Around 30% of people don’t respond to traditional antidepressants (TADs), such as SSRIs or SNRIs (McIntyre et al., 2023) and there is a pressing need for new treatments.
In recent years, psychedelic-assisted therapy (PAT) has generated a lot of excitement as a potential new treatment for MDD. PAT uses psychedelics, such as psilocybin (found in magic mushrooms) or LSD along with psychotherapy and results from early trials have been promising. Specifically, PAT seems to produce larger improvements in depression than those seen in trials of TADs (von Rotz et al., 2022). What’s more, PAT also seems to be effective in people with difficult-to-treat depression i.e., those who have had two courses of separate TADs but experienced little-to-no improvement in symptoms (Goodwin et al., 2022).
But there’s a snag: in placebo-controlled clinical trials of PAT, participants can guess whether they have received a placebo or psychedelic around 95% of the time – they are ‘functionally unblinded’ to the treatment they receive (Holze et al., 2023). Put simply, it’s hard to not realise that you’re tripping! This makes it difficult to separate out the placebo and true drug effects.
The authors of this review argue that the effectiveness of PAT and TADs can only be truly compared if like is compared with like, specifically comparing the results of PAT trials with those of open-label (unblinded) TAD trials. This is what they set out to do.
Psychedelic-assisted therapy is being explored as a treatment for depression, but trial unblinding makes it difficult to separate drug effects from expectations.
Methods
The authors aimed to test three hypotheses:
- H1: PAT will outperform open‑label TAD treatment.
- H2: Open-label TAD will outperform blinded TAD treatment.
- H3: Blinded and open-label PAT treatment will perform similarly.
To decide if one treatment was better than another, they selected a threshold representing a clinically noticeable difference in depressive symptoms. This corresponded to a difference in score of 3 on the HAM-D, a commonly used scale for measuring depression severity.
To explore these hypotheses, the team ran a systematic search using a single study database (PubMed) to identify all outpatient trials in adults with MDD that used either PAT or open‑label TADs. They then pooled results across studies using meta-analytic techniques to estimate the overall effect of each treatment on depressive symptoms.
For the comparison in H2, they drew on data from a previously published meta‑analysis of blinded TAD studies (Cipriani et al., 2018).
Results
The search identified 24 studies suitable for meta-analysis. This included 16 open-label TAD trials and 8 PAT trials (6 of which were formally blinded).
The authors used two statistical frameworks to test their hypotheses: Bayesian and frequentist. These are two different ways of looking at probability and assessing how certain we can be of a result. Here, the authors were trying to identify the ‘true value’ of the difference in effectiveness between PAT and TADs.
The Bayesian framework says: given the results we’ve seen, there’s a 95% chance the true difference falls within a certain range (the credible interval, CrI). The frequentist framework says: if I were to repeat this experiment 100 times, I can identify an interval (the confidence interval, CI) that would capture the true difference in 95 of those 100 repetitions.
H1 Results
Both Bayesian and frequentist frameworks failed to find support for H1; the evidence was in keeping with there not being any difference between PAT and open-label TADs. This was the same when only PAT studies of non-treatment-resistant depression were considered, and when PAT was compared against individual classes of TAD (SSRI and SNRI).
H2 Results
For H2, both frameworks suggested open-label TAD treatment was more effective than blinded treatment. However, the size of this effect was too small to be clinically meaningful.
H3 Results
Finally, both frameworks supported H3: i.e., there was no important difference in the effects of blinded and open-label PAT treatment.
This meta-analysis found no clear evidence that psychedelic-assisted therapy works better than antidepressants, with only small or no meaningful differences between treatments.
Conclusions
By pooling the results from multiple studies, the authors concluded that there was no clinically important difference in the effectiveness of PAT and open-label TAD treatment. The authors argue that this allows for a fairer comparison of the two treatment types, due to the functional unblinding that is seen with PAT.
Open-label TADs marginally outperformed blinded TADs, but not in a way that was clinically meaningful. Conversely, open-label PAT and blinded PAT performed similarly, supporting the idea that blinding doesn’t work in PAT trials.
The authors concluded that there was no clinically important difference in the effectiveness of PAT and open-label TAD treatment.
Strengths and limitations
The authors’ decision to use both frequentist and Bayesian methods to test their hypotheses is a strength as it allows them to ‘stress-test’ their findings. If the results of the two approaches agree, as they broadly did for all the hypotheses, that suggests the conclusions are more robust.
Several methodological decisions, however, deserve a closer look.
First, the timing of the final outcome in the PAT trials was earlier than in TAD trials (3.4 vs 8.1 weeks, respectively). The authors state this probably favoured PAT trials, although they do not explain why. They deliberately excluded TAD studies that measured outcomes earlier than 6 weeks, the reason given for this being that:
it is known that TAD’s effect take at least 6 weeks to start.
No such restriction was applied to PAT studies. In fact, the pure drug effect (as opposed to placebo effect) can be seen as early as 1 week in treatment with antidepressants – it is at its steepest point early on and gradually flattens over time (Taylor et al. 2006, Cheng et al. 2020). At 6 weeks, antidepressants are therefore approaching their maximal effect; using only studies with an end point after this could risk inflating the comparative effect of TAD in relation to PAT.
Second, the authors excluded antidepressant trials with fewer than 100 participants. They justify this by saying small trials contribute little to the pooled estimate while adding workload. However, they don’t report how many of these studies were removed for this reason. If it was a lot, that could potentially shift the results.
Third, the search strategy was unusually narrow. They only searched PubMed, even though standard meta‑analytic practice is to search multiple databases to avoid missing eligible studies. They also didn’t check for publication bias which is where studies with positive findings are more likely to be selected to be published. Both these decisions increase the risk that relevant studies were overlooked.
Finally, while the authors wish was to compare like with like, it is important to remember that PAT is a unique treatment paradigm with combined pharmacological and psychotherapy components. It may be interesting, therefore, for a future meta-analysis to compare PAT with studies of combined TAD and psychotherapy.
Methodological choices such as timing, study selection, and limited search strategy may have influenced results despite broadly consistent statistical findings.
Implications for practice
The large effect sizes reported in early psychedelic therapy trials have contributed to considerable hype, with some framing it as a ‘miracle treatment’ (Yaden et al., 2022). The results of this study do not support such an optimistic picture and instead place the efficacy of PAT at the same level of our current, imperfect antidepressants. Its findings are corroborated by another meta-analytic study that, using different methods, also suggested an equivalence between PAT and TADs (Hsu et al., 2024).
Such findings could pose an issue for the widespread clinical adoption of PAT. After all, PAT is expensive and time-consuming: the dosing sessions alone typically require two highly trained clinicians to sit with a single patient for around eight hours. If PAT is only as good as the antidepressants that we already have, why bother?
Nevertheless, PAT may provide other benefits that mean it is still in contention:
Firstly, as the authors of this paper note, monitoring changes in depressive symptoms may only capture a proportion of the benefits of PAT. A 6-month follow up of a study that compared escitalopram (a commonly prescribed antidepressant) and psilocybin head-to-head did not find any significant difference in the improvement in depressive symptoms between the treatments. However, psilocybin performed better on scores of functioning, psychological connectedness, and meaning in life, suggesting a broader recovery that may hold up better against relapse (Erritzoe et al., 2024).
Secondly, the original escitalopram/psilocybin study also found that the psychedelic was associated with more acute, transient side-effects, predominantly occurring on dosing days. Escitalopram, on the other hand, was associated with more chronic, day-to-day side effects (which makes sense given that escitalopram, but not psilocybin, is taken every day), and more sexual side-effects (which are often a big factor in antidepressant discontinuation) (Cahart-Harris et al., 2021).
Thirdly, some of the studies included in the meta-analysis demonstrated efficacy of PAT in difficult-to-treat depression, whereas the TAD studies were overwhelmingly focused on non-difficult-to-treat depression (which makes sense, because treatment success is defined according to the effect of TADs on a patient’s symptoms). Additionally, when the difficult-to-treat PAT studies were removed from the meta-analysis, this did not seem to affect the efficacy of PAT versus TAD. In other words, the results from the difficult-to-treat studies weren’t ‘dragging down’ the pooled efficacy, suggesting that they work similarly for difficult-to-treat and non-difficult-to-treat depression.
Therefore, one could argue that, while not replacing TADs, PAT may still be a viable alternative treatment for some patients; particularly those who might benefit from more ‘holistic’ changes to their behaviours and thought patterns, those who cannot tolerate TADs because of side effects, and those with difficult-to-treat depression.
From my own experience working on clinical trials of PAT, I’ve seen patients whose depression doesn’t improve after receiving what they may have implicitly hoped would be a “miracle treatment”. For some, this can deepen feelings of self‑blame (“what’s wrong with me that means it didn’t work?”) or hopelessness (“if even this didn’t help, what will?”). Findings from studies like this one may help to calibrate expectations around PAT, reduce the risk of that kind of disappointment, and ultimately helping patients.
Despite early hype, results indicate psychedelic therapy is similar in effectiveness to antidepressants, raising questions about cost and widespread clinical use, but it may still be relevant in specific patient groups and treatment settings.
Statement of interests
Luke Baxter is a study medic and researcher on clinical trials of psychedelic therapy for psychiatric conditions, and receives payment for that work. AI was used to help polish the text of this article to improve clarity.
Editor
Edited by Éimear Foley. AI tools assisted with language refinement and formatting during the editorial phase.
Links
Primary paper
Williams, Zachary J., Barnett, Hannah, & Szigeti, Balázs.Psychedelic Therapy vs Antidepressants for the Treatment of Depression Under Equal Unblinding Conditions: A Systematic Review and Meta-Analysis. JAMA Psychiatry. Published online March 18, 2026. https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2846479
Other references
World Health Organisation (2025) World mental health today: latest data. https://iris.who.int/bitstream/handle/10665/382343/9789240113817-eng.pdf
Carhart-Harris R, Giribaldi B et al. (2021). Trial of psilocybin versus escitalopram for depression. New England Journal of Medicine, 384(15), 1402-1411. https://www.nejm.org/doi/full/10.1056/NEJMoa2032994
Cheng Q, Huang J et al. (2020) Analysis of Time-Course, Dose-Effect, and Influencing Factors of Antidepressants in the Treatment of Acute Adult Patients with Major Depression, International Journal of Neuropsychopharmacology, Volume 23, Issue 2, Pages 76–87. https://academic.oup.com/ijnp/article/23/2/76/5644522
Cipriani A, Furukawa T et al. (2018) Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis, The Lancet; 391, 1357-1366. https://www.thelancet.com/article/S0140-6736(17)32802-7/fulltext
Erritzoe D, Barba T et al. (2024). Effect of psilocybin versus escitalopram on depression symptom severity in patients with moderate-to-severe major depressive disorder: observational 6-month follow-up of a phase 2, double-blind, randomised, controlled trial. EClinicalMedicine, 76. https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(24)00378-X/fulltext
Goodwin G, Aaronson S et al. (2022) Single-Dose Psilocybin for a Treatment-Resistant Episode of Major Depression. N Engl J Med. 2022 Nov 3;387(18):1637-1648. https://www.nejm.org/doi/10.1056/NEJMoa2206443
Holze F, Gasser P et al. (2023). Lysergic acid diethylamide–assisted therapy in patients with anxiety with and without a life-threatening illness: A randomized, double-blind, placebo-controlled phase II study. Biological psychiatry, 93(3), 215-223. https://www.sciencedirect.com/science/article/pii/S0006322322015530?via%3Dihub
Hsu, T.-W. et al. (2024) Comparative oral monotherapy of psilocybin, lysergic acid diethylamide, 3,4-methylenedioxymethamphetamine, ayahuasca, and escitalopram for depressive symptoms: systematic review and Bayesian network meta-analysis. BMJ 386, e078607. https://www.bmj.com/content/386/bmj-2023-078607.long
McIntyre, R et al. (2023) Treatment-resistant depression: definition, prevalence, detection, management, and investigational interventions. World Psychiatry 22, 394–412. https://onlinelibrary.wiley.com/doi/10.1002/wps.21120
Taylor M, Freemantle N et al. (2006) Early Onset of Selective Serotonin Reuptake Inhibitor Antidepressant Action: Systematic Review and Meta-analysis. Arch Gen Psychiatry. 63(11):1217–1223. https://jamanetwork.com/journals/jamapsychiatry/fullarticle/668229
von Rotz R, Schindowski E et al. (2022) Single-dose psilocybin-assisted therapy in major depressive disorder: A placebo-controlled, double-blind, randomised clinical trial. EClinicalMedicine. 28;56:101809. https://www.sciencedirect.com/science/article/pii/S2589537022005387?via%3Dihub
Yaden D, Potash J, Griffiths R. (2022) Preparing for the Bursting of the Psychedelic Hype Bubble. JAMA Psychiatry. 79(10):943–944. https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2795948
