Andrew Huberman interviews Stanford psychiatrist Dr Nolan Williams on psychedelics, brain stimulation, and depression treatment
Andrew Huberman speaks with Dr Nolan Williams, a dual-trained neurologist and psychiatrist at Stanford, about emerging treatments for depression and trauma.
Summary
Dr. Nolan Williams, a dual-trained neurologist and psychiatrist at Stanford, joins Andrew Huberman to discuss the neuroscience of depression and a range of treatments including transcranial magnetic stimulation (TMS), SSRIs, ketamine, psilocybin, MDMA, ibogaine, and ayahuasca. He argues that depression is best understood not as a chemical imbalance but as a circuit-level dysfunction — specifically, a failure of the left dorsolateral prefrontal cortex to regulate the anterior cingulate — and that this reframing is what he calls "Psychiatry 3.0." He presents Stanford's accelerated TMS protocol, which delivers the equivalent of seven and a half months of standard TMS in five days using spaced learning theory, achieving full remission in 60–90% of patients within one to five days. He also describes his ongoing research with former Navy SEALs and special forces personnel using ibogaine, reporting dramatic reductions in PTSD symptoms and moral injury, and explains the convergent brain-circuit changes seen across both TMS and psychedelic treatments.
Key Takeaways
FULL TRANSCRIPT
Depression as a Global Health Crisis and the Brain-Heart Connection
Andrew Huberman: Thanks for joining today. I'm really excited to have this conversation. I have a lot of questions about different compounds, psychedelics in particular. But before we get into that discussion, I want to ask you about depression, broadly speaking. I heard you say in a wonderful talk that you gave that depression is perhaps the most debilitating condition worldwide. Yet in contrast to other medical conditions like cancer, we actually have a fairly limited number of tools to approach depression — and yet the number of tools and the potency of those tools is growing.
Dr Nolan Williams: Depression is the most disabling condition worldwide. What's interesting about depression is it's both a risk factor for other illnesses and it makes other medical and psychiatric illnesses worse. Recently the American Heart Association added depression as the fourth major risk factor for coronary artery disease — alongside hypertension, high blood pressure, hyperlipidemia, high cholesterol, and diabetes. Those three have been on the list for a long time, and depression ended up being added as the fourth one.
A lot of what we're doing in the lab is measuring brain-heart connections. With transcranial magnetic stimulation — a form of brain stimulation — we can actually decelerate the heart rate. We can capture that heart rate deceleration over the mood regulatory regions, so it's actually a direct probe of that connection.
We've been very interested in a particular clinical set of problems around the most severe and highest acuity settings that people with depression end up in — emergency settings, inpatient units. The field really hasn't developed a way of consistently being able to treat that problem. People end up getting the same standard oral antidepressants they've been getting as outpatients. I came to this because I dual-trained as a neurologist and psychiatrist, went back and forth between the two, and saw that in neurology we have all of these ways of treating acute brain-based problems. I really wanted to emulate that in psychiatry and find ways to develop and engineer new brain-based solutions.
Huberman: Many people out there probably think of the relationship between the heart and the mind as kind of woo, or a soft biology. But here you're talking about an actual physical connection. What area of the brain is it?
Williams: The first place where the stimulation goes is called the dorsolateral prefrontal cortex — it's kind of the sense of control, the governor of the brain. What we know is that when you use a magnet — what we call Faraday's law, this idea of using a magnetic pulse to induce an electrical current in electrically conducting substances, in this case brain tissue but not skull or scalp or hair — you have a direct depolarization of cortical neurons, the surface neurons of the brain, in this dorsal prefrontal region. If you do that inside an actual scanner, which we can do, you can see that signal distribute down into the anterior cingulate, the insula, and the amygdala, and ultimately the tract goes into something called the nucleus tractus solitarius and ultimately into the vagus nerve and into the heart.
The heart very consistently seems to be the end organ of the dorsolateral prefrontal cortex. If you do the same stimulation over visual cortex or motor cortex, you don't get any of those findings. It's really specific to this control region of the brain. Our work and other groups' work — Martin Arns in the Netherlands — has shown the same connections. I think it's been replicated four or five times.
Where I think TMS is really interesting is this: we had a lot of patients who told me their therapist said they weren't trying hard enough in therapy. These are moderate to pretty severely depressed patients. As soon as we get them well with the TMS approaches — a rapid five-day approach — the next week they come in and say, "You know what I did all weekend? I looked at my therapy books and now I can understand it." So I actually see TMS as a way of providing exogenously the sorts of cognitive functions that in milder forms of depression we can pull off with psychotherapy — this idea of being able to turn that prefrontal cortex on and have it govern these deeper regions.
In depression, the deeper regions govern the prefrontal cortex. In one case it's like the coach telling the player what to do; in the other case it's like the player telling the coach what to do. You restore order to the game.
You restore order to the game, and what it looks like is that depression is a bunch of kind of spontaneous content that's semi-volitional, being generated out of this conflict detection system — the cingulate. In depression, it looks like the left dorsolateral does not sufficiently clamp down on it. What therapy appears to do is restore that. What we see with TMS over that region is that we exogenously do the same sort of thing — we restore the governance of the left dorsolateral over the cingulate area, and that is correlated with treatment improvement. The degree to which you can re-time or re-regulate the left dorsolateral over the cingulate, the more of an antidepressant effect you have.
TMS is almost like exercise for the brain. You're exercising this region over and over again with a physiologically relevant signal and turning that system on. What's interesting is we've had probably five or six folks who, if they remit early enough in the week — we have this very dense stimulation approach where we can stimulate people really rapidly over a five-day block — by Wednesday they're at zero on the depression scales, even better than most people walking around, really no anxiety, no depression at all. By Thursday, the first person who told me this came in and said, "I was driving back to my hotel and I decided to go to the beach and I just sat there and was totally present in the present moment for an hour. I read about this in my mindfulness books, but I experienced it last night and I've never experienced anything like this before."
I didn't tell any more patients about that. But about five more over the last couple of years, when they got well early in the week, by the end of the week they were going to the beach and having what people describe as a pretty mindful, present-moment experience. I don't have full scientific data to tell you what that is, but it's an interesting anecdote — that when you push people through this point of feeling clinically well, some of them end up reporting this additional set of features.
SSRIs and the Shift to Psychiatry 3.0
Huberman: I want to make sure that before we dive into ketamine and psilocybin, we touch on SSRIs — selective serotonin reuptake inhibitors — because we can't really have a discussion about depression without talking about them. My understanding is that SSRIs are powerfully effective for certain forms of obsessive-compulsive disorder and may also be effective for treatment of depression. How should we think about SSRIs? Are they useful?
Williams: SSRIs clearly work. Many meta-analyses prove that out — in a subpopulation of individuals they achieve great benefit for depression, for obsessive-compulsive disorder, for generalized anxiety disorder, panic, all these things. You can see an improvement in those symptoms with SSRIs.
The issue is that they don't work immediately. They don't work the same day you start taking them. That suggests that probably it's not exactly the serotonin being present that's directly driving the effect — it's much more likely that there are some brain plasticity effects. There's not a deficit of serotonin. You're not born with what people call a chemical imbalance. Psychiatry has known this. This is not actually new information — it's kind of a rehashing of things we've known for a while, though in the lay press it's hit in a way it didn't seem to grab attention before.
I think what's really important is the idea that the chemical imbalance model is wrong. What I'll call Psychiatry 1.0 — the Freudian era, psychotherapy and its origins — was a lot around your family, your experiences, and psychotherapy going in and helping you figure out or see certain cognitions that aren't helping you. Things like the schizophrenogenic mother — that was a concept at some point. We transitioned from that to Psychiatry 2.0, the chemical imbalance idea, the notion that something is chemically missing. The trouble for a patient is that it sends the message: there's something missing with me — whether it be experiences I had no control over as a child, or a chemical in my brain.
What I think is really powerful with TMS — and at a level even powerful with the psychedelic story — is that it says something different. TMS works and there's no serotonin coming in or out of the brain. We're doing a rapid form of TMS that works in one to five days, so it's very unlikely that there's some long-term upregulation of serotonin driving that. Our work actually pushes back on the serotonin hypothesis as being the center of depression, because it says: we're not giving anybody any serotonin. We're simply turning these brain regions on. We're focused on the circuitry. That's Psychiatry 3.0.
Neuromodulation is a really nice use case for Psychiatry 3.0 because it's a way to focally and directly perturb brain regions. But there are also a lot of groups doing neuroimaging before and after and seeing circuit-level changes for something like psilocybin or ketamine long after the drug is gone — in those same brain regions that converge. The subgenual default mode network connection that we see changing with our Stanford Neuromodulation Therapy technique is that same set of brain regions that ketamine and psilocybin seem to act on. These connections between brain networks shift. And so it refocuses the story on something that's highly correctable — it's basically electrophysiology, basically recalibrating a circuit that is recalibratable — instead of "I have something missing" or "I have a set of experiences from early in life that are going to forever trap me in these psychiatric diagnoses."
That's what's so powerful about Psychiatry 3.0 — focusing on the circuit — because it gets us thinking about psychiatric illnesses as something that are recoverable. People can get better. We've seen it with our TMS techniques and with some of the psychedelic work we've done, where people are actually at normal levels of mood for sustained periods of time.
Huberman: Within five days.
Williams: Within five or fewer days. And in the case of psychedelics, within a few days. We can get people out of these states — they're totally well, there's no drug in their system at that point. In the case of TMS, there was never a drug in their system. And it just tells us that it's fixable. It's just like an arrhythmia in the heart. It's like a broken leg. We can go in and do something and get somebody better.
What's empowering — and what a lot of patients have told me — is that even some people who relapse and need more stimulation or more psychedelics will say, "I don't fear that I'm chronically broken. I don't fear that the chemical imbalance is still imbalanced. I don't fear that things I couldn't control in my childhood are going to drive this problem forever." And I think that's what's so powerful about this.
Psilocybin, Trauma, and Why the Brain Holds On to Unhelpful Rules
Huberman: That brings me to the question about psychedelics and the frankly altered thinking and perception that occurs in high-dose psilocybin clinical sessions. Many people do report improvements in trauma-related symptomology and depression after taking psilocybin, because during those sessions something comes to mind spontaneously. They report a new way of seeing the old problem. The old problem could be the voice that says they're no good, that nothing will ever work out — or it could be even more subtle than that. Why do you think the brain would ever hold on to rules that don't serve us well?
Williams: I think it's an evolutionary neurobiology answer. We end up being a result of biology that's probably not that useful in the modern era. For PTSD, for example — a lot of veterans come back and experience PTSD symptoms that are not at all useful back home. They hear a loud noise and they're behind a car, or I've heard of folks jumping behind a trash can in the middle of San Francisco. But if you put them back on the battlefield —
Huberman: Highly adaptive.
Williams: That's highly adaptive. We hold on to those things from an evolutionary neurobiology standpoint. But what seems to alleviate that — for whatever reason — are these substances. Some newer, like MDMA; some that have been around for thousands of years, like psilocybin. They seem to have a therapeutic effect that's pretty long-lasting for these phenomena. It's curious that in the absence of these substances these things will keep going on and on, but in the presence of that exposure you see a resolution of the problem.
We have some work now treating Navy SEALs, and the anecdotes we're getting are that people are coming back saying these PTSD symptoms are finally gone. This idea that going into what's probably a highly plastic state, re-experiencing memories, and then reconsolidating them in that state — for whatever reason — may drive a therapeutic effect.
My business is to find treatments that help people, so I'm much more pragmatic about it. If this sort of thing — which has a lot of cultural baggage — ultimately ends up being therapeutic, if we can design trials that convince me and others that it is, then we should absolutely use it. And if it doesn't, then we clearly shouldn't.
MDMA for PTSD and Psilocybin for Depression: Clinical Trial Results
Huberman: What do the clinical trials generally say? Let's start with psilocybin and MDMA.
Williams: MDMA appears to — in one to a few sessions — have an anti-PTSD effect that seems to be outside of the standard assumed levels of PTSD improvement you can observe in individuals with this level of PTSD.
Huberman: So for people who have trauma, who do — and again we're talking about a clinical setting — one or two doses of MDMA, the standard MAPS dose being 150 to 175 milligrams, doing this with a physician in a controlled clinical trial, legally — they do it once or twice, and broadly speaking, what percentage of people with trauma report significant relief afterward?
Williams: About two-thirds of people had a clinically significant change in their PTSD.
Huberman: That's impressive. And how long-lasting was that?
Williams: It appears to last for a while. In the earlier trials where they followed people out, it seemed to last in the years range for some people. It's pretty compelling. Contrast that with ketamine, which on average only lasts about a week and a half for a single infusion — a much shorter duration.
Huberman: So they have to get repeated infusions of ketamine every ten days or so?
Williams: For some people. Or they end up getting a bunch of doses over a couple of weeks, and for some people that seems to last a while. But I think the psilocybin story for depression and the MDMA story for PTSD seem more interesting to me.
Huberman: So for psilocybin, what are the rough percentages for relief from depression?
Williams: In open-label studies it's closer to half to two-thirds of people getting better, depending upon their level of treatment resistance. In the blinded trials it was more like a third or so.
The Neurochemistry of Psilocybin
Huberman: Let's talk a little about the neurochemistry of psilocybin. What's going on when one takes psilocybin, and why is it interesting in light of depression?
Williams: David Nutt and Robin Carhart-Harris's work on neuroimaging of psychedelics were some of the first to do that work. To their great surprise, they thought there was going to be an increase in activity on psychedelics, and what they found was the opposite — an overall decrease in the level of activity in the brain. But they've also looked at connectivity, and there's this small-world versus large-world connectivity to think about. Small-world meaning much more focused cortical or subcortical function, and what you see is a difference in that level of engagement of brain regions — global connectivity kind of increases. It's still to be determined, and there's a lot of work that needs to be done, but it's certainly suggestive of pretty profound changes in brain activity and connectivity.
What we found to be really interesting is that the antidepressant effects of psilocybin have a particular connectivity change that we also see with our TMS approaches — the connectivity between the subgenual anterior cingulate and the default mode network. When we do effective Stanford Neuromodulation Therapy stimulation, we see a downregulation of the connectivity between the negatively valenced mood state — in depressed individuals — and the self-representation of the brain. You see that same connectivity change occur post-psilocybin, suggesting a convergent mechanism.
It makes sense: you've got an overconnected, negatively valenced conflict system that's kind of attached onto the self-representation, and people feel stuck. When you do whatever effective thing you do, it unpairs those two systems.
Ibogaine: The Longest-Acting Psychedelic and Research with Special Forces Veterans
Huberman: I want to ask you about ibogaine. Is it legal in the US as a clinical tool? Who's using it and for what purposes?
Williams: Ibogaine is one of the alkaloids you can extract from an iboga tree root bark that's typically grown in Gabon, Africa. What individuals taking ibogaine will say is that with eyes open they don't see anything, but with eyes closed they go back through and re-experience earlier life memories — and they're able to experience it from a place of empathy, not only for themselves but for others, with a kind of detached empathy, being able to see events almost as a third party even though they were there. Ibogaine is in no way a recreational substance. You're essentially having what they call a life review. They also call it ten years of psychotherapy in a night.
Huberman: How long does it last? Is it truly one night?
Williams: Depending upon how fast you metabolize it — sometimes 24 hours, sometimes 36 hours, sometimes shorter. But it is a long time. It's definitely the longest-acting psychedelic substance I know of.
Over the last couple of years we've been able to do this first-in-human, full neurobiological and clinical neurocognitive evaluation of what ibogaine is doing — in this case in special operations and special forces individuals, former Navy SEALs, former Army Rangers, that crew of folks — and look at the pre-post changes they're experiencing and totally quantitate all of that. We've captured all the clinical scales: depression scales, PTSD scales, all the standard stuff; neurocognitive batteries covering executive function, verbal memory, all of that; neuroimaging; and EEG. This will be the first human study of ibogaine for those measures.
The reason ibogaine hasn't been studied as much is that it has a cardiac effect. It seems that you can screen people out who have risk based on their electrocardiogram and reduce the risk quite a bit, which is what we did. But that's why people haven't really studied it as much. And nobody goes to a rave on ibogaine — there's no recreational use at all.
Huberman: It's not fun. People say that it's relieving, but it's hard work, because you're re-examining things.
Williams: Exactly. We see these folks afterward and — we haven't fully analyzed the data yet — but from what my people are telling me, it's pretty dramatic. People come back and they're doing a lot better. Soldiers experience something called moral injury, where maybe they accidentally caused an explosion that killed a child, or a civilian died — vocational risks of the job. They come back and say they've forgiven themselves, which is huge. Part of that is being able to see themselves in a different light, having empathy for themselves finally, and being able to have that experience of forgiving.
There's this Timothy Leary sociocultural construct that gets overlaid over psychedelics. What I think is that if you rid yourself of all those preconceived notions — the counterculture movement, all of that — and re-examine this as straight scientists looking at it: if we just discovered these substances today, we would say they are a huge breakthrough in psychiatry. They allow us to do a lot of the things we've been thinking about with SSRIs and psychotherapy, but combined — psychotherapy plus drugs in a substance that allows you to re-examine these things.
There's a lot to do to figure out if that's true. As it stands right now, we don't know if that statement is fully proven. But the hypothesis is that if it is true, this will be seen as a breakthrough because it allows you to do things you can't do with normal waking consciousness.
That's also why we have to really think about this carefully. These drugs can't be recreational drugs. They really shouldn't be recreational drugs — they're too powerful to be used in the context of recreation. This generation of psychedelic researchers is really clear about that. The 1960s folks were not clear about that. But this cohort really understands that in order to make this work, we have to recognize that if you need a prescription for an SSRI — which doesn't change your consciousness a whole lot, and we're very worried about that, and the doctor has to evaluate you for it regularly — then the idea that some of these substances would go outside of very strict medical supervision is kind of preposterous. If we're going to do this right, we've got to do it in such a way that's so protected and so safe that people understand these things are not recreational, and they're really for the pure purpose of powerfully changing cognition for a while and letting people have these seemingly therapeutic states.
Ayahuasca: Pharmacology, Safety, and the Brazilian Prison Study
Huberman: Tell me about ayahuasca as a plant medicine. Is it useful for the same sorts of conditions we've talked about? And could you tell me a little about the Brazilian prisoner study?
Williams: Ayahuasca is another psychedelic used as a sacrament in Brazil, Peru, Ecuador, and Colombia — a lot of South American countries. What they do is combine two plants together, where one of the two would effectively do nothing on its own, but the two-plant combination together is capable of producing a very profound psychedelic effect. What's really curious is that there are, as I understand it, ten to twenty thousand plant species in the Amazon, and somehow somebody —
Huberman: Someone tried them all.
Williams: — combined these two plants together in a certain proportionality and cooked them for five to ten hours, to the point where you cook the dimethyltryptamine out of one plant and the reversible monoamine oxidase inhibitor out of the other. The reversible monoamine oxidase inhibitor prevents the gastrointestinal breakdown of the dimethyltryptamine in such a way that it's then allowed to cross the blood-brain barrier and get into the brain. If you didn't add the reversible monoamine oxidase inhibitor, it would never cross into the brain. If you put people on a standard psychiatry-prescribed monoamine oxidase inhibitor that wasn't reversible, you'd throw them into serotonin syndrome. So there's this kind of sweet spot that ayahuasca practitioners have found — getting DMT into the brain from an oral source with this combination. It's curious.
That substance has been explored as an antidepressant agent and some studies have looked at that. It also appears to be very safe. There's a psychiatrist at UCLA Harbor who's done a lot of work with this, looking at children even who've been exposed to small doses of ayahuasca as a sacrament within Amazonian tribes, and found no neurocognitive effects — in children or in adults. It's part of various religions, including some merged with Catholicism in South America, which is very interesting. In some sects of Catholicism in Brazil it's used as a sacrament during religious ceremonies.
That became interesting to Brazilian researchers as to whether or not they could affect recidivism rates for prisoners in Brazilian prisons. They gave half the prisoners some sort of inert substance and half an ayahuasca session. The recidivism rate — the return-to-prison rate — in the ayahuasca-exposed individuals was statistically significantly lower than in the control group, suggesting that whatever is going on there seems to have an effect on whatever drives criminal behavior. I don't have the details on the exact nature of the crimes. I am also in no way saying that we should be giving psychedelics to people in prison — I think that's a very edgy thing to do and probably not something anybody should try. But it does bring up this curious question of what it is about that experience that would drive people to change those behaviors, and why people make those behavioral decisions.
Stanford Neuromodulation Therapy: Accelerated TMS and the Spaced Learning Protocol
Huberman: Before we wrap, I do want to give you the opportunity to talk about the SAINT study.
Williams: SAINT — or what we're calling SNT now — Stanford Accelerated Intelligent Neuromodulation Therapy, or now Stanford Neuromodulation Therapy. The idea is that TMS is a device that delivers a treatment, and the treatment is the protocol — the stimulation parameter set in a specific brain region for a specific condition. Whether it be transcranial magnetic stimulation, transcranial direct current stimulation, or deep brain stimulation, in all of those cases the device itself is a physical layer conduit of a stimulation protocol that's therapeutic for a given condition in a given brain region.
We decided to address the problem I talked about at the beginning — we don't have a treatment for people in high-acuity psychiatric emergency states. The idea was to engineer a treatment where we reorganize the stimulation approach in time to be much more efficient by utilizing something called spaced learning theory.
The idea for viewers: if you're cramming for a test, you write out sixty note cards and read each one for a minute until you get back to the first card — about an hour later. That's spaced learning theory. You need to see the material about every hour to an hour and a half, and that optimizes learning. What we found was that the old way of doing TMS — once a day, every day, five days a week for six weeks — didn't utilize spaced learning theory. It's like studying a little bit once a day for a month or two. You remember some of it, but it's not as potent as that week where you're cramming.
We realized that if we reorganized the stimulation in time, we could take the whole six-week course and actually do it in a day. We also figured out that people were underdosing TMS, because if you just kept going after six weeks out to months three, four, and five, more and more people got better. So it's not just one day — we're going to give five times the normal dose. We're going to give seven and a half months' worth in five days using spaced learning theory.
Huberman: Every hour for ten hours —
Williams: For five days.
Huberman: For five days.
Williams: So it's a fifty-hour block. It's ninety minutes of actual stimulation but spread out through the day in the same way as learning. What we've found is that within one to five days — in more cases than not, somewhere between 60 and 90% of the time depending on whether you're looking at open-label or trial data — people will go into full remission. They're totally normal from a mood standpoint at the end of this. With variable durability, which is the part we have to figure out now in terms of dosing and how to keep people well. But for some people we've had four years of remission, a year of remission.
It really is that cramming of the test. You're laying in that information to exactly the right spot. The signal is a simple one but a profound one: turn on, stay on, remember to stay on. You're sending this memory signal into the brain in such a way that you're taking it out of the hippocampus's hands and sending the same signal the hippocampus normally sends out — now into the prefrontal cortex — and utilizing the brain's own communication style to get it out of this state.
What's very cool is that when people exit out of that, they say they don't have any side effects and they feel back to normal.
Huberman: Thank you so much for taking us on this incredible voyage through the neurocircuitry underlying certain aspects of depression, the coverage of the different types of depression, the various therapeutic compounds, and how they work. You've shared so much knowledge, and even as I say that I very much want to have you back to talk about many other things we didn't have time to cover. To take the time to sit down with us and share all this knowledge that is really in service to mental health and human wellbeing — and in fact often averting suicidal depression — it's just incredible work and an incredible generosity. Thank you so much.
Williams: Absolutely. Thank you.