Mapping the phenomenology of intranasal 5-MeO-DMT in psychedelic-naïve healthy adults

Design and participants: The data derive from a Phase I, double‑blind, randomised, placebo‑controlled single ascending dose trial of intranasal BPL‑003 (5‑MeO‑DMT benzoate dry powder; 1–12 mg) in psychedelic‑naïve healthy adults aged 25–55 years. Fifty volunteers were enrolled; inclusion/exclusion criteria are reported in supplementary material. The paper also includes qualitative data from an additional open‑label cohort receiving a modified 12 mg formulation. Dosing context and acute measurements: Dosing took place in a controlled, supportive environment with a nurse and one psychedelic monitor present. Participants performed preparatory breathing exercises, listened to a prepared playlist during sessions, and both participant and monitor rated subjective drug intensity (SDI) on a 0–10 Likert scale every 2 minutes for up to 90 minutes post‑dose. If participants were unresponsive a rating of 10 was recorded. Microphenomenological interviews and qualitative analysis: Consenting participants took part in one‑to‑one microphenomenology interviews approximately 2 hours after dosing, conducted in person or by video call, and audio recorded. Microphenomenology is a semi‑structured guided‑recall method that focuses on how an experience unfolded in time and on fine‑grained sensory, emotional and cognitive details. Interviews were transcribed verbatim and de‑identified. The researchers used an inductive, microphenomenology‑inspired analytic approach adapted to the longer duration of 5‑MeO‑DMT sessions: transcripts were chronologically reorganised, minimal meaning units ('descriptemes') were identified, clustered into categories, and iteratively reviewed by two researchers until consensus; all authors approved the final map of categories. Quantitative NLP analysis: Participant utterances (interviewer speech excluded) formed a corpus of 9,774 utterances. Sentences were embedded using a pre‑trained Transformer model (NV‑Embed‑v2 producing 4096‑dimensional embeddings). Two complementary topic modelling strategies were applied. In the unsupervised pipeline, embeddings were reduced with UMAP and clustered with HDBSCAN; clusters were curated to remove non‑phenomenological material and labelled using c‑TF‑IDF distinctive terms and concise labels generated by Llama 3 70B. The extracted text inconsistently reports the resulting topic count (it states both 28 and 42 topics in different places); the exact number is therefore unclear in the extraction. In the supervised pipeline, semantic similarity (cosine similarity) between each sentence embedding and researcher‑defined reference sentences (derived from the qualitative analysis) was computed; similarities below 0.50 were discarded, remaining values normalised to 0–1 and averaged per participant to yield theme scores. Temporal modelling: where participants' recall of intensity could be verified against recorded SDI traces, utterances were anchored to 2‑minute timestamps to model the temporal evolution of topic prevalence. Dose ranges for descriptive dose‑dependency analyses were defined as Low (1–4 mg), Medium (6–8 mg) and High (10–12 mg).

Sample and interviews: Of 50 enrolled participants, 40/50 (80%) were interviewed; the extracted text reports that 32 of these had received 5‑MeO‑DMT and 8 received placebo. Four participants declined interviews and one cohort lacked interviewer availability; demographic details are presented in the paper's tables (not fully reproduced in the extraction). Time course and intensity: SDI ratings and interview reports converged on a consistent temporal pattern. Initial effects began within 0–2 minutes post‑administration, intensity typically escalated from 2 to 8 minutes, and peak intensity occurred around 8–15 minutes. A variable plateau followed and effects usually dissipated between 15 and 40 minutes, with consciousness‑altering effects largely ending by 45–60 minutes and some residual effects persisting up to 90 minutes. Higher doses tended to be associated with somewhat longer peaks and more protracted dissipation, but there was notable interindividual variability (some low‑dose participants reported intense effects and some high‑dose participants reported mild effects). Qualitative themes and experiential structure: The analysis identified core experiential domains including sensory and bodily sensations, emotional modulation, cognition, metacognition, arousal and perceived relations between self, others and environment. Typical onset (0–2 min) involved nasal/administration sensations (burning, stinging, blocked nose), which faded as psychedelic effects emerged. During 2–8 min intensity increase many participants initially retained partial awareness of external environment and body, but awareness diminished as intensity rose; prominent early themes included altered bodily control, feelings of separation from the body, and variable anxiety related to loss of control. The peak (8–15 min) was marked by intense, often wave‑like experiences across emotional, physical and perceptual domains: low doses tended to produce primarily affective changes, whereas higher doses more often elicited mystical experiences, ego‑dissolution or non‑duality, void states, rebirth, visions, memory reliving and extreme emotions. Time perception was universally altered and many participants described the experience as difficult to put into words. Emotions and self‑experience: Emotional phenomenology ranged from relaxation and contentment at low doses to ‘‘immense’’ emotional intensity at higher doses, including sadness, grief, joy and relief; some participants reported reliving traumatic memories. Most participants reported altered spatial self‑consciousness (feeling in a void, ocean, universe or white/black space) and either profound aloneness or a sense of unity/connectedness; complete ego‑loss was uncommon, with many retaining some sense of self even at high intensity. Bodily sensations were prominent: intense experiences often involved loss of bodily sensation, while lower intensities featured somatic pleasure, synaesthesia‑like experiences and strong breath‑related phenomena. Cognition and metacognition: High‑intensity peaks commonly involved reductions or absence of thought and ineffability; medium and lower doses preserved more metacognitive capacity, allowing participants to observe and attempt to regulate their experience (e.g. breathing, positive thoughts, therapeutic touch). Surrendering or ‘‘letting go’’ was repeatedly identified as a key process associated with more positive experiences; resistance to surrender tended to coincide with more challenging sessions. Return and reflection phases: As effects waned (15–40 min) participants gradually regained bodily and environmental awareness, agency and cognitive control. Many reported lingering connectedness, gratitude, or bittersweetness, and some described concrete reflections on life priorities. End‑phase reports (40–90 min) typically described relaxation, calmness and/or ordinary thoughts about plans and daily life; a minority reported disappointment when their expectations were not met. Quantitative topic modelling findings: The unsupervised topic analyses identified a set of content‑relevant topics that broadly aligned with the qualitative themes (visual patterns/colours, intensity, physical sensations, emotions, metacognition). The extracted text inconsistently reports the number of identified topics (both 28 and 42 are mentioned); irrespective of the precise count, the authors report dose‑dependent trends: topics referencing visual phenomena, positive emotions and overall intensity tended to increase with dose, while themes such as bodily relaxation, fear and discomfort were reported more at lower doses. The authors emphasise that visual reports largely reflected elementary perceptual changes (light, colour, simple patterns) rather than complex hallucinations. Dynamic topic modelling, using temporally anchored utterances, captured the rapid somatic onset and the timing of different experiential themes across the session. Supervised modelling and temporal phenotypes: Semantic similarity analyses to predefined reference sentences corroborated individual variability and suggested dose‑related increases in the expression of key themes. The time‑resolved analyses yielded three exploratory temporal phenotypes proposed by the authors: (1) somatic‑dominant onset phase (0–5 min) with peak physical discomfort and respiratory awareness; (2) a cognitive‑emotional transition (8–20 min) in which a bimodal pattern of difficulty surrendering emerged (peaks ~9–10 and 19–21 min); and (3) a late‑phase emotional processing period (28–35 min) with resurgence of crying and sadness as lucidity returned. The authors present these as hypothesis‑generating temporal markers that could inform clinical support strategies (for example, enhanced breathing guidance during minutes 2–10 and targeted facilitation of surrender at the identified resistance points).

The authors interpret their findings as the first systematic microphenomenological characterisation of acute intranasal 5‑MeO‑DMT experiences in a clinical trial context, highlighting a rapid onset, short duration, pronounced somatic and emotional effects, and the prominence of surrendering or ‘‘letting go’’ as a determinant of experiential quality. They report that challenging episodes, though distressing, often afforded opportunities for emotional breakthroughs or therapeutic insight when participants felt supported by preparation and monitor presence. The researchers position these results relative to other psychedelics by noting both overlaps (mystical experiences, ego‑dissolution, emotional catharsis) and distinguishing features: 5‑MeO‑DMT's brief time course, predominance of interoceptive/somatic effects over complex visual imagery, infrequent entity encounters compared with N,N‑DMT, and strong emotional intensity. They suggest these phenomenological differences could have implications for therapeutic applications and for designing support protocols. The authors emphasise methodological strengths: the use of microphenomenology shortly after dosing for fine‑grained recall, a relatively diverse, psychedelic‑naïve sample, and the novel combination of qualitative interviews with NLP topic modelling, which they argue provided objective validation and temporal enrichment of the qualitative map. They note several limitations acknowledged in the paper: the Phase I dose‑response design is underpowered to draw definitive dose comparisons; healthy volunteer sampling limits generalisability to clinical populations; interviews were conducted both online and in person (possible but unquantified effects); participants' introspective ability varied; and the longer duration of 5‑MeO‑DMT sessions required adaptation of a method typically applied to very short experiences. The authors frame many of their temporal and dose‑dependent findings as exploratory and hypothesis‑generating, and they recommend future studies combining microphenomenology with neurophysiological measures (EEG, HRV) and testing in patient populations to elucidate neural correlates and potential therapeutic mechanisms. They also highlight practical implications for trial preparation and monitoring: emphasising expectation management, breathing guidance and the importance of supportive monitoring to facilitate surrender and mitigate acute anxiety.

O. and colleagues conclude that microphenomenological interviews combined with NLP provide a detailed diachronic and synchronic map of acute 5‑MeO‑DMT effects in psychedelic‑naïve healthy adults. Key characteristics include rapid onset, short duration, intense emotional and somatic experiences, alterations in self‑experience, and the central role of surrendering to the experience. Challenging experiences frequently co‑existed with opportunities for emotional processing and insight. The authors present these findings as a foundational phenomenological characterisation to inform future mechanistic, clinical and methodological research on 5‑MeO‑DMT, recommending integration with neurophysiological measures and exploration in clinical populations.