Psychedelics match antidepressants for depression - but don't beat them

For years, psychedelic-assisted therapy has ridden a wave of optimism. Psilocybin trials reported effect sizes that dwarfed those of conventional antidepressants. Investors poured money into startups. Regulators fast-tracked reviews. The narrative was simple and seductive: psychedelics work better than anything else we have for depression.

But a study published March 18 in JAMA Psychiatry offers a sobering counterpoint. When researchers from UC San Francisco, UCLA, and Imperial College London leveled the playing field between psychedelics and traditional antidepressants - eliminating the distortion caused by participants knowing which drug they received - the apparent advantage of psychedelics vanished entirely. Both treatments reduced depression scores by about 12 points on standard clinical scales. The difference between them was negligible.

The blinding problem that haunts psychedelic research

Clinical trials depend on blinding. Ideally, neither the patient nor the clinician knows who got the real drug and who got the placebo. This matters enormously because belief in a treatment can itself produce measurable improvement - the well-documented placebo effect that has complicated drug development for decades.

Traditional antidepressant trials manage this reasonably well. An SSRI pill looks identical to a sugar pill, and side effects are subtle enough that many participants genuinely cannot tell which group they belong to. Blinding holds. The comparison between drug and placebo remains meaningful.

But psychedelics present an obvious problem: when a patient takes psilocybin and experiences vivid hallucinations, dissolved ego boundaries, and a profound sense of cosmic unity, they know they did not receive a placebo. Virtually every participant in a psychedelic trial can tell whether they got the real drug or not. This is not a minor methodological nuisance. It is a fundamental flaw in the experimental design that has never been adequately addressed.

This near-universal unblinding inflates the treatment group's improvement while simultaneously deflating the placebo group's outcomes. Participants who realize they got a placebo may feel disappointed, even hopeless - and their depression scores reflect that disappointment. The result is an artificially wide gap between drug and placebo that makes psychedelics look far more effective than they might actually be.

Comparing apples to apples for the first time

Balazs Szigeti, a clinical data scientist at UCSF's Translational Psychedelic Research Program, wanted to cut through this distortion. He did not expect to undermine the case for psychedelics. He expected to strengthen it.

His team devised a straightforward approach: instead of comparing psychedelic trials to blinded antidepressant trials (the usual benchmark), they compared them to open-label antidepressant trials - studies in which every participant knew they were receiving an active drug. No deception, no guessing.

The logic was clean. In open-label antidepressant trials, participants benefit from the same expectation effects that inflate psychedelic results. Both groups know they are getting a real treatment. Both groups expect to improve. That levels the psychological playing field in a way that no previous comparison had attempted.

The team, which included co-first author Zachary J. Williams of UCLA and Hannah Barnett of Imperial College London, pooled data from psychedelic trials using psilocybin and LSD for major depression, then matched them against open-label trials of traditional antidepressants. They measured depression outcomes using validated clinical rating scales.

Virtually identical outcomes

The results surprised the researchers. Both types of treatment reduced depression scores by approximately 12 points on standard scales. The difference between psychedelics and antidepressants was statistically and clinically negligible.

Szigeti did not set out to debunk psychedelics. Quite the opposite. He went in hoping to demonstrate that even when compared fairly, psychedelic therapy would still outperform conventional antidepressants. The data said otherwise, and he has been candid about his own surprise.

This finding does not mean psychedelic therapy fails. Patients in psychedelic trials improved substantially - a 12-point reduction in depression scores is clinically meaningful by any measure. But so did patients taking standard antidepressants when they knew what they were receiving. The two treatments appear to work about equally well when the confounding variable of unblinding is removed from the equation.

Why psychedelics looked better than they were

The study's analysis helps explain a pattern that has puzzled researchers for years. In psychedelic trials, the gap between the treatment group and the control group is typically large - sometimes dramatically so. In antidepressant trials, that gap is typically small, often barely clearing the threshold for statistical significance. This has been widely interpreted as evidence that psychedelics are simply more potent drugs.

But the researchers argue this interpretation confuses a measurement artifact with a real clinical difference. The large gap in psychedelic trials reflects the impossibility of blinding, not the superiority of the drug. The treatment group benefits psychologically from knowing they received psilocybin or LSD. The control group suffers from knowing they received an inert substance during what was supposed to be a transformative therapeutic experience. In antidepressant trials, where blinding holds reasonably well, neither group gets that extra psychological push or drag.

When you strip away expectation effects by ensuring both groups know they are receiving active treatment, the drugs perform the same. The supposed advantage of psychedelics was, at least in part, an illusion created by broken blinding.

What this means for patients and clinicians

The findings do not suggest that patients should avoid psychedelic therapy. For patients who have not responded to conventional antidepressants - a group that includes roughly one in three people with major depression - psychedelics may still represent a valuable alternative. Not because they work better, but because they work through a different mechanism and may reach people whom SSRIs and SNRIs have failed.

But the results do challenge the narrative that psychedelic therapy represents a dramatic advance over existing treatments for depression. The enthusiasm that has surrounded these drugs - and the billions of dollars invested in psychedelic medicine companies over the past several years - has been built partly on effect sizes that now appear inflated by a methodological artifact rather than genuine pharmacological superiority.

The study also raises practical questions about how regulatory agencies should evaluate psychedelic therapies going forward. If blinding is effectively impossible in psychedelic trials, how should the FDA weigh evidence from studies where every participant knows their treatment assignment? The standard playbook for drug approval was not designed for drugs whose effects are inherently unblindable. So does this work in real-world settings where patients choose their treatment? That question remains open.

Limitations of the analysis

Several important caveats apply. This was a meta-analytic comparison, not a head-to-head randomized trial. The psychedelic data and antidepressant data came from different studies conducted at different times, with different patient populations, different inclusion criteria, and different treatment protocols. A direct comparison trial - randomizing patients to psychedelic therapy or open-label antidepressants within the same study - would provide considerably stronger evidence, but no such trial has yet been conducted.

The analysis focused specifically on major depression. Psychedelics are also being investigated for PTSD, alcohol use disorder, tobacco addiction, and end-of-life anxiety, and the comparative landscape for those conditions may look quite different. The results should not be extrapolated beyond depression without additional data.

The study received no external funding. One co-author, Williams, disclosed consulting fees from Roche. The other authors reported no conflicts of interest, which is notable in a field where financial entanglements between researchers and psychedelic companies have drawn increasing scrutiny.

A field confronting its own expectations

The psychedelic medicine field now faces an uncomfortable reckoning with its own optimism. For years, advocates have pointed to large effect sizes as proof that these drugs outperform conventional treatments. This analysis suggests those effect sizes were measuring something real - but that something was the power of expectation, not pharmacological superiority.

That distinction matters enormously for how research funding is allocated, how trials are designed, and how patients make treatment decisions. Psychedelic therapy may well earn a durable place in the psychiatric toolkit. But its place may be alongside antidepressants as one more option in a limited arsenal, rather than above them as a categorically superior treatment.