How Low-Doses Of Psychedelics Compliment High Dose Experiences: Observational Evidence In Naturalistic Settings

Psychedelics produce marked changes in consciousness through effects on serotonergic and other neurotransmitter systems, and renewed clinical interest has focused on their therapeutic potential for disorders such as depression, anxiety, and substance dependence. Parallel to clinical research, a distinct practice of repeated low-dose usage—commonly called microdosing—has emerged in lay and online communities; practitioners typically report sub-perceptual dosing intended to enhance mood, cognition, creativity and workplace performance. Most evidence about microdosing to date is observational, with a small number of experimental studies that sometimes report physiological or neuroelectric changes even when subjective perceptual effects are absent. This paper by Ferenstein responds to two gaps. First, the literature tends to treat low and high-dose psychedelic use as separate strategies, whereas many users report combining them strategically. Second, forthcoming regulatory changes in parts of the United States will expand non-clinical facilitation of psychedelic experiences, making it important to understand how low-dose education and coaching are being used in real-world settings. The study aims to characterise how people describe low-dose (and higher-dose) psychedelic use in relation to wellbeing, relationships and work performance, using detailed interview material and natural language processing to map semantic patterns across accounts.

The study is a secondary, mixed qualitative–computational analysis of semi-structured telephone interviews collected by a private-sector company that provides psychedelic education and coaching. Participants were users of the company who volunteered feedback via various outreach methods including email, Slack and snowball sampling. The extracted text does not give a formal inclusion/exclusion checklist or a demographic breakdown of participants. Interviews lasted 30–90 minutes and probed the impact of low-dose (micro) practices on job performance, productivity, behaviour and mental wellbeing; examples of questions included frequency and dose of psilocybin or LSD and perceived effects on work and relationships. Audio recordings were transcribed using a commercial automatic service (Otter.ai) and then manually reviewed and corrected by a trained transcriptionist. Interviewer turns were excluded from the corpus during text processing. Text pre-processing and analytic steps were performed in R using quanteda and related packages. The authors tokenised text into unigrams and bigrams, lowercased tokens, removed numbers, punctuation and a customised stopword list, and applied stemming. They excluded the top 20% most frequent tokens and rare tokens present in fewer than the 10th percentile of interviews to improve signal-to-noise. For topic discovery, the investigators combined unsupervised and semi-supervised methods: unsupervised topic models and word-embedding based summarisation were used for exploration, while the reported analyses emphasise results from seeded Latent Dirichlet Allocation (LDA). Seeded LDA incorporated a theory-driven list of psychological themes and seed words derived from a character-strengths framework; the extracted text indicates there were 24 psychological themes but does not specify the exact number of unique seed words. Document weights from seeded LDA were compared with unsupervised structural topic model outputs using Spearman rank correlations. The corpus was constructed at the paragraph/turn-taking level so that segments of interviews describing different dose experiences could be analysed independently.

The analysable dataset comprised 41 coded interviews. Reported psilocybin dosages ranged broadly—from about 0.1 grams to 5 grams of dried mushrooms—so the authors operationalised a pragmatic dichotomy for coding: psilocybin reports of 600 mg or below were classified as "low" and anything larger as "high". For LSD, the cutoff for "low" was set at 30 micrograms. The authors note they intentionally used the term "low" rather than "micro" because some participants reported perceivable drug effects during their low-dose routines. Qualitative excerpts and thematic coding indicate a pattern in how participants used different dosing strategies. High-dose experiences were most often associated with spiritual insight, belief re-appraisal and processing of deeper mental-health issues; examples include retreat or ayahuasca ceremonies that participants described as "powerful" and effective for releasing grief. Low-dose regimens were reported to be beneficial for behavioural change, relationship development and managing episodic stress, and to support sustained changes through repeated application. In several narratives a high-dose experience appeared to catalyse insight, while subsequent low-dose use helped to integrate and operationalise that insight in day-to-day contexts such as parenting or management. One participant described a high-dose sense of "interconnection with all living beings" and then said microdosing allowed her to translate that into more tolerant managerial behaviour. Substance-specific distinctions emerged in the interviews. Psilocybin was more commonly reported to improve interpersonal relations and conflict resolution in both personal and professional domains; respondents described increased empathy, vulnerability and outreach to colleagues. By contrast, LSD was frequently associated with sustained individual focus and solitary cognitive work, such as writing and deep research. Participants who used both substances contrasted a "people lens" on psilocybin with a more intellectual, problem-focused stance on LSD. The seeded LDA analyses produced keyword sets consistent with these qualitative themes, though the authors emphasise the models were noisy given the open-ended interview material. A low-dose topic seeded with terms like "micro" and "microdosing" returned terms such as "creative", "writing" and "working". A high-dose topic seeded with terms like "macro", "grams" and "retreat" returned spiritual or insight-related keywords such as "understood", "shamanic", "powerful" and "ayahuasca". In behavioural calibration, about 20% of participants reported at least one low-dose attempt had no perceptible effect, often during dose-finding adjustments.

Ferenstein frames the findings as evidence that low-dose and high-dose psychedelic experiences are frequently used in complementary ways: high doses tend to be described as facilitating deep insight and belief reappraisal, while low doses are reported to support the translation of those insights into sustained behavioural change, improved relationships and better management of episodic stress. The authors highlight substance-specific roles in users' accounts—psilocybin for enhancing interpersonal empathy and workplace relations, and LSD for enhancing focused, solitary cognitive work—and suggest these differences may help explain why users report productivity gains in the context of improved mental health rather than as purely pharmacological enhancement. The paper situates these results within broader social and regulatory developments, noting that state-level reforms and a growing market for low-dose education could shape how large numbers of people access psychedelics outside formal medical contexts. The authors caution that their NLP models were noisy and that the present study did not test psychological mechanisms experimentally. They therefore call for further clinical validation and more rigorous research on low-dose regimens as adjuncts to high-dose therapy, particularly regarding workplace stress, integration practices and the intersection of mental health and job functioning. The extracted text does not provide an extensive limitations section beyond noting noise in the topic models and the absence of mechanistic testing, and the authors emphasise that more research is warranted to substantiate and clarify these observational findings.