A systematic study of microdosing psychedelics

Microdosing denotes the regular ingestion of very small, sub‑perceptual amounts of serotonergic psychedelics (commonly LSD or psilocybin) purported to confer benefits such as improved mood, creativity, focus and wellbeing. The phenomenon has attracted substantial popular and media interest and large online communities, but peer‑reviewed empirical work is scarce. The broader psychedelic literature indicates that higher doses of classic psychedelics produce lasting changes in mood, cognition and perception through serotonergic (5‑HT2A) mechanisms and altered network connectivity, providing a plausible biological rationale for microdosing, yet the evidence specific to low, repeated doses is limited and potentially confounded by expectancy effects. Polito and colleagues designed an initial exploratory programme to characterise self‑reported short‑ and longer‑term effects of microdosing and to gauge the role of expectancies. Study One was a six‑week, observational tracking study of people who microdose, combining daily brief ratings with comprehensive psychometric batteries at baseline and after six weeks. Study Two surveyed expectations about the effects of a hypothetical six‑week microdosing regimen in a larger sample from the same online communities, to assess whether popular expectations align with the effects observed in Study One. Nine psychological domains were selected a priori for investigation: mood, attention, wellbeing, mindfulness, mystical experience, personality, absorption, creativity and sense of agency.

The research comprised two linked, internet‑based studies recruiting participants from online microdosing communities (reddit.com/r/microdosing, bluelight.org and Facebook groups). Both studies excluded people reporting a history or current diagnosis of primary psychotic, mood, anxiety, dissociative or substance disorders. All contact was automated and anonymous; ethics approval and electronic consent were obtained. Study One was a longitudinal observational study open April 2016–April 2017. Of 1,181 initial visitors, 251 completed the baseline battery; after exclusions for duplicate responses and those who did not microdose during the study period, 98 participants provided at least one daily report and were included in the daily‑ratings analyses, while 63 participants completed both baseline and post‑study psychometric batteries and were included in the long‑term analyses. Over 42 days participants received daily emails with a short set of items (Connectedness, Contemplation, Creativity, Focus, Happiness, Productiveness, Wellbeing) rated on a 5‑point Likert scale; they also completed a comprehensive baseline and 6‑week battery (~50 minutes) covering the nine domains. Key instruments included the DASS‑21 for depression/anxiety/stress, Mind Wandering Questionnaire (MWQ), Quality of Life Inventory (QOLI, modified due to a coding error), MAAS, Hood Mysticism Scale, M5P personality questionnaire, Tellegen Absorption Scale, Creative Personality Scale, and Sense of Agency Rating Scale (SOARS). Participants reported substances and doses; some reports were excluded for implausible dosages or concurrent use of higher‑dose substances. Study One statistical approach used linear mixed‑effects models. For daily ratings Day type (Baseline, Day0 dosing day, Day1, Day2) was a categorical fixed effect with subject intercepts as random effects; each daily measure was treated as a separate family and Holm‑Bonferroni correction was applied within measures. For long‑term measures the models included Time (baseline vs post), prior microdosing Experience (none vs some) and number of Doses during the study as fixed effects, plus Time x Experience and Time x Doses interactions; again subject intercepts were random effects and multiple comparisons were corrected. Conditional R2 estimates were reported as a relative effect‑size index. Study Two (April–July 2017) was a cross‑sectional expectations survey. The team developed items corresponding to each Study One subscale and asked participants (n = 263) to imagine regularly microdosing for six weeks and to predict directional change (decrease/stay same/increase, scored −1/0/+1) and rate confidence (1–5). Weighted expectation scores combined direction and confidence (confidence weights scaled so a 5 carried twice the weight of a 1). Analyses included one‑sample tests against 0 to determine significant expectations and rank‑order comparisons (Kendall's tau‑b) between naïve and experienced participants and between expectations and observed Study One effects.

Study One—sample and dosing: Ninety‑eight participants supplied 1,792 daily reports across the 42‑day period, comprising 489 initial microdosing reports; after excluding three implausible high doses and eight instances of concurrent high‑dose substance use, 478 microdosing‑day reports remained. Reported substances were predominantly LSD (48.1% of microdosing reports) and psilocybin (47.1%), with small proportions of mescaline and a variety of novel serotonergic compounds. Participants reported a mean confidence of 5.9/7 in their dose estimates. Among the 63 who completed pre‑ and post‑study batteries, 31.7% had never microdosed before and participants averaged 5.0 microdoses during the study (SD 3.6, range 1–19); mean inter‑dose interval was 6.7 days (SD 4.8). Daily ratings: Linear mixed‑effects analyses across 98 participants showed a significant increase from Baseline on dosing days (Day0) for all seven daily measures (Connectedness, Contemplation, Creativity, Focus, Happiness, Productiveness, Wellbeing), all p < .001. These increases were generally not maintained on Day1 (the day after dosing). Two days after dosing (Day2) Focus (t = 2.31, p = .044) and Productiveness (t = 2.61, p = .019) showed small but significant elevations relative to baseline. Conditional R2 values were reported as relative indices of variance explained; the authors caution interpretation of effect‑size estimates in mixed models. Long‑term psychometrics (n = 63): Analyses of baseline versus post‑study measures identified several significant Time effects. Depression decreased (DASS Depression; t = −3.97, p = .001) and Stress decreased (DASS Stress; t = −3.36, p = .004). Mind wandering (MWQ) decreased (t = −2.49, p = .047), indicating improved attentional stability. Absorption (TAS) increased (t = 4.46, p < .001). A small increase in Neuroticism (M5P) was observed (t = 2.70, p = .027). There were no significant Time x Experience or Time x Doses interactions, suggesting changes were not significantly moderated by prior microdosing history or the number of doses during the study. Null results were reported for mindfulness, mystical experience, overall quality of life (QOLI), creativity, and most personality domains; some trends (e.g. Effortlessness on SOARS, Time x Doses for Conscientiousness) were noted but not significant after correction. Exploratory debrief: Rescaled 1–100, the mean personal meaningfulness rating (Q1) was 38.95 (SD 25.71), spiritual significance (Q2) 31.69 (SD 27.86). A single item asking whether microdosing changed personal wellbeing (Q3) gave a mean of 78.55 (SD 18.05), with 87.09% reporting some increase in subjective wellbeing, though this item did not correlate with the post‑study QOLI total score (t(57) = 1.62, p = .110). Study Two—expectations (n = 263): The sample included 152 (57.8%) naïve and 111 (42.2%) experienced microdosers. Weighted expectation scores (direction × confidence) were highly similar for naïve and experienced participants (Kendall's τb = .934, p < .001). Participants expected significant change (increase or decrease) for all queried variables (all p <= .01). Specifically, they expected decreases in depression, anxiety, stress, mind wandering, neuroticism and involuntariness; and increases in quality of life, mindfulness, mystical experiences, extraversion, agreeableness, conscientiousness, openness, absorption, creativity and effortlessness. Expectations aligned directionally with several Study One effects (decreased depression, decreased stress, decreased mind wandering, increased absorption) but contradicted the observed increase in neuroticism. Crucially, the rank order of expected effects did not correlate with the rank order of observed changes in Study One (Kendall's τb = .283, p = .139), indicating that the most strongly expected changes (e.g. creativity, wellbeing, mindfulness) did not map onto the largest observed effects.

Polito and colleagues interpret the pattern of findings as indicating that microdosing is associated with short‑lived boosts in a broad set of self‑reported psychological functions on dosing days, with limited carryover into the following days. Over a six‑week observational period, they observed decreases in reported depression and stress, reduced mind wandering, and increased absorption, alongside a small increase in trait neuroticism. The authors note that the domains showing longer‑term change are not the ones most heavily emphasised in media and online narratives (for example, creativity and overall wellbeing did not show reliable increases on standardised measures), suggesting a partial disconnect between popular expectations and systematic self‑reports. The expectation survey (Study Two) revealed strong, consistent beliefs among both naïve and experienced community members that microdosing produces broad benefits. However, the lack of correspondence between the rank order of expectations and the observed pattern of change in Study One argues against a simple expectancy‑only account of the findings. The authors also highlight a surprising absence of dose–response relationships: changes were unrelated to the number of doses reported during the study and to prior microdosing experience, a result the investigators caution may reflect imprecision in self‑reported dose and frequency rather than a genuine lack of dose dependence. Several limitations are emphasised. This was an observational, self‑report study subject to sampling bias (recruitment from pro‑microdosing online communities), potential inaccuracies in participants' dose estimates, and some concurrent use of higher‑dose psychoactive substances. The absence of a placebo or controlled dosing arm means expectancy and other non‑pharmacological influences cannot be definitively ruled out. The authors therefore call for controlled, dose‑specified experimental research—ideally with placebo conditions—to disentangle pharmacological effects from expectancy and to characterise safety, dose–response relationships and individual differences in response. They also note the legal and bureaucratic barriers that constrain such research. In terms of implications, the investigators suggest microdosing may produce modest improvements in mental health and attentional stability for some individuals but also that not all effects are positive—highlighted by the unexpected increase in neuroticism and by qualitative reports of intensified negative emotions in a minority of participants. The paper concludes that the most promising avenues for future confirmatory research are the effects on depression/stress, attentional capacities (mind wandering and absorption), and the conditions under which adverse responses occur; rigorous, placebo‑controlled trials with well‑characterised dosing are required to test these possibilities further.