Effectiveness and Safety of Psychedelic Drugs in Treating Mental Health Disorders: A Review of Clinical Trial Summaries

This study design followed an umbrella review framework, synthesizing evidence from previously published meta-analyses of randomized controlled trials. This review was conducted according to a pre-prepared, unregistered protocol.

We followed the guidelines according to PRISMA (). In February 2025, the following databases were searched for systematic reviews with meta-analyses of randomized controlled trials (RCTs): PubMed, Scopus, and Web of Science. The search was performed by two independent groups of investigators. The eligible papers, their quality, and selected data were assessed. Figure S1

The search query was as follows: ("psychedelic" OR "mescaline" OR "lysergic acid diethylamide" OR "LSD" OR "psilocybin" OR "ayahuasca" OR "N,N-dimethyltryptamine" OR "dimethyl-tryptamine" OR "DMT" OR "methylenedioxymethamphetamine" OR "MDMA" OR "3-Meo-DMT") AND ("systematic review" OR "meta-analysis"). Searches were not limited by any filters, e.g., study type, language, or publication data. A manual search of the reference lists of included reviews was performed in addition to the digital search to ensure that no relevant articles were missed. Disagreements were resolved by discussion and consensus.

This umbrella review used the PICOS framework to develop detailed inclusion/exclusion criteria, as shown in Table 1. Based on this model and additional information, the following criteria of inclusion were developed: (1) agreeing with PICOS and (2) publications in English. The excluded criteria were as follows: (1) not confirmed with PICOS; (2) not an original article; (3) not in English; (4) full text was not available; and (5) unpublished.

Two reviewers independently screened articles by title, abstract, and then full text. Discrepancies were resolved through the additional consultation of a senior reviewer.

Systematic reviews that included both randomized and non-randomized evidence were eligible only if they contained at least one RCT relevant to the outcome of interest. In such cases, only RCT-derived data were extracted and analyzed, consistent with the standardized umbrella review methodology. Reviews containing exclusively non-randomized studies were excluded.

Data related to the effects of psychedelics on mental disorders were extracted independently by two reviewers using a tailored form. The form included authors, year of publication, country, study design, number of studies included in the meta-analysis, type of psychedelics, diagnosis, and outcomes regarding efficacy, dosage used, and duration of treatment, as well as risks and safety. To ensure consistency in data extraction across reviews, we applied a predefined hierarchy for selecting outcomes from each meta-analysis. Priority was given to (1) the primary clinical efficacy endpoint (e.g., depression severity scales, CAPS for PTSD); (2) remission or response rates, when available; (3) standardized mean differences or effect sizes convertible to Hedges' g; and (4) safety outcomes reported in a comparable format. When multiple eligible measures were provided within a single meta-analysis, we extracted the outcome with the most complete reporting and highest methodological relevance. Adverse events were classified using a unified framework consistent with the clinical trial methodology, distinguishing among (1) any adverse event, (2) treatment-emergent adverse events, and (3) serious adverse events. When inconsistent reporting prevented quantitative pooling, adverse events were summarized using a structured, itemized narrative to improve comparability across studies.

To ensure the quality of this umbrella review, the following assessment tools were used: the Assessment of Multiple Systematic Reviews 2 (AMSTAR2) and GRADE. AMSTAR2 provides a comprehensive assessment of systematic reviews and meta-analyses from analyses from several perspectives, including literature search, statistical analysis, bias, and conflict of interest. It is the most widely used tool for assessing the methodological quality of meta-analyses worldwide and has been used to assess meta-analyses including RCTs. The quality and certainty of evidence were evaluated by two teams of researchers using AMSTAR2 and GRADE (the evidence marked as high, moderate, low, or very low), respectively (Table 2).

In the selected studies, effect sizes were converted to Hedges' g to ensure consistency and comparability across studies. This standardized metric accounts for small sample size bias and allows for the integration of the results derived from various statistical approaches (e.g., standardized mean differences, t-values, or Cohen's d). By transforming all outcomes into Hedges' g, the meta-analytic synthesis maintains methodological rigor and uniformity in effect size interpretation. Quantitative synthesis was conducted only when sufficient and methodologically comparable data were available, allowing effect sizes or summary statistics to be harmonized across meta-analyses. In several areas, particularly for safety outcomes and secondary endpoints, data were incomplete, reported using non-comparable measures, or lacked the statistical information required for effect-size conversion. In these cases, a structured narrative synthesis was applied to avoid producing potentially misleading pooled estimates. To ensure an appropriate level of clinical utility for each mental disorder, the existing evidence was summarized into several best-documented recommendations.

To minimize the risk of double-counting participants from overlapping primary trials, we systematically screened all included meta-analyses for shared RCTs by comparing trial identifiers, study populations, and registration numbers. When the same RCT appeared in more than one meta-analysis, effect sizes were extracted from the most comprehensive or methodologically rigorous review or from the review providing the largest set of complete RCT data. When numerical disentanglement was not feasible, overlapping data were synthesized qualitatively to avoid inflation of precision.

Heterogeneity across meta-analyses was assessed using the I² statistic and the χ² test. When I² exceeded 50% or when clinical or methodological differences suggested meaningful diversity between studies, random-effects models were applied. Fixed-effect models were used only in datasets with low heterogeneity. In cases where pooling was deemed inappropriate due to substantial heterogeneity (e.g., differences in dosing regimens, diagnostic criteria, or population characteristics), the results were synthesized narratively following a structured approach. Analyses were performed using RevMan 5.4 software and STATA 18.

Following a comprehensive search of three databases, a total of 1611 articles were obtained. Following the removal of duplicates, the number of articles was reduced to 877. Following a thorough review of the titles and abstracts, 179 articles were selected for full-text evaluation. The full texts of the 176 potentially relevant articles were then downloaded and evaluated in full. Following this process, a total of 23 meta-analyses that satisfied the specified criteria were selected for inclusion in the systematic review. This process is described in the PRISMA flowchart (Figure 1). Selected meta-analyses have been grouped in Table 2 depending on the psychiatric disorder studied.

As an umbrella review aims to synthesize and evaluate evidence to inform the potential clinical utility of interventions, three key therapeutic areas were identified and examined based on the gathered data: the use of psilocybin in major depression (MD) cases, MDMA in post-traumatic stress disorder (PTSD) cases, and LSD in alcohol use disorder (AUD) cases. Each of these areas was selected due to the growing body of evidence supporting their efficacy and safety profiles, as well as their relevance to the current clinical research and potential future treatment frameworks. The review explores the strength of the evidence, treatment outcomes, and safety considerations for each compound within its respective clinical context.

The umbrella review included 23 meta-analyses that evaluated the clinical utility of psychedelic-assisted therapies across three key indications. A summary of the results of this work is presented in. Overall, the methodological quality of the included reviews was generally low, as reflected by the AMSTAR2 ratings, while the certainty of evidence (GRADE) varied from low to moderate. Tables S1–S4 (Supplementary Materials)

In the area of psilocybin for depression, nine meta-analyses assessed its efficacy, safety, dose–response relationships, and administration (Table S1). All were rated as low quality according to AMSTAR2, except for the studies by Salvetti et al. (2024) [10] and Perez et al. (2023) [8], which provided moderately graded evidence regarding dosing and clinical efficacy. The GRADE certainty of evidence ranged from low to moderate, with moderate ratings typically assigned where pooled analyses and consistent outcomes supported the antidepressant effect of psilocybin, especially in the short-term follow-up.

For MDMA-assisted therapy in PTSD cases, ten meta-analyses were included (Table S2). While most were rated as low or critically low on the AMSTAR2 scale, a few demonstrated higher methodological rigor (Table 2). These reviews also yielded moderate certainty of evidence (GRADE), particularly in relation to symptom reduction, response and remission rates, and short- to mid-term safety profiles. However, many reviews either lacked pooled safety analyses or had limitations in assessing long-term adverse effects. Notably, Colcott et al. (2024) provided the most comprehensive and methodologically sound assessment of adverse events across both Phase 2 and Phase 3 trials [24].

In the case of LSD for alcohol use disorder, two reviews [6,26] supported the short-term efficacy of LSD in increasing abstinence and reducing alcohol consumption (Table S3). Both were rated as low quality (AMSTAR2) and had low certainty of evidence (GRADE), largely due to methodological concerns. Despite this, both reviews indicated a consistent short-term therapeutic effect.

Four meta-analyses provided sufficient data to be meta-analyzed. Data are depicted in Figure 2. Pooled findings confirm that psilocybin and psychedelic-assisted therapy significantly reduce depressive symptoms—Hedges' g =1.05, CI: [0.84; 1.26], p < 0.001.

Most psilocybin studies for depression utilized moderate to high doses, typically in the range of 10–30 mg. When normalized to a 70 kg adult, these doses range from approximately 0.14 to 0.5 mg/kg. This reflects common clinical protocols. Only one study used strictly body-weight-adjusted dosing. Moderate doses are most common, balancing efficacy and tolerability. Table 3 summarizes the psilocybin doses in major depression.

According to Swieczkowski et al. (2025), doses of 0.215 mg/kg and 25 mg were particularly effective, with 25 mg outperforming 10 mg (RR—risk ratio = 5.62) [7]. Similarly, Perez et al. (2023) reported a significant dose–response relationship (p < 0.0001), with an estimated ED50 of 8.23 mg/70 kg and ED95 of 24.68 mg/70 kg [8]. Li et al. (2022) [3] found that high doses (30–35 mg/70 kg) showed greater efficacy (Hedges' g = 3.059). The short-term subgroup (≤1 month) displayed stronger effects (Hedges' g = 1.534) compared to long-term treatment (Hedges' g = 1.123) [3]. Ko et al. (2023) [11] analyzed various psychedelics and found a significant reduction in depressive symptoms from Day 1 (SMD—standardized mean difference = −1.36) to Week 5 (SMD = −3.12), with the strongest effect observed in Weeks 3–5. However, the effect was no longer significant by Weeks 6–8 (SMD = −1.52; p = 0.14), suggesting a diminishing impact over time [11]. Menon et al. (2024) demonstrated superiority over control conditions at all evaluated timepoints: Day 2 (SMD = −1.11), Day 7 (SMD = −0.72), Day 14 (SMD = −0.67), and Day 42 (SMD = −0.56) [13].

Follow-up durations in the studies ranged from 2 weeks to 12 months, mostly covering short to medium terms. These findings highlight the nuanced efficacy of psilocybin in varying contexts and dosages, particularly its pronounced short-term benefits in alleviating depressive symptoms. However, the durability of these effects over extended periods remains uncertain. Furthermore, comprehensive research is needed to delineate its optimal therapeutic window and to address variability in individual responses. These considerations are critical as psilocybin therapy continues to evolve as a potential standard of care for major depressive disorder.

Psilocybin was associated with a higher risk of adverse events, though no significant increase in serious adverse events was observed. Common short-term adverse events included headache, anxiety, nausea, and confusion [8,9,13]. These findings highlight psilocybin's significant short-term antidepressant effects and manageable side effects, despite needing further study for long-term safety. Data variability prevented meta-analysis.

MDMA-assisted psychotherapy has been evaluated in numerous studies and meta-analyses, consistently demonstrating strong therapeutic effects in the treatment of PTSD. Four meta-analyses provided sufficient data to conduct data synthesis. The results are depicted in Figure 3.

MDMA significantly reduced PTSD symptoms compared to the placebo or active placebo, with a large effect size (Hedges' g = 1.24 95% CI: 1.08, 1.41, p < 0.001).

PTSD studies using MDMA-assisted therapy report moderate to high doses, typically in the range of 75–125 mg. While some studies use body-weight-based dosing, most apply fixed-dose regimens. For a 70 kg adult, this corresponds to approximately 1.0–1.8 mg/kg. The consistency of dosing across studies reflects established therapeutic protocols. There is a limited exploration of dose–response curves in meta-analyses. Table 4 summarizes the MDMA doses in PTSD cases.

Yang et al. (2024) focused on clinical efficacy and dosing and found that high-dose MDMA (125 mg) was more effective than lower doses (e.g., 75 mg) [2]. On the other hand, Illingworth et al. (2021) [19] reported significant reductions in PTSD symptoms at both 75 mg (MD = −46.90; p < 0.00001) and 125 mg (MD = −20.98; p = 0.002) when compared to the active placebo. Analyzed studies clearly reported using it for 2 to 3 sessions, in line with the therapeutic protocols. Regarding the follow-up data—one long-term follow-up study (17–74 months) reported lasting relief in most participants, with only two cases of relapse [19]. Tedesco et al. (2021) supported these findings, showing that improvement was sustained up to 32 months (SMD = 0.81) [23].

Colcott et al. (2024) reported that MDMA-assisted therapy was associated with an increased likelihood of experiencing any adverse event during both the medication session (OR—odds ratio = 1.67) and the following week (OR = 1.59), with an even greater risk in Phase 3 trials (OR = 3.51) [24]. Adverse events were more frequent with higher doses and included jaw clenching, muscle tightness, nausea, appetite loss, chills, jitteriness, and blurred vision. Most reported adverse events were mild and resolved by 6 months; one serious event (suicidal ideation) was noted as potentially drug related. Importantly, there were no significant differences between groups in the incidence of treatment-emergent adverse events (RR = 1.03) or serious adverse events (RR = 0.95) [18]. Data variability prevented meta-analysis.

Sicignano et al. (2024) and Krebs et al. (2012) both provide evidence that LSD-assisted therapy can significantly improve outcomes related to alcohol misuse and abstinence, particularly in the short term [6,26]. Sicignano et al. (2024) found that, in randomized, double-blind, placebo-controlled trials, both LSD (OR = 1.99; 95% CI: 1.10–3.61) and psychedelics generally (OR = 2.16; 95% CI: 1.26–3.69) significantly increased the odds of achieving abstinence or reducing alcohol use [6]. Krebs et al. (2012) reported similar findings, with LSD rapidly improving alcohol misuse with improvement remaining significant up to 6 months in the follow-up (OR = 1.66) [26]. Overall, LSD-assisted therapy appears to offer meaningful short- and mid-term benefits for alcohol misuse, though long-term efficacy and safety require further study. Data were insufficient to perform a meta-analysis.

We identified another area of mental disorders that may benefit from psychedelic therapy: neurodevelopmental conditions such as autism. MDMA has shown potential therapeutic effects, including improved social functioning, enhanced social bonds, and feelings of love and friendship [5,27]. These results are presented in Table S4 (Supplementary Materials). However, the data are currently insufficient to draw definitive conclusions or to conduct a meta-analysis.

Among the available research results on psychedelic-assisted therapy in major de-pression cases, psilocybin comes to the fore. This umbrella review with meta-analyzed data has shown that psilocybin is effective in MD cases (Hedges' g = 1.05, 95% CI: 0.84; 1.26, I. = 88%, p < 0.001). Importantly, the therapeutic effect can last up to 6 months [35], after just one dose [16]. Psilocybin induces subjective experiences with mystical characteristics, such as an increased sense of oneness with the universe, transcendence of time and space, loss of self, and euphoria. Importantly and significantly, numerous observations confirm that mystical experiences are correlated with long-term improvements in well-being, psychosocial functions, and reductions in anxiety and depressive symptoms. There have been many inconsistencies around psilocybin dosage. Some researchers point to the therapeutic potential of even the lowest doses [36]. However, other meta-analyses indicated a clear relationship between therapeutic effect and dose. This controversy was compounded by the recreational use of psychedelic substances in various doses, as well as their use in the so-called "underground". Available data were too heterogenous to be meta analyzed, however the current evidence indicates that the therapeutic effect is dose-dependent, i.e., the higher the dose, the better the therapeutic outcomes. Effective doses for major depression are in the range of 25–36/70 kg, with higher doses required for treatment-resistant depression [3,8]. The optimal therapeutic dose, beyond which no further therapeutic gain is observed, depends on the specific patient population and potential confounding factors, like age and previous psychedelic experience. In general, the more severe and drug-resistant depression, the higher the doses required. Yet, another issue remains the administration schedule. According to Perez et al. (2023) [8], no significant difference between single- and two-dose psilocybin administration was observed. Moreover, the therapeutic effects of psilocybin in treating depression can be seen within a day, unlike antidepressants [8]. Only ketamine works so quickly [37]. However, psilocybin, unlike ketamine, has at the same time low potential for addiction and physical dependence [38]. Furthermore, the efficacy of psilocybin as compared to esketamine seemed to be similar with a better safety profile [39].

The Canadian Network for Mood and Anxiety Treatments (CANMAT) Task Force in a 2023 report summarized the results on psychedelic-assisted therapy for major depression [40]. CANMAT suggests psilocybin may effectively treat major depression, however the strength of the evidence has been described by this organization as weak (evidence level C). Thus, psilocybin remains an experimental option, available only through clinical trials or special programs. Similar regulations have been introduced in Australia.

In addition to the challenge of registering a substance like psilocybin for medicinal use, there is also the issue of finding a legal manufacturer. Psilocybin, which is present in certain mushrooms, poses challenges for stable synthesis and large-scale production.

The recommended PTSD treatment includes cognitive behavioral therapy (CBT) and selective serotonin reuptake inhibitors (SSRIs) like sertraline or paroxetine. CBT with exposure therapy is effective for certain patients but demands significant time and therapist expertise. One in four patients quits this therapy mid-treatment. Pharmacotherapy has limitations, with only 40–60% of patients showing an improvement. There is significant interest in psychedelic-assisted psychotherapy, particularly with ketamine and MDMA for PTSD treatment. MDMA-assisted therapy, a hallucinogen and psychostimulant, has proven more effective at reducing PTSD symptoms compared to placebo, as measured by the clinician-administered PTSD Scale (CAPS). In this umbrella review, we found that MDMA significantly reduced PTSD symptoms compared to the placebo or active placebo, with a large effect size (Hedges' g = 1.24 95% CI: 1.08, 1.41, p < 0.001). The difference in the effectiveness of the therapy—the size of the effect—was very large—nearly 2 times more effective than the standard of treatment used so far. The improvement lasted between 2 to 32 months. Moreover, this therapy has been deemed safe and well-tolerated by patients [23]. Apart from temporary side effects like bruxism, jaw clenching, anxiety, headache, nausea, and a brief loss of appetite, no serious or permanent effects have been reported. The effective dosage ranges from 40 mg to 187.5 mg. In general, the number of sessions and follow-up length are more consistently reported in MDMA studies compared to psilocybin studies. This may reflect the more advanced regulatory status of MDMA-assisted therapy in PTSD cases. This highlights a greater emphasis on long-term outcome monitoring in MDMA trials.

The U.S. FDA's refusal of registration, as of 13 August 2024, raised eyebrows. The decision pertains to methodological issues in studies on psychedelic-assisted psychotherapy, suggesting these should be addressed differently than the current standards. One contentious issue was the lack of standardized procedures for MDMA therapy, with varying doses and regimens used across different centers. Functional unblinding was another problem, as most participants guessed correctly whether they received MDMA or a placebo. The FDA proposed using active placebos. They also expressed concerns regarding inadequate monitoring of side effects, insufficient laboratory check-ups, and unclear long-term addiction risks. Therefore, it is essential to establish a standardized protocol for MDMA-assisted psychotherapy that includes optimal dosing and safety measures.

The research suggests potential benefits of psychedelic therapy for treating alcohol abuse and neurodevelopmental disorders. Studies on using psychedelics to address alcohol misuse are in the early stages [6,26]. LSD-assisted therapy has shown statistically significant short-term improvements in alcohol misuse and abstinence outcomes, but efficacy appears to decline over time, with no significant benefits detected at the 12-month mark. Although the initial results are promising, the current evidence is insufficient to recommend this therapy. Similarly, the research on neurodevelopmental disorders [5,27] indicates further trials with standardized session reporting and addressing adverse events are needed to validate long-term effects.

Current evidence suggests that psychedelic treatment is generally well-tolerated [12,24]. Commonly reported adverse events in psychedelic-assisted therapy include nausea, headaches, and anxiety for various indications and substances used. No studies have reported any serious side effects. There was a temporary increase in heart rate and systolic and diastolic blood pressure. Longer-lasting side effects included headaches (following psilocybin and MDMA), fatigue, depressed mood, and anxiety (following MDMA). The most frequent side effect of psychedelics was headache. However, it remains essential to monitor the safety of therapy according to the established protocol, which includes performing blood pressure measurements, ECGs, and basic laboratory tests, often omitted in previous studies. Reporting quality of adverse events varies between studies and meta-analyses. Most adverse events data are not pooled quantitatively, and many reviews report narrative summaries only.

However, in this umbrella review, we observed that reporting of safety outcomes was fragmented across trials and meta-analyses. Definitions of adverse events varied, with some studies relying on spontaneous reports and others using structured assessments. Serious adverse events were infrequently reported and often insufficiently detailed to allow quantitative synthesis.

Beyond the randomized controlled trials and meta-analyses included in this umbrella review, a broader body of clinical and observational literature suggests that psychedelic-assisted therapies are not universally beneficial and that a subset of patients may experience deterioration or clinically significant adverse psychological effects [41,42]. Reported phenomena include exacerbation of anxiety or depressive symptoms, emergence or worsening of suicidal ideation, and, more rarely, psychotic or dissociative symptoms [41,42]. Although such events appear to be infrequent, they highlight the importance of conservative patient selection, careful screening, preparation and integration, and long-term monitoring of outcomes.

The regulatory process for a substance's medicinal use, when it is on the illegal list, is complex. Predicting all scenarios and consequences of the final decision poses significant medical, legal, and social challenges. We may need to rethink the current paradigm as it no longer aligns with today's regulatory system and health needs. Nevertheless, the primary focus should always be on ensuring patients have access to effective modern therapies.

This umbrella review has several important limitations. At the level of the primary literature, the evidence is characterized by substantial heterogeneity in study design, dosing protocols, treatment settings, outcome definitions, and adverse event reporting, as well as an uneven maturity of evidence across diagnostic categories. Most trials are small, include highly selected populations, and provide limited long-term follow-up, which restricts the generalizability and durability of observed effects. Limitations specific to this umbrella review include the absence of formal protocol preregistration, limited pooling restricted to methodologically comparable data, and the risk of overlapping primary trials across meta-analyses. Although steps were taken to mitigate these issues, they reduce the independence and precision of pooled estimates. Consequently, conclusions regarding efficacy and safety should be interpreted with caution.