We have a medical treatment that’s been shown to massively help people with depression, yet the actual dosing of the compound seems to be only a part of the beneficial effects. When it comes to psilocybin-assisted therapy (PAT), preparation and integrations are proving to be just as important.
The main reason is because this field is so new, so compartmentalised, and so starved of funding for so long that a dozen institutions have been independently inventing the wheel. And the result is a pile of promising data with little standardisation for best practices.
That’s exactly what a new peer-reviewed meta-analysis published in Neuroscience & Biobehavioral Reviews set out to address. It looks beyond “does psilocybin work for depression?” (spoiler: yes) and instead focuses on which variables produce the best therapeutic outcomes.
It Works. Now Let’s Talk About How.
The study found a large, statistically significant antidepressant effect for psilocybin-assisted therapy across all seven trials. In meta-analysis speak, that’s an SMD of -0.79 – which puts psilocybin-assisted therapy firmly in the territory of “clinically meaningful.” This is consistent with earlier reviews from Goldberg and colleagues (2020) and Haikazian and colleagues (2023). This headline isn’t new.
But what is new, and what makes this paper interesting, is the subgroup analysis. The authors broke down the 7 trials by their methodological choices and looked at whether those choices correlated with better or worse outcomes. And the differences they found were striking.
“The antidepressant efficacy is optimised when administering bodyweight-adjusted doses, offering 8–10 hours of preparation, 8–10 hours of dosing sessions, 4 or more hours of integration — and non-manualized psychotherapy.”
The Four Things That Actually Seem to Matter
Here’s what the data pointed to. Please note that the authors are careful to flag that these are correlational findings from a small pool of studies, not causal laws. But the signals are clear enough to be worth taking seriously.
01 — Bodyweight-adjusted dosing was associated with an effect size of -1.36 vs -0.57 for fixed doses. So not giving everyone the same 25mg pill (i.e treating humans like the individuals they are) appears to make a real difference.
02 — Longer preparation sessions (8–10 hours vs 2–4 hours) were linked to larger effects. The set and setting conversation isn’t just hippie wisdom. It’s building therapeutic alliance, managing expectation, and preparing the nervous system.
03 — Longer dosing and integration sessions also correlated with better outcomes. The experience doesn’t end when the psilocybin wears off. What you do with it afterwards (the meaning-making) may be as important as the trip itself.
04 — Non-manualized psychotherapy outperformed scripted, manualized approaches. This one is counterintuitive, and the authors themselves flag it may partly be a reporting artefact. But it raises a question about whether rigid protocols suppress something essential in the therapeutic dynamic.
Why the “Non-Manualized” Finding Is the Interesting One
Let’s look the manualized vs non-manualized result for a moment, because it’s the most nuanced of the lot.
Studies using a formal manual to guide the psychological support showed a smaller effect size (SMD -0.57) than those without one (-1.36). On the surface, that looks like a win for intuition over protocol. But the authors offer an important caveat: several studies that claimed to use manuals were vague about what that actually meant in practice. And studies labelled “non-manualized” may simply have been under-reporting their methodology.
There’s also the facilitator variable. The quality of the guide (their training, their presence, and their experience) is almost certainly doing the heavy lifting here. And yet most studies gave this almost no page space. Some listed facilitators with bachelor’s degrees. Others had PhD psychiatrists. Davis et al. (2021) described facilitators with “varying levels of education” spanning social work to medicine.
What the finding probably tells us isn’t “throw out the manual.” It’s more likely implying that authentic therapeutic presence matters more than rigid adherence to script. Which, if you’ve ever been on either side of a meaningful conversation about difficult emotions, will surprise exactly no one.
The Regulatory Problem
In 2024, the FDA rejected MDMA-assisted therapy partly because of concerns about inconsistent therapeutic protocols and unblinding bias across trials. Sound familiar?
Psilocybin is not far behind MDMA in the regulatory queue. And if this field wants a different outcome, it needs to do something MDMA-assisted therapy couldn’t: present a coherent, standardised body of evidence that regulatory bodies can actually evaluate. Right now, you have 22-site multinational trials sitting next to single-site studies with 19 participants. You have 2-hour prep sessions next to 10-hour ones. It’s a fascinating mess.
This paper is, in part, a polite but clear message to the field to sort it out. Define best practice. Build consensus. We know it works, now we just need to find out how to make it work reliably and safely at scale.
“The evidence is in. Psilocybin helps with depression. The question is no longer whether — it’s whether we’re doing it right.”
The Sample Problem
One more finding worth flagging is that, across all 7 studies in this meta-analysis, an average of 91% of participants were White. In some studies, that figure reached 94%.
For a treatment being positioned as a potential response to a global mental health crisis, that’s a big blind spot. Psilocybin’s therapeutic effects almost certainly interact with cultural context, lived experience, and social environment in ways we don’t yet understand.
The Takeaway
If you’re someone following psychedelic therapy with a clinical interest – perhaps as a potential patient, a mental health professional, or just someone who cares about where this is going – here’s the takeaway in plain language:
Psilocybin-assisted therapy is real. The antidepressant effects are substantial and reproducible across multiple randomised controlled trials. But the container matters. How long you spend preparing. How much time you have to process the experience afterwards. Whether the person sitting with you is a warm, skilled human presence or someone reading from a script. These things seem to change outcomes significantly.
We’re not at the “book your psilocybin therapy appointment at your GP” stage. We’re still in the stage of figuring out what the best version of that appointment looks like. This paper is a meaningful step in that direction — and one of the more honest, rigorous attempts to answer that question so far.
The research is moving fast. Keep up.
The Spore Report covers psychedelic science, policy, and culture — clearly written, no hype. If this kind of deep-dive is your thing, it’s worth your inbox.
