Rapid and sustained antidepressant effects of vaporized N,N-dimethyltryptamine: a phase 2a clinical trial in treatment-resistant depression

Introduction

Major Depressive Disorder (MDD) is a prevalent mental health condition globally affecting over 185 million people [1]. Traditional antidepressants often take over 3 weeks to exert their effects [2], which leaves patients vulnerable to persistent depressive symptoms and an increased risk of suicide during this latency period [3]. Approximately one-third of patients with depression are classified as Treatment-resistant depression (TRD) [4], with only 50% responding to the first antidepressant, response rates dropping to 30% with a second treatment and just 10% for subsequent therapies [5]. The high prevalence and the limitations of current treatments underscore the urgent need for novel, effective, and scalable therapeutic approaches that can be integrated into public health systems [6, 7].

Serotonergic psychedelics such as psilocybin, ayahuasca, and N,N-dimethyltryptamine (DMT) have shown promising evidence of rapid antidepressant effects in patients with MDD [8–13]. Our research group has conducted clinical trials for TRD using ayahuasca [8, 14, 15], a DMT-containing brew traditionally used by Amerindians for medicinal and spiritual purposes for centuries [16, 17]. Evidence from our open-label and randomized, placebo-controlled trials indicate that ayahuasca can produce rapid antidepressant effects, with improvements starting one day after a single session and lasting at least seven days [8, 14, 15].

However, the long-lasting psychedelic effects of ayahuasca, psilocybin, mescaline, and lysergic acid diethylamide (LSD) typically last several hours [18–21], posing practical and economic challenges for their widespread clinical use. This context highlights the need for rapid-acting and short-lasting alternatives, such as parenteral DMT [22, 23].

DMT, an indole alkaloid, occurs naturally across various species, including plants, fungi, animals, and humans [24]. The compound was first synthesized by Manske in 1931 and subsequently isolated from Mimosa tenuiflora roots by Gonçalves-Lima in 1946 [25]. Although its biosynthesis, metabolism, and biological roles remain partly elusive, DMT is understood to interact with multiple neuroreceptors, notably serotonin 2A, 2C, and 1A receptors, as well as sigma-1 and Trace Amine-Associated Receptors (TAAR) [24–26]. When ingested orally, DMT is rapidly metabolized by monoamine oxidase (MAO), rendering it inactive unless paired with an MAO inhibitor (MAOi), as seen in the traditional brew ayahuasca. This combination induces potent psychedelic effects but also increases the risk of serotonergic syndrome when used alongside conventional antidepressants, making oral administration less ideal for clinical settings. Hence, a parenteral route for DMT administration might offer a more controlled and safer application in therapeutic contexts.

The first explorations into the human effects of isolated DMT date back to the 1950s, with intramuscular administration [27], followed by intravenous studies in the 1990s [28, 29]. These investigations revealed that while DMT induces experiences similar to other psychedelics, such as visual and sensory alterations, its effects have a much quicker onset (2–5 min) and shorter duration (10–60 min) when administered intramuscularly or intravenously [27, 30]. Moreover, a recent study explored its potential antidepressant properties following intravenous administration [11].

However, a non-invasive route might be safer and would simplify the clinical use of DMT; therefore, we first conducted an open-label, single-ascending, fixed-order, dose-response study to assess the feasibility of inhaled DMT in healthy volunteers [31]. Our findings indicated that vaporized DMT mildly and transiently increased blood pressure and heart rate without serious adverse events (SAEs). The effects were dose-dependent, provoking an altered state of consciousness, vivid imagery, body distortions, changes in auditory perception, cognition, mood, arousal, and positive valence, lasting less than 20 min.

The present study aimed to investigate vaporized DMT’s efficacy, safety, and tolerability in patients with TRD. We hypothesized that DMT would provide rapid and sustained antidepressant effects. This could offer a novel therapeutic approach that could be integrated into clinical practice, with significant advantages over traditional antidepressant treatments and better cost-effectiveness compared to other long-lasting psychedelics.

Materials and methods

Results

Discussion

This study aimed to investigate the antidepressant effects, safety, and tolerability of vaporized DMT in patients with TRD, based on the optimized dosing regimen identified in our phase 1 clinical trial [31].

The antidepressant effects of inhaled DMT were rapid. Significant improvements were observed as early as one day post-administration (D1), with a marked reduction in depression severity. The MADRS scores showed a substantial decrease from moderate to severe depression to no clinically relevant symptoms on average by D1. These antidepressant effects were not only immediate but also sustained, with significantly lower depression scores persisting for up to three months post-intervention (M3), the final follow-up period.

On average, patients experienced a reduction of 23.28 points on MADRS by the 1st day after dosing, reflecting a clinical response rate of 78.57% and a remission rate of 71.43%; at D7, the response rate peaked (85.71%). In the 3rd month, there still was a significant reduction, with 57.14% of patients remaining in the response range and 35.71% in the remission range. Notably, over the three months of MADRS evaluations by psychiatrists, the results consistently matched the self-reported PHQ-9 scores. The rapid onset of antidepressant effects observed in our study, with significant reductions in MADRS and PHQ-9 as early as D1, aligns with the emerging paradigm of psychedelic treatments [38] and rapid-acting antidepressants [39]. For comparison, the SYNAPSE study [40] with esketamine (plus oral antidepressants) achieved a 32% remission rate after 74 days of multiple dosages, whereas our study reached 35% remission after 84 days (M3) with a single-day intervention.

An exploratory study on the antidepressant effects of intravenous (I.V.) DMT in seven patients with MDD reported a moderate effect size (Hedge’s g = 0.75) with a mean reduction of 4.5 points on the Hamilton Depression Rating Scale (ranging from 0 to 52) one day after dosage [11]. In contrast, our study achieved a higher effect size (Hedge’s g = 2.76) on the MADRS score at D1, which was partially sustained for three months.

Beyond the administration route (I.V. vs. vaporized), notably, D’Souza et al. [11] utilized a lower dose of DMT (~21 mg), which may account for the more modest antidepressant response observed. However, it is challenging to draw definitive conclusions about the impact of dosage due to the absence of pharmacokinetic (PK) data in both studies. The intravenous route likely provides full bioavailability, while the inhaled route may be influenced by factors such as pulmonary capacity and losses associated with the vaporization device. Both studies assessed subjective experience intensity using a Visual Analog Scale (VAS). D’Souza et al. reported a mean VAS score of 94.3 with 0.3 mg/kg I.V. DMT, whereas we observed a mean score of 84.3 with 60 mg inhaled DMT. Despite these differences, the subjective intensity of the experience was comparable, highlighting the complexity of dose-response relationships in psychedelic treatments. Additionally, our study provided a structured setting with preparation and psychological support, potentially contributing to the stronger effects observed.

A recent study investigated the safety and efficacy of vaporized 5-methoxy-N,N-dimethyltryptamine (5-MeO-DMT) in 16 patients with TRD. At D7 vs. baseline, they found a MADRS reduction of 24.4 points; similarly, we found a reduction of 21.14 points [33]. Both studies highlight the rapid antidepressant effects of short-lasting psychedelic tryptamines using the same vaporizer device. However, our study used a single ascending fixed-dose approach as opposed to an individualized dosing regimen used by Reckweg et al. [33]. The positive outcomes in both studies reinforce the viability of using vaporized short-lasting psychedelics as a promising therapeutic approach.

Ayahuasca, a DMT-containing brew, has also demonstrated rapid and significant antidepressant effects in TRD, with response rates reaching up to 77% and reductions in suicidality observed within 24 h post-administration [8, 41]. These effects have been linked to changes in neural networks involved in emotional regulation and a decrease in inflammatory and stress biomarkers [42–44]. Compared to ayahuasca, vaporized DMT provides a shorter duration of action, avoids the need for MAOi coadministration, and offers advantages in scalability, cost, and clinical applicability while maintaining promising antidepressant potential.

Furthermore, we also observed a positive impact on severe outcomes associated with depression, such as the risk of suicide. In this regard, the results of MADRS-SI and BSI are particularly noteworthy. At baseline, 86% of patients exhibited suicidal ideation, with 43% assessed as being at severe risk. Remarkably, by the day after dosing, no patient demonstrated severe suicidal ideation. This rapid improvement was largely sustained, with only 21% of patients showing any signs of suicidal ideation at the 3-month follow-up. While most participants showed reductions in suicidality, one (P12) worsened to severe SI (from D7 to M1), and others (P6 and P13 at M3, and P14 at D7 and M1) experienced transient changes, with SI returning to baseline by follow-up. Notably, no participants developed severe SI beyond baseline levels.

Additionally, a significant correlation was found between the reduction in depression scores (MADRS-nonSI) and the BSI and MADRS-SI scales at D7, suggesting that the improvement in suicidality was likely mood-dependent rather than an independent effect. Ketamine has been well-established for its rapid antisuicidal effects [45]. Lopes et al. reported a Hedge’s g of 0.5 in suicidality reduction one day after the first ketamine dose, measured by the BSI [46]. Our study found a comparable effect size of 0.67 on D1. For MADRS-SI, Lopes reported a Hedge’s g of 0.5 on D7, while our study observed a stronger effect of 1.34 with DMT. These findings indicate that DMT may have potential as a candidate for further exploration in antisuicidal treatments.

Regarding the psychedelic effects, the onset seemed rapid, emerging within a few seconds, peaking around 2 min, and lasting for 10–20 min; the effects appeared intense and well tolerated, as measured subjectively by the VAS. Participants rated the experience as overall pleasant, with high satisfaction (HRS item 46) and positive valence, while there was a desire to repeat within a week (HRS item 48) and have regular sessions (HRS item 49). The relatively low frequency suggests minimal addiction risk. However, this evaluation is indirect and does not fully address addiction potential. As psychedelics have a low potential for misuse [47], this finding probably points to good tolerability rather than addiction concerns.

Our results on the ASC scale highlighted substantial effects in Complex Imagery and Elementary Imagery, which are hallmarks of DMT experiences [48, 49]. High scores were also observed in Blissful State and Insightfulness, grouped under Oceanic Boundlessness. Scores for Anxiety and Impaired Cognition were low, indicating good tolerability. Spiritual Experience did not show high scores, aligning with MEQ results, where only two individuals rated>60% [50, 51].

In our study, we observed a marginal positive correlation between reductions in MADRS scores with Complex Imagery but no association with MEQ scores. These marginal correlations may reflect variability in absorbed DMT doses due to incomplete inhalation, coughing, or vaporizer deposition. Mystical experiences are associated with sustained therapeutic outcomes in psilocybin treatments [52, 53]. On the other hand, Palhano-Fontes et al. found perceptual (visual) changes, not MEQ scores, to correlate with decreased depressive symptoms after ayahuasca treatment, pointing to a potential variable impact of different psychedelics [8]. These intriguing trends likely lack significance due to the small sample size. PK studies are needed to clarify dose-response relationships and optimize delivery methods.

Regarding visual phenomena, it may confer an ontological sense to the psychedelic experience [54, 55]. Another possible explanation is that these aspects could enhance the psychotherapeutic integration process by providing rich narrative content. The saying “seeing is believing” underscores the significance of vision in affirming reality. Interestingly, closed-eye visuals induced by ayahuasca activate brain regions to a similar extent as regular open-eye vision in ordinary states of consciousness [43], which may enhance therapeutic benefits by amplifying expectancy effects: patients who experience vivid imagery might perceive the treatment as more effective, potentially enhancing the broad therapeutic impact. The efficacy of psychedelic therapies in treating depression is influenced by a combination of factors that underscore their complexity, including the direct pharmacological effects of the substances, patient expectations, and the psychotherapeutic integration of the experiences during session [56].

The physiological findings confirmed that DMT was safe with respect to vital signs. During the acute phase (+0 to +22 min), there were mild, temporary, and self-limiting increases in both blood pressure and heart rate without any clinical consequences. These alterations were similar to those typically seen during moderate-to-vigorous exercise in healthy individuals without cardiac issues [57]. All patients returned to baseline values in sub-acute effects (+30 and +120 min). Regarding biochemical parameters, while some significant changes were noted following DMT administration, none were clinically relevant.

Regarding AEs, no serious AE was reported. The most commonly reported AEs fell into three categories: Respiratory, Thoracic, and Mediastinal Disorders, Nervous System Disorders, and Gastrointestinal Disorders across all time points. Most AEs occurred during the acute phase, encompassing both dosing sessions (1 and 2), which accounted for most of the complaints. The most frequently reported single medical concept was pharyngeal discomfort, followed by cough on D0 and headache on D1. Pharyngeal discomfort included symptoms such as burning throat, throat clearing, throat discomfort, and scratchy throat. This effect was reported in 57.1% of sessions on D0 (16 out of 28 sessions) but resolved for all but one participant by D1 and D7. These results align with our phase 1 trial [31], where pharyngeal discomfort occurred in 48.1% of sessions (26 out of 54 sessions) but did not persist beyond D14, the next medical evaluation. These findings suggest an irritative effect of inhaled DMT vapor, primarily restricted to the acute phase and rarely persisting [31].

This trial has some limitations. The open-label design and small sample size present a major caveat, as it does not allow for the disentanglement of placebo effects from DMT effects. The need for repeated doses over time was not assessed, and longer follow-up periods are desirable. Also, the absence of PK data hinders understanding of dose-blood concentration relationships. The characterization of TRD relied partly on self-reported treatment histories, which may lack precision, particularly for self-referred participants. The sample’s homogeneity in ethnicity, educational, and socioeconomic status limits generalizability. Furthermore, the lack of comprehensive cardiac assessments, such as QT interval monitoring using a 12-lead ECG, represents another limitation. Similarly, while implemented to minimize risks, the upper age limit of 60 years restricts the generalizability of findings to older populations with potentially higher cardiovascular risk. While most participants experienced therapeutic benefits, individual variability in response was observed. One participant (P12) showed symptom worsening at D7, closer to the intervention, suggesting a possible temporal association with the treatment. Conversely, another participant (P13) exhibited worsening only at M3, likely influenced by external life events rather than the intervention itself. These findings emphasize the importance of contextualizing individual outcomes and considering extra-pharmacological factors in the evaluation of psychedelic treatments. Future studies should aim to identify predictors of both positive and negative responses, integrating life circumstances into their analyses to improve patient selection and ensure safety.

Future directions include increasing participant numbers, a more diverse population, and employing randomized, placebo-controlled designs for more robustness. Future studies will also evaluate the impact of cumulative dosing regimens as a strategy to sustain and potentially enhance antidepressant responses over longer periods and should focus on identifying predictors of both positive and negative responses to improve patient selection and safety monitoring. Additionally, extending psychological integration sessions may help reinforce therapeutic gains. Enhanced cardiac safety monitoring, including QT interval assessments and validated addiction risk assessments, will provide a more comprehensive understanding of the risks, along with PK evaluations to clarify the relationship between administered and absorbed DMT doses, enabling optimization of inhalation procedures and dose standardization. To further refine the DMT administration protocol to optimize session duration, evaluations of tachyphylaxis should test repeated DMT administrations at shorter intervals (e.g., less than 30 min apart), allowing for a condensed procedure of 1–2 h.

The integration sessions on D1 and D7 were designed to support meaning-making, sustain benefits, and reinforce positive patterns to prevent relapse. While we hypothesize that these sessions contribute significantly to the therapeutic response, our study design does not isolate their effects from those of the DMT session itself. Future research will aim to disentangle these contributions, further refining the balance between pharmacological and psychological components in interventional psychiatry.

Our findings suggest that inhaled DMT has a rapid antidepressant and antisuicidal effect, is safe and well-tolerated. Vaporized DMT offers a noninvasive alternative suitable for outpatient care, with advantages over longer-acting psychedelics like psilocybin and LSD, such as shorter sessions and lower healthcare costs. Notably, DMT can potentially be used without the need to discontinue traditional antidepressants, which enhances its practicality by avoiding the need for tapering. These features make it a practical and scalable option for treating mood disorders within public health systems, addressing the high prevalence of depression and the ongoing need for more accessible treatments.

Supplementary information

Author contributions

MFC, FPF, IW, NGC and DBA designed the study; EA, FPF and DBA acquired the authorizations; SRBS, EJP, and JVCM extracted and purified the substance; MFC, FPF and DCB selected the participants; MFC led the session, administered the substance and served as head of psychiatry for the study; HB, RB, and SL provided psychological support; DM, ET, and RFV served as psychiatrists to medical follow-up; RA, RKAM, and FA provided nursing support; MFC, HB, RB, SL, DM, ET, RFV, RA, RKAM, and FA acquired the data; FPF, IW, NCG, and DBA provided research assistance; JACN, LAACN, LFNF, LDC, RBCB preprocessed the data; FPF analyzed the data; MFC, FPF, IW and DBA wrote the article; all authors reviewed the manuscript and approved the final version.

Funding

We acknowledge the institutional support of the Federal University of Rio Grande do Norte (UFRN), the University Hospital Onofre Lopes (HUOL), the CAPES Foundation, and the National Council for Scientific and Technological Development (CNPq) for providing support for this work. This study did not receive any specific funding and was primarily supported by the voluntary participation and efforts of the co-authors.

Data availability

The data that support the findings of this study are available on request from the corresponding authors, FPF and DBA.

Competing interests

DBA has served as a scientific and clinical advisor for Biomind Labs from June 2021 to December 2022. MFC has served as the Head of the Psychiatric Research Unit at Biomind Labs Inc. from January 2022 to May 2023; both positions concluded before the start of the study on October 9, 2023, and no other relationships or activities could appear to have influenced the submitted work. The remaining authors have nothing to disclose.

Supplementary information

The online version contains supplementary material available at 10.1038/s41386-025-02091-6.