and retrospective naturalistic studies reporting improvements in depression and anxiety with psilocybin. While previous research has compared wellbeing changes following psilocybin use across two age groups (Kettner et al., 2024), to our knowledge no study had jointly examined depression, anxiety, and life satisfaction, nor formally tested age as a moderator of these outcomes. We report here, for the rst time, that age moderates the association between classic psychedelic use and mental health outcomes: speci cally, younger users exhibited lower anxiety and depression scores compared with non-users, a pattern not observed in older participants. These ndings align with observations from independent clinical trials, where younger cohortshave historically reported more pronounced reductions in depressive symptoms compared to older populations. While a formal systematic comparison of these disparate trials has remained absent from the literature, our study provides the rst direct evidence within a single, age-diverse sample to suggest that age-related responses to psychedelic may not be uniform. Our nding of age moderation may also help account for divergent results across recent naturalistic studies. For example, a recent study reported higher levels of mixed anxiety and depression among psilocybin users in a sample with a mean age of approximately 48 years. This age group is comparable to our older participant group, which showed mental health scores similar to non-users. In contrast, studies reporting bene cial effects of psilocybinexamined younger samples, with mean ages of 27 and 38 years, respectively, closely aligning with the age groups in which we observed lower anxiety and depression relative to non-users. Taken together, our ndings suggest that age is a critical variable in psychedelic outcomes. Although the cross-sectional nature of the present study warrants caution, these results underscore the need for future clinical research to formally examine age as a key factor in uencing the e cacy of psychedelic interventions. Several converging processes may account for the observed age-related pattern. At the molecular level, the intensity of psychedelic experiences -and their associations with mental health -depends on 5-HT2A receptor activation. 5-HT2A receptor density declines substantially with age, which is likely to reduce the brain's capacity to enter profound psychological states that facilitate emotional recalibration. Consistent with this account, the age-related reduction in mystical-type experiences observed in our sample, from approximately 75% in younger participants to around 50% in the older cohort, broadly matches the expected decline in 5-HT2A receptor availability, estimated at 7-17% per decade across the adult lifespan. Together, these ndings support the idea that age-related molecular changes may contribute to attenuated psychedelic responsiveness. Beyond molecular mechanisms, age is associated with reduced neural plasticity and cognitive exibility, as well as a shift toward more segregated and stable brain network architecture from mid-adulthood onwards 46 . These factors may limit the brain's ability to reorganize in response to salient experiences, making the integration of psychedelic states (and their mental health correlates) less pronounced in older adults. Psychologically, older adults possess more mature ego structures, which may resist the temporary state of ego dissolution often invoked by classic psychedelics 48 , further constraining the potential impact of these experiences. Taken together, pharmacological, neural, and psychological factors likely converge to produce the age-dependent patterns observed in our study, highlighting chronological age as a critical moderator of mental health associations with classic psychedelic use.
The inclusion of a distinct group of non-classic psychedelic users, a design choice previously unimplemented in naturalistic research, was a key element of our analysis. This structure served two key purposes: rst, it provided a stringent comparison group to control for a general 'openness' to altered states; second, it enabled a direct examination of the pharmacological divergence between classic and non-classic substances. This latter objective is particularly salient given the partially overlapping, yet mechanistically distinct, neuroplasticity-enhancing properties of classic and non-classic substances. A notable nding was that non-classic psychedelics showed more favorable anxiety outcomes in older adults. Mechanistically, these age-related differences likely re ect the stability of speci c neurotransmitter systems. For instance, the main molecular targets of non-classic psychedelics show a weaker age-related decline in density than 5-HT2A. This raises the possibility that non-classic compounds may have greater therapeutic utility in older populations, particularly as classic psychedelics may also present increased cardiovascular risks in these cohorts. For anxiety and depression, Classic-only and Classic+ users showed similar trends and appeared distinct from Non-classic-only, suggesting that use of classic psychedelics, regardless of co-use, may be the primary driver of these associations. However, Classic+ users were more likely to report mystical-type experiences in older age than those who had used only classic substances. It is possible that co-use patterns may shape subjective experiences and potentially mental health outcomes. However, we cannot fully rule out the possibility that some of the associations in the present study re ect differing baseline psychological traits between psychedelic class users. Future studies should determine whether the observed associations are driven by speci c compounds, especially within the pharmacologically heterogeneous non-classic category (e.g., MDMA vs ketamine) or by interactions between speci c substances.
In line with recent and ongoing clinical trials showing anxiolytic properties of psilocybin and LSD, our study demonstrates that these associations persist in naturalistic settings. Furthermore, our ndings indicate that mystical experiences may be more crucial for depression than anxiety. Introducing mystical experiences in the statistical model abolished the age x drug type interaction for depression, whereas the interaction for anxiety remained. This suggests that the psychological mechanisms through which naturalistic psychedelic experiences in uence mental health outcomes may differ between these two conditions. Psychedelics use did not in uence broader wellbeing. This dissociation suggests that positive affect and general wellbeing may be more stable traits and less sensitive to episodic psychedelic use than negative affect. An alternative explanation is methodological: many participants reported high wellbeing, raising the possibility of ceiling effects, and the WHO-5 offers a limited assessment of this multi-dimensional construct. By contrast, non-classic psychedelics were positively associated with life satisfaction. Their common use in relational, social, or therapeutic environments may enhance interpersonal connection and positive autobiographical memory -factors closely tied to life satisfaction. Distinguishing whether these effects re ect drug pharmacology, contextual factors, or user characteristics will require more granular future work.
Several methodological issues warrant caution when interpreting the ndings. The cross-sectional design precludes causal inference and is vulnerable to selection and self-selection bias. Systemic differences in psychological traits, including emotional stability, personality, and risk tolerance may exist between individuals open to classic psychedelic use and those who are not. Consequently, the observed associations may partially re ect preexisting psychological resilience or a higher baseline for openness among those inclined toward psychedelic use, rather than the effects of the substances themselves. However, the use of a 'Non-classic-only' comparison group, which includes individuals with similar risk pro les and psychological openness to users of classic psychedelics, partially addresses this potential selection bias. Furthermore, all outcome measures relied on self-reporting, and the study did not account for socioeconomic factors, which may in uence both psychedelic use and mental health outcomes. Nonetheless, our large and age-diverse sample enabled robust modelling of age-by-drug-class interactions. Examining psychedelic use within real-world polydrug patterns increases ecological validity
Create a free account to open full-text PDFs.
Rockhill, K. M., Black, J. C., Ladka, M. S. et al. · Annals of Internal Medicine (2025)
Nayak, S., Jackson, H., Sepeda, N. D. et al. · Frontiers in Psychiatry (2023)
Nygart, V., Pommerencke, L. M., Haijen, E. et al. · Journal of Psychopharmacology (2022)
Raison, C. L., Jain, R., Jain, S. et al. · Frontiers in Psychiatry (2022)
Carter, B., Cleare, A. J., Rucker, J. et al. · Neuropsychiatric Disease And Treatment (2023)
Kettner, H., Roseman, L., Gazzaley, A. et al. · The American Journal of Geriatric Psychiatry (2024)
Russ, S. L., Carhart-Harris, R. L., Maruyama, G. et al. · Psychology of Consciousness Theory Research and Practice (2019)
Barrett, F. S., Johnson, M. W., Griffiths, R. R. · Journal of Psychopharmacology (2015)
Griffiths, R. R., Johnson, M. W., Richards, W. A. et al. · Psychopharmacology (2011)
Kometer, M., Schmidt, A., Bachmann, R. et al. · Biological Psychiatry (2012)
Carhart-Harris, R. L., Goodwin, G. M. · Neuropsychopharmacology (2017)