Psychedelics for major depression—From controlled research settings into broader clinical use

Abstract

Psychedelics have shown promising antidepressant effects in recent phase 2 randomized controlled trials, with phase 3 studies underway. While therapeutic outcomes are encouraging, the underlying mechanisms—from receptor-level pharmacology to subjective experience and context—remain only partially understood. We highlight clinical and mechanistic advances and outline priorities for future research.

Psychedelics have shown promising antidepressant effects in recent phase 2 randomized controlled trials, with phase 3 studies underway. While therapeutic outcomes are encouraging, the underlying mechanisms—from receptor-level pharmacology to subjective experience and context—remain only partially understood. We highlight clinical and mechanistic advances and outline priorities for future research.

Main text

Introduction

Psychedelics are currently experiencing a renaissance in psychiatry, marked by a surge of modern clinical trials. As of June 1, 2025, more than 150 active and recruiting studies are registered on ClinicalTrials.gov, reflecting an unprecedented surge in clinical interest. This revival is fueled by growing evidence that psychedelics may offer rapid and sustained therapeutic benefits—particularly in conditions that have proven difficult to treat using conventional pharmacological approaches, such as treatment-resistant depression (TRD) and addiction.

Among the emerging therapeutic indications, major depression—especially TRD—has been the primary focus of modern clinical trials. The most robust findings have come from phase 2 double-blind randomized controlled trials (RCTs), most prominently involving psilocybin, which have progressed into phase 3 development (e.g., NCT05624268). Recent trials have also begun to explore the antidepressant effects of other psychedelics, including ayahuasca and lysergic acid diethylamide (LSD), with early results suggesting converging clinical potential despite differing pharmacological profiles.

While these trials mark a critical step forward, the mechanisms underlying these clinical effects remain elusive. In this commentary, we argue that while the clinical promise of psychedelics is clear, a more comparative, mechanistic, and context-aware research agenda is urgently needed to enable safe and scalable implementation.

Recent advances in psychedelic clinical trials

Psilocybin

Psilocybin, the psychoactive compound found in Psilocybe mushrooms, has emerged as a leading candidate in psychedelic therapy for depression due to its favorable duration (4–6 h) and promising early efficacy data. In comparison, at (presumably) psychoactive-equivalent doses, other psychedelics such as LSD present with longer-lasting acute effects of 8–11 h.¹ This shorter duration for psilocybin allows for more contained clinical sessions and reduces the logistical burden on both patients and clinicians.

Recent phase 2 double-blind RCTs have shown that psilocybin is generally well tolerated and leads to significant reductions in depressive symptoms in both major depressive disorder (MDD) and TRD (Table 1). While various trial parameters, such as primary outcome measures (i.e., psychometric scales used to evaluate symptom severity, like the Montgomery-Åsberg Depression Rating Scale [MADRS]), the timing of assessments post-administration (e.g., 3 vs. 6 weeks), placebo (e.g., inactive/active or low drug dose), and the number of preparation and integration sessions, varied across studies, the clinical outcomes consistently favored a full psilocybin dose over the comparator (e.g., placebo/low-dose psilocybin). Specifically, both response rates (defined as a 50% reduction in symptom severity from baseline at the primary endpoint) and remission rates (achieving a symptom score below a predefined threshold on assessment instruments) showed clear benefits for psilocybin. Notably, psilocybin was found to be at least as effective as the serotonin-reuptake inhibitor (SSRI) escitalopram in a head-to-head study, even outperforming the antidepressant on several secondary outcomes.² Remarkably, despite the low number of administrations (i.e., typically one or two) longitudinal follow-ups suggest effects can be sustained for 6–12 months.³

Table 1.

Phase 2, double-blind RCTs with psychedelics for MDD and TRD

Clinical trial	Study design	Population	Sample size	Dose	Comparator	Primary endpoint timing	Primary endpoint	Response rate	Remission rate	NCT number
Psilocybin
Carhart-Harris et al.2	RCT, double-blind	MDD	59	2 × 25 mg psilocybin	escitalopram (10–20 mg/day)	6 weeks	QIDS-SR-16	70%	57%	NCT03429075
Goodwin et al.4	RCT, double-blind	TRD	233	25 mg psilocybin	10 mg psilocybin, 1 mg psilocybin	3 weeks	MADRS	37%	29%	NCT03775200
von Rotz et al.5	RCT, double-blind	MDD	55	0.215 mg/kg psilocybin	placebo (mannitol)	2 weeks	MADRS	58%	54%	NCT03715127
Raison et al.6	RCT, double-blind	MDD	104	25 mg psilocybin	100 mg niacin	6 weeks	MADRS	58%	44%	NCT03866174
LSD
Müller et al.7	RCT, double-blind	MDD	61	100 μg LSD (first dose), 100 or 200 μg LSD (second dose)	2 × 25 μg LSD (small dose)	6 weeks	IDS-C	40%	32%	NCT03866252
Ayahuasca
Palhano-Fontes et al.8	RCT, double-blind	TRD	29	0.36 ± 0.01 mg/mL of N,N-DMT, 1.86 ± 0.11 mg/mL of harmine, 0.24 ± 0.03 mg/mL of harmaline, and 1.20 ± 0.05 mg/mL of tetrahydroharmine	Placebo (water, yeast, citric acid, zinc sulfate, and aramel colorant)	7 days	MADRS	64%	36%	NCT02914769

Overview of phase 2, double-blind, randomized controlled trials (RCTs) investigating the efficacy of various psychedelics—including psilocybin, ayahuasca, and LSD—in major depressive disorder (MDD) and treatment-resistant depression (TRD). The table summarizes key trial characteristics including study design, population, sample size, dosing strategy, comparator condition, primary outcome measure, and timing of the primary endpoint, as well as reported response (a ≥50% reduction in depression severity from baseline) and remission rates (a score below a specific threshold on depression rating scales [e.g., MADRS ≤10]). Only trials with a primary endpoint focused on depressive symptomatology were included. QIDS-SR-16, 16-item Quick Inventory of Depressive Symptomatology – self-report; MADRS, Montgomery-Åsberg Depression Rating Scale; HAM-D, Hamilton Depression Rating Scale; IDS-C, Inventory of Depressive Symptomatology – clinician-rated.

While generally well-tolerated, we note that adverse events (AEs), like headache and nausea, are relatively common in psychedelic therapy with psilocybin.¹ More rarely, serious adverse events (SAEs) have also been reported, such as worsening of depression or suicidal ideation. AEs and SAEs may increase in frequency with larger sample sizes, i.e., in phase 3 trials. Importantly, a subset of patients does not respond to psilocybin, prompting interest in other psychedelics with distinct pharmacological profiles.

LSD

LSD, the most potent and widely recognized classic psychedelic, was extensively studied in the 1950s and 60s for its antiaddictive and antidepressant properties. However, until recently, modern clinical data remained scarce: only one contemporary phase 2 double-blind RCT had demonstrated that LSD could reduce anxiety and associated depressive symptoms in patients with anxiety disorders,⁹ leaving its antidepressant potential in MDD largely unexplored.

Müller and colleagues⁷ addressed this gap by conducting the first modern clinical trial assessing LSD’s antidepressant effects in MDD. In a recently published phase 2 double-blind trial in patients with moderate-to-severe MDD, they compared two low-dose (25 μg) LSD administrations with two high-dose administrations (100 μg followed by 100 or 200 μg) spaced four weeks apart. The primary outcome measures were changes from baseline to two weeks after the second dose on both the clinician-rated (IDS-C) and self-rated (IDS-SR) versions of the Inventory of Depressive Symptomatology (IDS). At this primary endpoint, the high-dose group (n = 30) showed a statistically significant greater reduction in IDS-C scores compared to those of the low-dose group. However, when adjusting for baseline depression severity—significantly higher in the high-dose group—the result was attenuated and reached only a trend-level significance (p = 0.08). Despite this attenuation, the consistent pattern across multiple outcome measures suggests potential clinical relevance, warranting further investigation in larger trials. Secondary endpoints, including other time points on the IDS-C/IDS-SR, the Beck Depression Inventory, and the Symptom Checklist-90, consistently favored the high-dose group. Notably, reductions in depressive symptoms persisted for up to 12 weeks post-treatment. A challenge in interpreting these findings lies in disentangling specific therapeutic effects from nonspecific factors such as expectancy and contextual influences. The study did not include an inert placebo, and while the 25 μg dose was intended as a low, presumably subtherapeutic comparator, it elicited unexpectedly strong acute subjective effects. As the authors acknowledge, this may have contributed to symptom improvement in the low-dose group beyond what would be expected from a traditional placebo. However, it also raises the broader possibility that some of the observed antidepressant effects—across both groups—may have been driven in part by nonspecific psychological mechanisms activated by the psychoactive experience itself. While blinding in this study was better than in previous psychedelic trials, the use of an active comparator with noticeable effects still complicates the separation of pharmacological efficacy from expectancy-driven improvement. Future studies would benefit from including both pharmacologically inactive and carefully titrated low-dose comparators to better disentangle these effects.

Importantly, LSD was generally well tolerated, with adverse events comparably distributed across groups. Nonetheless, four SAEs were reported during the trial. Of these, three were considered possibly related to the intervention, involving worsening of depressive symptoms—two in the low-dose group and one in the high-dose group. Additionally, two participants in the high-dose group discontinued participation after the first administration due to challenging subjective experiences. Individuals with heightened sensitivity to prolonged and challenging psychedelic experiences may therefore benefit more from shorter-acting or less intense alternatives.

These findings suggest LSD may be effective for some patients, but its extended duration and intensity may limit its generalizability. A phase 3 trial is underway (NCT06941844).

Ayahuasca

Ayahuasca has also been shown to be effective in reducing depressive symptoms in patients with TRD.⁸ The Amazonian brew combines several plants, including the N,N-dimethyltryptamine (DMT)-containing Psychotria viridis and Banisteriopsis caapi, rich in monoaminoxidase inhibitors—like harmine and harmaline—which enable the oral activity of DMT.

A phase 2 clinical trial with ayahuasca in 29 TRD patients demonstrated significant reductions on the MADRS one week post-administration.⁸ Response rates were significantly higher in the ayahuasca group, and remission rates showed a trend toward significance at one week post-administration (p = 0.054).

Despite a similar duration of action to psilocybin, ayahuasca is associated with frequent gastrointestinal AEs, including nausea and vomiting—observed in over half of the above-mentioned trial’s participants. However, no SAEs were reported in this trial. Nevertheless, the complexity of ayahuasca preparation and administration limits its scalability. Ongoing pharmaceutical efforts (e.g., NCT04716335) aim to develop synthetic, dose-controlled formulations of DMT and harmine to standardize treatment delivery.

Mechanisms of action: Current hypotheses and challenges

Receptor targets and downstream effects

The precise mechanisms by which psychedelics exert antidepressant effects remain incompletely understood. A central hypothesis implicates agonism at the serotonin 2A receptor (5-HT2AR), a shared pharmacological property of all classic psychedelics.¹ Supporting this view, preclinical work has shown that 5-HT2AR activation facilitates neuroplasticity, a key process in recovery from depression. However, alternative routes to neuroplasticity have been proposed. For instance, psychedelics also upregulate brain-derived neurotrophic factor (BDNF) and activate its receptor TrkB, a mechanism that may be independent of 5-HT2AR agonism¹⁰; however, these findings require further confirmation. This raises the question of whether psychedelic-like benefits could emerge in the absence of psychedelic experiences. Trials are now testing this idea by co-administering 5-HT2AR antagonists such as risperidone to block subjective effects while preserving downstream plasticity (e.g., NCT05710237). If successful, such strategies could broaden the applicability of psychedelic therapy to populations for whom the psychedelic state is contraindicated, including those with psychosis spectrum disorders.

Importantly, while 5-HT2AR agonism is a unifying feature, psychedelics vary widely in 5-HT2AR binding affinities—sometimes by orders of magnitude. Furthermore, each psychedelic has a distinct neuropharmacological profile with activity at several receptors (Figure 1). Whether these differences merely influence duration and intensity or contribute to therapeutic outcomes remains unknown. To date, no clinical trial has directly compared antidepressant efficacy across various psychedelics, making it unclear whether compound-specific pharmacodynamics shape clinical response.

Figure 1.

Receptor affinities of classic psychedelics with clinically demonstrated antidepressant effects

The chart is displaying the relative binding affinities of LSD, psilocybin, and N,N-dimethyltryptamine (DMT) across selected serotonergic and non-serotonergic receptors and transporters. Data points reflect published in vitro affinities for several serotonergic, dopaminergic, and other relevant receptor sites (reviewed in Holze et al.¹). Higher values represent stronger affinities (i.e., lower Ki). This receptor diversity may contribute to compound-specific therapeutic and subjective effects.

In light of recent developments, interest has grown in so-called short-acting, ultra-fast psychedelics such as intravenous or inhaled DMT and 5-methoxy-N,N-dimethyltryptamine (5-MeO-DMT). Preliminary evidence suggests that very brief, yet intense, interventions (15–30 min) are well tolerated and may yield rapid antidepressant effects (for a review see Ramaekers et al.¹¹). Larger phase 2 RCTs are currently ongoing (e.g., NCT05870540).

In brief, the assumption that 5-HT2AR agonism fully accounts for therapeutic outcomes is no longer tenable. Future studies must explicitly test the relevance of alternative pathways, including head-to-head studies comparing distinct psychedelics, non-hallucinogenic analogs, and downstream plasticity mechanisms.

Neural signatures and treatment outcome biomarkers

Understanding how psychedelics exert their antidepressant effects at the systems level remains a key translational challenge. Functional neuroimaging provides a valuable window into these mechanisms, offering in vivo insights into changes in functional connectivity, network dynamics, and neuroplasticity. While most existing studies have been conducted in healthy volunteers, early findings suggest several consistent principles relevant to therapeutic action.

One such principle is the disruption of thalamic gating, reflected in increased thalamocortical connectivity, particularly with sensorimotor cortices.¹² This effect may enhance the flow of bottom-up information and reduce the dominance of top-down priors, potentially supporting therapeutic effects. Another frequently reported phenomenon is reduced modularity of large-scale networks—particularly the default mode network—accompanied by increased between-network connectivity (see for instance Avram et al.¹³). These changes are most pronounced in transmodal cortices, i.e., brain regions integrating information from multiple sensory modalities and contributing to higher-level cognitive processes such as executive functions, social cognition, and mentalizing.

Neuroimaging data from some of the above-discussed RCTs are beginning to link such changes to treatment outcomes. For instance, psilocybin therapy has been shown to increase between-network connectivity in patients with depression, an effect correlated with symptom reduction.¹⁴ More recently, Wall and colleagues¹⁵ demonstrated that psychedelics may engage emotional processing circuits differently from conventional antidepressants; while SSRIs such as escitalopram dampen amygdala reactivity to emotional stimuli, psilocybin does not—possibly enabling deeper emotional engagement during therapy. These findings lend support to hypothesized mechanisms of action underlying therapeutic outcomes.

However, despite these promising insights, direct evidence for neuroplastic changes in humans—as assessed through structural or functional neuroimaging—remains limited. Studies explicitly targeting such changes, particularly in clinical populations, are urgently needed. Moreover, to disentangle the contributions of distinct receptor targets, comparative neuroimaging studies across different psychedelics and related compounds are essential. Such studies remain rare¹³ and are entirely lacking in patient populations.

Future research should prioritize longitudinal designs and precision imaging (i.e., collecting large amounts of fMRI data from the same individual) in clinical samples. The integration of biological markers of plasticity—e.g., sleep spindle activity assessed via sleep electroencephalogram (EEG) and polysomnography—may further illuminate the relationship between neural changes and sustained therapeutic outcomes.

In our view, neuroimaging—particularly in patients—remains an underutilized but essential tool in resolving many of the questions posed above, especially when paired with multimodal biomarkers and longitudinal designs.

Subjective experience and contextual modulation

In addition to their neurobiological effects, psychedelics reliably induce altered states of consciousness (ASCs) that may contribute directly to their therapeutic action.¹ ASCs often include ego dissolution, intensified emotional processing, and a re-evaluation of autobiographical memories or entrenched cognitive frameworks (https://alteredstatesdb.org/). Several studies have linked such psychological phenomena—particularly ego dissolution and so-called mystical-type experiences—to improved clinical outcomes (reviewed in Holze et al.¹). However, these associations are typically correlational and complicated by retrospective reporting. It remains unclear whether specific experiential features (e.g., ego dissolution, mystical-type experiences, etc.) are necessary, sufficient, or merely epiphenomenal to therapeutic change.

Indeed, as mentioned above, these psychological phenomena may not be required for therapeutic efficacy. This claim is supported by in vitro and animal studies and, if confirmed in humans, could broaden the clinical applicability of psychedelic compounds to populations currently excluded from psychedelic therapy.

An additional, underexplored dimension concerns the role of set and setting. While widely acknowledged to influence the acute psychedelic experience, their contribution to long-term clinical outcomes remains poorly understood. For example, younger age and state anxiety may increase the likelihood of challenging experiences, but it is unclear whether such experiences are always detrimental or could potentially be therapeutic as well. Notably, most modern RCTs share a highly structured and supportive design: participants are treated in comfortable, non-clinical settings, accompanied by curated music and therapeutic presence. The specific contribution of these contextual psychotherapeutic elements to clinical outcomes is difficult to disentangle.

Ongoing efforts, including studies with embedded neuroimaging protocols (e.g., NCT06626139), are beginning to address these questions systematically. Understanding how context and subjective experience interact with pharmacological action will be key to optimizing psychedelic therapy across diverse clinical populations.

Another crucial aspect of psychedelic administration, though beyond the scope of this discussion, is the psychotherapeutic framework involved. Researchers are actively exploring optimal therapeutic approaches, relevance of the personal psychedelic experience of the therapists (NCT05570708), timing, and duration of integration sessions, all of which are vital for refining treatment protocols and ensuring lasting benefits.

Conclusions

With the recent study by Müller and colleagues, yet another psychedelic compound has demonstrated antidepressant efficacy in a phase 2 double-blind RCT. This adds to the growing body of evidence that supports the therapeutic potential of serotonergic psychedelics in affective disorders. Larger phase 3 trials are now underway and will be crucial not only for confirming efficacy but also for systematically assessing safety profiles and identifying patient populations that are most likely to benefit—or to be at risk.

Throughout this commentary, we have reviewed recent clinical trials that begin to address key open questions: are there specific neurobiological mechanisms that predict response? If so, do these mechanisms differ across various psychedelics? To what extent are the psychological features of the psychedelic experience necessary for therapeutic change? How much do contextual variables, such as set and setting, shape clinical outcomes?

We have emphasized the urgent need for comparative studies, particularly head-to-head trials across different psychedelic compounds and administration strategies. Such studies are notably absent yet would provide invaluable insight into the relative benefits of longer-acting psychedelics such as psilocybin and LSD vs. ultra-short-acting compounds like 5-MeO-DMT or DMT. These fast-acting agents could drastically reduce the cost and logistical complexity of psychedelic therapy, potentially enhancing scalability and accessibility.

At the same time, we argue that combining psychedelic interventions with multimodal neuroimaging (fMRI, PET)—and including EEG and polysomnography—will be instrumental in illuminating the biological substrates of therapeutic change. These tools may also help distinguish general neuroplastic effects from compound-specific or context-dependent mechanisms.

Finally, longitudinal designs will be crucial for both understanding long-term efficacy and identifying delayed adverse effects—especially as these interventions move from controlled research settings into broader clinical use.

In sum, the field is advancing rapidly, and current studies are beginning to untangle the intricate interplay between pharmacology, neurobiology, and experience. With psychedelic therapies on the verge of broader clinical deployment, the field must now shift toward integrative, comparative, and scalable approaches grounded in mechanistic clarity and real-world feasibility.

Declaration of interests

The authors declare no competing interests.