5 Questions for Robin Carhart-Harris

In 2014, Robin Carhart-Harris became the first person to legally administer LSD in the UK since the drug was outlawed in 1971. His work studying the brain activity of those participants — as well as participants using MDMA, DMT, and psilocybin — has established Carhart-Harris as one of the best-known scientists in psychedelic science. After earning a PhD in psychopharmacology, Carhart-Harris got a job at Imperial College London, where he spearheaded some of the earliest neuroimaging work looking at how psychedelics affect brain activity. Over the years, he’s authored dozens of psychedelic papers, including studies probing the feasibility of using psilocybin to treat depression and the neuroscience of ego dissolution.

In 2019, he founded the Centre for Psychedelic Research at  Imperial College London, then the world’s first research institute dedicated to psychedelic work. In March 2021, Carhart-Harris announced another new initiative: helming the psychedelics division at the University of California at San Francisco’s Neuroscape, a neuroscience center. The Microdose talked with Carhart-Harris about the research questions he and his team at UCSF are drawn to, and how these areas of study can help inform psychedelic therapy practitioners and their patients.

You spent over a decade at Imperial College London, and helped establish the institution as a leader in psychedelic research — it must have been a big decision to leave. What precipitated the transatlantic move?

Culture is changing around psychedelics and policies are changing around psychedelics, and they seem to be changing faster in certain pockets of the States than anywhere else. I wanted to be close to that. One state up, in Oregon, psilocybin therapy is being legalized — strides in a similar direction could soon be happening in California. That could present lots of interesting opportunities for learning and healing. 

And then there are other silly reasons: I quite like the weather here, and it was a good time for my family to make a move. Plus, there's something about a move that really freshens things up — all that novelty and healthy uncertainty is useful for resetting all your models and assumptions so that you can learn new things. 

Speaking of new things: are there new directions you’ll be exploring at UCSF?

We're designing a great program of research that will further scrutinize the psychedelic experience and its different substates: states of struggle, states of letting go, of release, of emotional breakthrough, and how they influence the trajectory of the therapeutic process. 

We’ll also be systematically examining how different extra-pharmacological factors — set and setting and other contextual factors — influence the therapeutic process. The different variables are quite rich. You've got your music, you've got your relationship with the guides, you've got scent. Maybe at the beginning or end of therapy, you can use visual stimuli such as nature footage to induce therapeutically useful experiences, like a sense of interconnectedness with nature. Perhaps we can use kinesthetics — a warm breeze or a cool breeze, or the grounding influence of touch when it is done skillfully and appropriately — to help ground and anchor people. Then there’s the interpersonal stuff: what about community support afterwards? What about taking up meditative practice afterwards to cultivate healthy change and remain open-minded and avoid old unhealthy habits? 

Sets and setting won't be one size fits all. We might find one size that's pretty good for most people. If we can personalize both ahead of sessions, but also be responsive and dynamic with the package such that it might change for a given individual if their feelings change: they might want to change up the music. Maybe things need to be done around the therapeutic alliance between practitioners and patient mid-process. All this finessing is where psychedelic therapy needs to go. It needs to show, loudly and clearly, that this is a combination treatment, meaning it's not pure pharmacotherapy, not pure medicine. It's about combining pharmacological actions that increase your sensitivity with extra-pharmacological factors like set and setting. We need to establish that first and then learn how to optimize that setting, and that's the program of research that we're drawing up at Neuroscapes. 

There has never been a more exciting – or bewildering – time in the world of psychedelics. Don’t miss a beat.

That combination of factors is so important to study — but it’s also really difficult to carefully control the messy variables of human experience. What might study design for this kind of work look like?

There was a time when I think we were right to do an open label study on treatment-resistant depression, because it moved the needle and brought investment money in. [Editor’s note: Typically, it’s considered good scientific practice to make sure neither researchers nor participants know what treatment they’re receiving, but in an open label study, both participants and researchers know what treatment participants are receiving.] And there was a time to do a nice, controlled, double-blind, randomized controlled trial to show how good psilocybin treatment is relative to current standard practice — we did that in our New England Journal of Medicine trial, despite the very conservative framing of the results in the abstract. If people look at the data, they'll see that psilocybin therapy looks pretty damn good next to our current set of leading drug treatments for depression. That was necessary to do to set the context for what psilocybin therapy could be. But perhaps now is a time to take some of those experimental control constraints off a little bit, so that we can accelerate a degree of exploration to better understand how to improve psychedelic therapy. If you continue to constrain and fix all your parameters, how can you learn? If all you look at is one treatment session model — for example, X number of hours of prep and integration — then you get kind of stuck. We have to lift the constraints. It’s time for some hybrid designs, maybe multiple designs: naturalistic research, pragmatic trials, and citizen science initiatives.

I know naturalistic research is studying people in their usual context, and citizen science initiatives recruit people to collect data on themselves or the world around them. What does a pragmatic trial design look like?

A pragmatic trial is different from a double blind, randomized controlled trial in that it tries to replicate the conditions of clinical care. If you have a licensed treatment and you're just giving it to people, that data is real world evidence. But if you're doing the trial prior to licensing, you might call that a pragmatic trial. As a clinician, considering the patient that's in front of you, you might feel that another round of treatment would be beneficial at three or six months in — but in a double-blind randomized controlled trial, you couldn't do that because it would affect the scientific rigor of the design. In a pragmatic trial, you could. Pragmatic trials compromise on scientific rigor in the interest of ecological validity — the studies would replicate natural conditions more faithfully, and enable more exploration because you can try things; you're not constrained by your highly controlled experimental design. 

At this point, it seems quite likely psychedelic medicalization will catch on beyond Oregon. How do you see your work — past and future — playing into that?

There's a danger that practitioners might think, “Well, Hopkins and Imperial and NYU did X, Y and Z, and that seemed to work quite well. Let's just do that.” But we didn’t design those studies in the belief that they are the optimal designs for therapeutic care. They were just appropriate designs for the questions that we asked at that time. But it's amazing, especially when the field is really young, how influential that early work is. It gets too heavily weighted; it's too influential. It's going to be a healthy situation now that there are so many different teams around the world doing different trials and different designs. There’s going to be a huge amount of learning in the next few years. 

This interview has been edited and condensed for clarity and length.