Acute dose-dependent effects and self-guided titration of continuous N,N-dimethyltryptamine infusions in a double-blind placebo-controlled study in healthy participants

N,N-dimethyltryptamine (DMT) is a short-acting serotonergic psychedelic whose acute psychoactive effects are primarily mediated at 5-HT2A receptors. Prior clinical work has most commonly delivered DMT by intravenous bolus or by combining a bolus with a continuous infusion; boluses tend to produce very rapid, intense and sometimes poorly tolerated effects, whereas some continuous infusions without a bolus have produced milder, gradually stabilising effects. However, dose-dependent acute effects and detailed pharmacokinetics of longer continuous DMT infusions across a range of infusion rates remain poorly characterised, and formal testing of within-session self-titration has not been reported in a controlled setting. Erne and colleagues therefore designed a double-blind, placebo-controlled, within-subject crossover study to map acute subjective, autonomic, endocrine and pharmacokinetic responses to four continuous intravenous DMT infusion rates (0.6, 1.2, 1.8 and 2.4 mg/min) administered over 120 min, and to evaluate a subsequent open-label, self-guided titration session in which participants could adjust dose rate at prespecified times to reach a desired level of subjective effect. The study aimed to inform dose selection for future research and to test whether acute subjective states can be rapidly and safely modulated by within-session dose adjustment.

The investigators conducted a double-blind, placebo-controlled, crossover study with five randomized sessions per participant (placebo and four DMT infusion rates: 0.6, 1.2, 1.8, 2.4 mg/min), followed by a non-randomized sixth session for self-guided titration. Twenty-two healthy adults (11 men, 11 women; mean age 30 ± 5 years, range 25–45) completed the protocol. Washout intervals between sessions were at least 7 days. Inclusion/exclusion criteria were standard for healthy volunteer psychedelic studies and included screening for psychiatric history, drug use, pregnancy and significant medical issues; most participants had some prior exposure to psychedelic or recreational drugs. DMT hemifumarate of confirmed high purity was prepared under Good Manufacturing Practice. For randomized sessions, infusions delivered a total volume of 50 ml over 120 min starting at t = 0; the non-randomized titration session used a similar volume but began at 1.2 mg/min and permitted dose adjustments. In the self-titration session participants could change the infusion rate at 40 and 80 min by ±0.2, ±0.4 or ±0.6 mg/min, with a maximum allowed rate of 2.4 mg/min. Placebo vials contained sterile water and were indistinguishable from active vials; at the end of sessions participants retrospectively guessed their assignment to assess blinding. Study sessions lasted about 5 h and were performed in a calm hospital room with two investigators present. Repeated measures included subjective effect ratings (0–10 Likert scales for "any drug effect", "good drug effect", "bad drug effect", "anxiety"), the 5-Dimensions of Altered States of Consciousness (5D-ASC), and multiple mystical-experience questionnaires (MEQ30/40, PES48), with denser sampling in the titration session. Autonomic measures (blood pressure, heart rate) were taken repeatedly. Adverse effects were recorded using the List of Complaints. Plasma samples for DMT and metabolites were collected frequently up to 180 min in randomized sessions, and concentrations were measured by validated HPLC–MS/MS. Endocrine assays included plasma oxytocin, cortisol and prolactin, and serum BDNF at specified time points. Pharmacokinetic parameters were estimated by non-compartmental methods; two participants who stopped infusions prematurely were excluded from PK analyses. Peak or peak-change values from repeated measures (Emax, ΔEmax) were analysed using within-subject analysis of variance (ANOVA) with dose as the factor, followed by Tukey post hoc tests; significance was set at p < 0.05. Data processing used standard software (Phoenix WinNonlin for PK, R for statistics).

All 22 participants completed the randomized sessions and the self-titration session, with repeated subjective, physiological and plasma measurements collected as planned. DMT produced dose-dependent subjective effects that were distinguishable from placebo and manifested rapidly. For the 1.2, 1.8 and 2.4 mg/min rates, subjective effects rose over the first 20–30 min and reached a plateau that persisted until infusion termination at 120 min; at 0.6 mg/min subjective effects rose only modestly and then declined slightly between 50 and 120 min. After stopping the infusion, subjective effects subsided rapidly and largely resolved within 20 min. A ceiling effect for positive subjective experience was observed around 1.8 mg/min: "good drug effect" and many 5D-ASC and MEQ30 scales did not increase significantly between 1.8 and 2.4 mg/min. In contrast, the highest rate (2.4 mg/min) produced significantly greater "bad drug effect", "fear" and anxious ego-dissolution on the 5D-ASC than lower rates and placebo (p < 0.05). Three infusions were stopped prematurely in two participants because of overwhelming psychedelic effects and intense fear (twice during 2.4 mg/min, once during 1.8 mg/min); both participants had prior psychedelic experience. Autonomic effects were dose-dependent: systolic and diastolic blood pressure increased with dose compared with placebo, and heart rate rose significantly at 1.2 mg/min and above. Only the 2.4 mg/min condition increased overall adverse effect scores relative to placebo. Common complaints across sessions included headache, tiredness, impaired concentration, weakness, hot flashes and palpitations. One serious adverse event potentially related to DMT was an exacerbation of preexisting tinnitus that led to participant discontinuation; another participant withdrew after two sessions due to anxiety about future higher doses. Endocrine responses showed dose-dependent increases in serum prolactin and cortisol: prolactin rose significantly at 1.2 mg/min and above, and cortisol increased at 1.8 mg/min and above versus placebo. Serum BDNF and plasma oxytocin did not change across dose rates, and BDNF did not differ between early and later sessions. Pharmacokinetics were dose-proportional: mean Cmax values were reported as 26, 51, 78 and 105 ng/ml for the 0.6, 1.2, 1.8 and 2.4 mg/min rates respectively, with Tmax around 86–102 min depending on dose. After stopping the infusion at 120 min two distinct half-lives were observed: an early t1/2α of about 6.3–7.1 min and a later t1/2β of about 18–19 min, with the switch from early to late elimination occurring approximately 20 min after infusion end at low-nanogram-per-millilitre concentrations. Plasma concentrations continued to rise from 30 to 120 min and steady-state was not attained at any tested rate. Concentration–effect analyses showed clockwise hysteresis—subjective effects stabilised or declined despite further increases in plasma DMT—consistent with moderate acute pharmacological tolerance during continuous infusion. Blinding was only partially retained: immediately after sessions misclassification rates for placebo and the four active rates were reported (3/22, 5/22, 8/22, 11/22 and 7/22 misclassifications respectively), often confusing adjacent dose levels. In the self-titration session (open-label), participants began at 1.2 mg/min and could adjust at 40 and 80 min; the group mean dose reached about 1.9 mg/min and peak "any drug effect" averaged 8.2 ± 0.5. Approximately one-third increased to the maximum 2.4 mg/min and another third increased to 1.6 mg/min; one participant reduced to 1.0 mg/min and one kept 1.2 mg/min. Self-titration produced strong subjective effects comparable to the blinded 1.8 mg/min condition, with low ratings of "bad drug effect" and minimal "fear." Participants' subjective responses adapted within 5–10 min after dose adjustments.

Erne and colleagues interpret the findings as demonstrating that continuous intravenous DMT infusions produce rapid, dose-dependent psychedelic effects that typically plateau within 20–30 min and decline quickly after infusion cessation. The investigators emphasise a practical ceiling for positive subjective effects at about 1.8 mg/min, whereas the highest tested rate (2.4 mg/min) selectively increased anxious ego-dissolution and negative subjective effects. They therefore suggest that continuous infusions without an initial bolus provide a more gradual onset and better tolerability than bolus administration, which can provoke abrupt, overwhelming experiences. The authors further note clinically relevant pharmacokinetic features: DMT clearance is rapid, with a short early half-life that parallels the rapid decline of subjective effects after infusion stop, and a later, longer half-life likely reflecting distributional kinetics. Plasma concentrations rose throughout the 120 min infusion without reaching steady-state at the tested rates, and concentration–effect hysteresis indicated moderate acute tolerance during continuous administration. These PK/PD observations inform dose selection and modelling considerations and argue that acute effects can be rapidly modulated because of the short elimination half-life. Regarding self-titration, the study is presented as the first formal test of within-session dose adjustment for a psychedelic. When empowered to titrate, participants—most of whom had prior psychedelic experience and were familiar with the study procedures—tended to choose dose paths that produced strong but well tolerated effects comparable to the blinded 1.8 mg/min condition; negative effects were low and subjective effects adapted within minutes of each adjustment. The investigators suggest that gradual, participant-controlled escalation and prior experience may underlie the good tolerability of titration and that this approach could be valuable to individualise dosing in therapeutic contexts, particularly given interindividual PK/PD variability. The authors acknowledge strengths and limitations: a within-subject, double-blind design across four doses with balanced sex representation and standardised psychometric measures are strengths, as is the inclusion of a self-titration paradigm. Limitations include the fact that self-titration was tested only in participants with recent DMT experience, the controlled laboratory setting and the exclusive enrolment of healthy volunteers; responses may differ in patients or in less controlled environments. The investigators propose that lower active doses such as 0.6 mg/min might serve as an "active placebo" in future therapeutic trials to improve blinding, but they stop short of clinical claims beyond the observed acute and PK/PD data.

DMT administered as continuous intravenous infusions over 120 min showed dose-proportional pharmacokinetics and produced rapid, dose-dependent subjective effects that plateaued within about 30 min and dissipated quickly after stopping the infusion. A ceiling for positive subjective effects was observed at 1.8 mg/min, whereas negative and anxiogenic effects increased at 2.4 mg/min. Self-guided within-session titration allowed participants to attain strong but well-tolerated psychedelic effects comparable to the 1.8 mg/min condition, suggesting a potential means to individualise dosing in future research or therapeutic applications.