Psilocybin microdosers demonstrate greater observed improvements in mood and mental health at one month relative to non-microdosing controls

The paper summarises renewed public and scientific interest in psilocybin, not only at full psychoactive doses but also at substantially lower ‘‘microdoses’’ that are intended not to disrupt normal functioning. Prior cross-sectional and a few prospective studies have reported that microdosing (commonly 0.1–0.3 g dried mushrooms, several times per week) is associated with self-reported improvements in mood, cognition and reductions in stress, depression and anxiety. However, existing longitudinal research is limited by small samples, inconsistent use of non-microdosing control groups, difficulties maintaining placebo/blind conditions, high rates of prior psychedelic experience among participants, and little investigation of combinations of microdosed psilocybin with other substances (so-called ‘‘stacking’’), such as lion's mane (Hericium erinaceus; HE) and niacin (vitamin B3; B3), which have been anecdotally combined with psilocybin in the community. K. and colleagues designed a large, naturalistic, prospective observational study to extend this literature. The study follows self-selected participants recruited via an iOS application, comparing individuals who microdose psilocybin-containing mushrooms against non-microdosing comparators over approximately 22–35 days. The authors aimed to examine changes in mood and mental health (including depression, anxiety and stress), and cognitive and psychomotor functioning, to investigate whether effects varied by age, gender or presence of mental health concerns, and to explore whether stacking psilocybin with HE and B3 was associated with different outcomes.

The researchers conducted a naturalistic, observational prospective study integrated within an iOS application. Eligible participants were 18 years or older, English-literate and able to use an iOS device. Recruitment targeted both individuals who were microdosing and individuals not engaged in microdosing through media and presentations related to psychedelic use. Participants completed a baseline assessment and a follow-up assessment 22–35 days later; both assessments used equivalent schedules and took about 20–30 minutes. Informed consent was obtained and the study received institutional ethics approval. Hypotheses and outcomes were not pre-registered. Assessments included self-report measures of mood and mental health and brief cognitive and psychomotor tasks adapted from Apple ResearchKit. Mood was measured with the Positive and Negative Affect Schedule (PANAS); depressive, anxiety and stress symptoms were measured with the DASS-21 (scores converted to percentages for one PANAS negative item that was missing due to technical error). Psychomotor and cognitive performance were assessed using an adapted finger tap task (number of alternating taps in 10 s), an adapted Corsi Block-Tapping spatial span task (number correct), and an adapted Paced Auditory Serial Addition Test (PASAT; total correct). For statistical analysis the authors used mixed-effects linear models across eight outcomes (DASS depression, DASS anxiety, DASS stress, PANAS positive, PANAS negative, finger taps, spatial span, PASAT). Models included continuous Time (days since baseline) and a dichotomous Microdose group factor (non-microdosers = 0, microdosers = 1). To limit model complexity and preserve subgroup sizes, only one moderator was included per model: Age (dichotomised as 55 and over versus under 55) was used in models of cognitive and mood outcomes, while presence of Mental Health Concerns (self-reported current psychological, mental health or addiction concerns) was used in DASS models. Models retained variables that predicted outcomes or were part of higher-order significant interactions. Where a Microdose*Time interaction was found, full factorial models examined three-way interactions with Gender and either Age or Mental Health Concerns. Supplementary analyses included removal of outliers (>2 standard deviations), exclusion of participants who reported microdosing at baseline to examine incident microdosing effects, and a set of stacking analyses comparing Psilocybin-only microdosers to Psilocybin+HE and Psilocybin+HE+B3 subgroups. Chi-squared tests and random trimming were used to address subgroup imbalances in age, microdose frequency and dosage when comparing stacking groups.

Across the principal mental health outcomes, the authors report small-to-medium improvements among microdosers relative to non-microdosers over the ~30-day interval. Mixed-effects models identified significant Microdose*Time effects for depressive, anxiety and stress symptoms (DASS domains) and for positive and negative affect (PANAS), and these effects were generally robust to removal of outliers and to analyses restricted to participants who were not microdosing at baseline. The Microdose*Time reduction in depressive symptoms was moderated by gender, with stronger reductions among females than males (Microdose*Gender*Time: F(1,1016)=6.61, b=0.17). Interactions between Microdose and Mental Health Concerns were non-significant, indicating that improvements associated with microdosing were observed both in participants who did and did not report current mental health concerns. For illustrative numerical changes, among microdosers reporting mental health concerns at baseline mean (SD) DASS depression changed from 18.85 (12.03) at baseline to 11.73 (9.85) at Month 1; anxiety from 11.04 (8.48) to 7.46 (6.68); and stress from 19.93 (9.71) to 13.91 (9.02). Among respondents without mental health concerns, depression changed from 10.40 (9.78) to 6.65 (7.60); anxiety from 6.53 (6.50) to 4.81 (5.57); and stress from 13.96 (9.12) to 9.78 (7.50). On mood measures (PANAS), Microdose*Time effects were observed and were equivalent across age groups (Age*Microdose*Time not significant for either positive or negative affect). Stacking (addition of HE or HE+B3) did not influence PANAS outcomes. Cognitive and psychomotor results were mixed. The finger tap task showed a main microdosing effect: microdosers demonstrated greater improvement in tapping performance than non-microdosers (F(1,886)=9.09, b=-0.24, p=0.03). This effect remained when excluding participants who were microdosing at baseline and after outlier removal. No main effects were observed for the spatial span task (F(1,944)=0.24, p=0.63) or PASAT (F(1,775)=0.21, p=0.65). Supplementary stacking analyses found no differences between Psilocybin-only and Psilocybin+HE microdosers on the DASS domains or finger tap. In contrast, comparison of Psilocybin-only versus Psilocybin+HE+B3 showed relatively greater improvement in finger tap scores with the addition of both HE and B3 (F(1,732)=3.93, b=-0.51, p<0.05). A three-way interaction with Age indicated that this stacking-related psychomotor improvement was driven by older respondents (Psilocybin only vs Psilocybin+HE+B3 * Time * Age: F(1,732)=8.4, b=0.6, p=0.004), with follow-up tests suggesting effects concentrated in participants aged 55 and over. These stacking findings were robust to controlling for subgroup differences in age, microdose frequency and dosage. The authors note that subgroup sample sizes for stacking comparisons—especially involving B3—were relatively small.

K. and colleagues interpret their findings as broadly consistent with earlier longitudinal studies of psychedelic microdosing: microdosers experienced improvements in mood and reductions in symptoms of depression, anxiety and stress over about one month, with effect sizes described as small to medium. The study is presented as the largest prospective investigation of psilocybin microdosing to date and one of the few to include a non-microdosing comparison group, lending weight to the observed associations. The authors highlight that participants who reported mental health concerns at baseline showed an average reduction in depressive symptoms that shifted from the moderate range to the mild range over the study period, a result the authors consider noteworthy given the public health burden of depression. The authors emphasise limits to causal inference in a non-randomised, unblinded observational design. They acknowledge that both microdosers and non-microdosers knew their group assignment, so expectancy or placebo effects could contribute to differences; they also note broader practical difficulties in achieving effective blinding with psychoactive compounds. The authors argue that naturalistic large-cohort designs remain valuable for examining consistency of effects across demographic and clinical subgroups, even if they do not fully isolate direct pharmacological effects. Regarding cognitive and psychomotor outcomes, the authors report mixed evidence: no effects on spatial memory or processing speed, but improved finger tap performance among microdosers, particularly in older participants combining psilocybin with HE and B3. They caution that these stacking-related psychomotor findings are novel, based on relatively small subgroups, and may be anomalous; the study lacked sufficient numbers to disentangle the effects of B3 alone versus the full combination, and the age cut-off at 55 was chosen for pragmatic power reasons. Consequently, the authors call for replication before drawing clinical implications. The authors list additional limitations: observational and exploratory design, potential response bias from self-selection and recruitment through pro-psychedelic venues, restriction to Apple iOS users, lack of assessment of dose and dosing practices, and no systematic measurement of adverse effects or interactions with common antidepressants or anxiolytics. They recommend future research with more systematic recruitment, randomised and placebo-controlled designs where feasible, and studies that assess dosing, potency, adverse effects and interactions to better define safety, efficacy and optimal practices for microdosing.