How does MDMA actually work? Our new study recruits psychotherapists to help answer that question
- Rosalind McAlpine, Elisa Liberati, and Sunjeev Kamboj
- 12 hours ago
- 7 min read
Written by Rosalind McAlpine, Elisa Liberati, and Prof Sunjeev K Kamboj
This blog was written by the co-researchers about their upcoming study, which forms part of Drug Science and University College London’s strategic collaboration to advance our understanding of MDMA-assisted psychotherapy.
Of all the psychedelic treatments now under investigation, MDMA-assisted psychotherapy is among the most advanced. Two Phase III trials have supported its efficacy and safety for post-traumatic stress disorder (PTSD), and by the usual standards it met or exceeded the bar for a treatment to be considered empirically supported. Even so, in 2024 the US Food and Drug Administration (FDA) declined to approve it, citing a range of limitations.
Underneath that decision sits a question the field has not properly answered: even when MDMA-assisted therapy works, we cannot say clearly how it works, or what part the psychotherapy plays in it. That gap is the reason for the study we have just begun at UCL, an unusual group of participants: experienced psychoanalytic psychotherapists. The protocol is now published in PLOS One.
The unresolved role of the therapy
MDMA-assisted therapy is a combination treatment: MDMA is given within a course of psychotherapy. That structure creates a missing evidence piece that clinical trials have not yet been able to address. They compare MDMA plus psychotherapy against a placebo drug plus psychotherapy, which means the drug’s contribution and the therapy’s contribution cannot be pulled apart during the analysis. And because the FDA regulates drugs but not the practice of psychotherapy, a central component of the treatment is excluded from consideration by the system assessing it.
This has left a real question open: how much does the psychotherapy matter in MDMA-assisted therapy, and what kind of psychotherapy should be used? Some researchers have gone as far as to question whether co-administered psychotherapy is necessary at all. Others have argued for a particular model; for example, Yaden and colleagues have proposed that cognitive behavioural approaches should be the default framework for all psychedelic treatments. Overall, there is a general scepticism toward the therapy used in the main MDMA trials (e.g., a non-directive, “inner-directed” approach drawing on somatic and internal family systems methods) that seems to stem from its exclusion in mainstream psychological research.
The difficulty with settling any of this is that the arguments are being made without an agreed account of what MDMA actually does to a person psychologically. It is hard to say which elements of the therapy are essential, or which form of psychological therapy best complements the drug, when the drug's own psychological mechanism has not been specified.
We can describe MDMA’s effects. We can't yet explain them
There is no shortage of description. The literature tells us that MDMA increases things like self-compassion, openness and empathy, and dampens the response to fear. These are real and important observations. But they are essentially a list of components – the raw materials from which a mechanistic model might be built, not the model itself.
What is missing is everything that would turn that list of components into an explanation: the order in which these changes occur, how one leads to another, and how they are embedded in the relationship between a person and their therapist. Current accounts, as noted in the recent protocol, remain at the level of descriptive behavioural psychopharmacology. The order in which various changes occur, the reason why they occur, and the context in which they occur are underspecified, and the field lacks both a shared vocabulary and validated measures for these processes. That is why proposals to install one school of therapy as the default are premature: they tend to reason at the level of separable ingredients (e.g., openness, engagement, awareness) rather than explaining, as a process, how MDMA and psychotherapy actually interact.
It helps to see what a process-level explanation would look like. Rather than simply noting that MDMA increases trust, one might hypothesise that MDMA reduces defensive avoidance, which enhances a person’s ability to recognise and accept trustworthy information (what we call epistemic trust), which in turn encourages both trauma processing and therapeutic engagement, enabling further change. That is a sequence, with an order and a set of dependencies that can be tested. Producing hypotheses of that kind, rather than another inventory of effects, is what our study is designed to do.
Why psychoanalytic psychotherapists?
Building a process-level account depends on the quality of the descriptions you can gather, and here most people are limited: asked what just happened in their own mind, some may struggle to say. Psychoanalytic and psychodynamic clinicians are trained precisely against that limitation, which is why we are recruiting them as our participants.
Their orientation treats the therapist’s own inner responses as a source of clinical information; for example, countertransference, or the emotional reactions a patient stirs in the clinician, is understood as a window onto the patient's internal world. Working this way requires a cultivated capacity for self-reflection, supported by a training that mandates extensive personal therapy, usually at least twice weekly over several years. The result is a group of people with an unusually developed ability to observe their own mental states, to describe them with precision, and to bring theoretical frameworks to bear on what they notice. It is this that makes their first-hand accounts valuable as high-fidelity reports of what MDMA does. We are treating them, in effect, as expert observers of their own experience.
Around 25 clinicians will each receive two doses of MDMA, up to 120 mg, at least a week apart, under clinical supervision. Deliberately, no formal psychotherapy is carried out during the sessions, so that we can zoom in specifically on what MDMA does on its own, and participants are not led toward traumatic material, a point we return to below.
Professor Alessandra Lemma, one of the researchers on the study, explained:
"Psychoanalytic training teaches you to treat your own reactions as data, for example, that a shift in one's countertransference - a flicker of anger or boredom - tells us something about what is happening in the room with a patient. Under MDMA, these clinicians are better equipped to name the exact moment a defence such as intellectualisation or avoidance loosens, and to trace how that shift changes what they notice next, whether a fragment of memory surfacing or another person's words suddenly reading as trustworthy. That kind of process precision, moment to moment rather than before and after, is exactly what has been missing from the field's vocabulary so far."
Measuring what psychoanalysis has always tracked
The study collects experience at several timepoints and in several ways. Brief phenomenological interviews take place during the dosing sessions themselves, capturing the experience as it unfolds rather than in retrospect; deeper interpretative phenomenological interviews follow afterward, exploring how participants make sense of what happened; some may also involve a micro-phenomenological interview aimed at the fine detail of the moment MDMA’s effects first become apparent. Participants keep daily reflective journals throughout.
Alongside these, the study applies measures drawn from the psychoanalytic tradition itself. Using the Relationship Anecdote Paradigm interview, we will score each participant’s Core Conflictual Relationship Theme – Luborsky's method, often described as the first operational way of measuring transference, which captures the relational template a person carries into their close relationships. We will assess reflective functioning, the measurable capacity to understand behaviour in terms of underlying mental states. And we will track shifts in attitudes, alongside validated questionnaires covering compassion, empathy, epistemic trust, attachment and interpersonal problems. One concrete hypothesis is that MDMA may produce changes in the participant’s Core Conflictual Relationship Theme (particularly in what a person wishes for from others) with reflective functioning acting as a moderator of that change.
The point of gathering description and structured measurement together is to end up with something more than a set of testimonies: a model of MDMA’s psychological action that links what participants actually experience to constructs that can later be operationalised and tested in trials.
What this study won’t tell us
In keeping with Drug Science’s preference for evidence over enthusiasm, it is worth being clear about the limits:
This is not a clinical trial, and it will not show whether MDMA treats PTSD or any other condition.
It is open-label and uncontrolled, so expectancy and demand effects will shape what participants experience; we address this analytically rather than pretend it away.
Participants are healthy clinicians rather than patients, and they will not be processing trauma, which leaves out something central to how the treatment is actually used.
The psychodynamic training that makes these participants such acute observers is also a lens that may shape what they notice and how they frame it. For that reason, we separate their first-hand descriptions from their interpretations at every stage, eliciting them through different methods and at different times.
On the decision not to invite trauma processing: it would be ethically questionable to steer healthy volunteers into distressing material they would not otherwise have chosen to revisit. It also serves the science, letting us observe MDMA’s psychological effects without conflating them with the exposure-based work that clinical trials cannot separate out.
As study researcher Professor Sunjeev K Kamboj explained:
Psychopharmacology researchers have been slow to adopt phenomenological approaches to addressing questions about drug mechanisms. We’ve generally been more interested in studying brain changes or measuring behavioural responses that can be fed into computational models. But these methods haven’t brought us much closer to understanding how these drugs work in different psychological disorders and how (or even if) psychotherapy should be incorporated into these treatments. We think that qualitative approaches can help us narrow down the set of hypotheses worth testing.
Why it matters
If we could describe MDMA’s mechanism as a process rather than a list, the questions the field is currently arguing about would become tractable: which elements of the surrounding therapy are doing the work, for whom, and whether the changes MDMA sets in motion extend beyond trauma to other conditions. The intended output is a testable model grounded in theory, a bridge between first-person experience and the kinds of constructs that future clinical and experimental trials can measure. The approach is also meant to be a transferable template for studying the mechanisms of other psychedelic compounds through disciplined first-person inquiry rather than neurobiology alone. It may even help generate new measurement tools, built around long-standing psychoanalytic ideas that MDMA seems poised to affect – the loosening of defences, access to previously guarded material, the sense of a corrective emotional experience.
MDMA remains a Schedule 1 drug in the UK, and the study runs under a Home Office licence in a university unit with long experience of supporting people through altered states. Careful, ethically grounded psychedelic research is possible here, within a restrictive framework. Recruitment is underway, and we expect results in 2028.
Establishing that a treatment works has taken this field a long way. Explaining how it works (and doing so precisely enough to test the explanation) is the harder task, and it is the one worth turning to next. Our study hopes to do just that.
