Psychedelics for Treating Mental Disorders | Dr. Matthew Johnson

Huberman opens with the question that should be simplest and turns out not to be: what qualifies a substance as a psychedelic? Johnson answers by taking the word apart. As a label it is not pharmacologically satisfying, because it spans several classes of compounds, and its real content sits at a high level of abstraction: these drugs can profoundly alter a person's sense of reality, and acutely, the sense of self. He is also candid that the boundary is contested inside the field, and that not everyone will accept the wide version of the category he is sketching. The definition, in other words, is a description of an effect rather than a statement about a mechanism.

The classes Johnson places under the umbrella: the classic psychedelics, meaning LSD, psilocybin (found in what are called magic mushrooms, over 200 species known so far), DMT (present in dozens of plants), and mescaline (in peyote and other cacti such as San Pedro); the NMDA antagonists, meaning ketamine, PCP and dextromethorphan; and edge cases such as salvinorin A, a kappa-opioid agonist from Salvia divinorum. Within the classic group he separates two structural starting points, the tryptamines (psilocybin, DMT) and the phenethylamines.

Chemistry alone does not settle membership, and Johnson returns to this point repeatedly. At the receptor level, the classic psychedelics share one property: they act as agonists or partial agonists at the serotonin 2A receptor. The NMDA antagonists do something entirely different yet produce a large overlap in subjective effects. MDMA is a third case again, structurally a phenethylamine like mescaline but also like amphetamine, and it does not work by occupying the postsynaptic receptor at all. Johnson describes MDMA as acting on what he calls the pitcher side, releasing more of the endogenous neurotransmitter rather than mimicking it in the glove.

It means touching within. It alludes to the idea that it can really put someone in touch with their emotions.

The entactogen language matters for the rest of the conversation. Johnson notes the older term empathogen, meaning it can afford empathy, but says entactogen is the term he tends to focus on, and that the primary mechanism is serotonin release, with some release of other monoamines as well. This is a different pharmacological strategy from the 2A agonists, which is exactly why he resists building a taxonomy on molecular shape: you can keep the same backbone and profoundly change how a compound interacts with a receptor.

Huberman tries to break the experience into components: hallucination in either modality, synesthesia, and a disturbance in the brain's computation of space and time. Johnson treats the space-time framing as an interesting hypothesis rather than a settled account, and reaches for comparative pharmacology to test it. Alcohol, he points out, badly degrades orientation in time and space, and so do the benzodiazepines, which behave much like alcohol; if the criterion is loose enough to include everything that scrambles orientation, it stops distinguishing anything.

We're prediction machines, and so much of that is top-down, and psychedelics have a good way of, loosely speaking, dissolving those models.

The model language is made concrete with an example Huberman supplies: a child learns that a thrown ball falls down rather than up, and that expectation is not innate but acquired. Johnson agrees that a ball rising into the sky would be astonishing to a four-year-old and to an adult alike, and adds dryly that it would probably also be a pretty psychedelic experience. What the drug does, on this account, is prevent the violated prediction from being filtered out, so that the error signal arrives as surprise and recognition rather than being quietly absorbed.

Johnson is careful about how far that idea can be pushed. He mentions credible but very atypical reports of people who believed they could fly and jumped, while noting that the late-1960s propaganda inflated the frequency, and that alcohol causes far more fatal falls every year. His own laboratory has a milder version on video: a volunteer who appeared to be trying to dive through a painting on the wall. She was fine, but reviewing the footage it looked as though she genuinely expected to pass through it.

The perceptual mechanism he keeps returning to is dishabituation. Habituation is the simplest form of learning, the decline in response to a repeated stimulus, and without it nobody could function. Psychedelics appear to suspend it locally, which is why a hand can become the most amazing miracle in the universe while the person holding it is standing at a curb in Manhattan. Johnson's point is not decorative: it is that the drug is not adding content so much as removing the suppression that makes ordinary content usable.

Huberman extends this with a vision-science observation: stare at any fixed target long enough and it fades, which is why microsaccades exist. The default brain jumps around. On psychedelics, the therapeutic work looks like the opposite of that default, drilling into one narrow percept and expanding it from the inside. Johnson adds the receptor caveat that keeps the biology honest: serotonin itself cannot be the whole story, since otherwise a meal would be a trip.

Clearly, serotonin itself is not psychedelic, or else we'd be tripping, all of us all the time.

What is known downstream is thinner than the confidence of the public conversation suggests. Johnson describes 2A activation followed by increased glutamate transmission as a likely commonality, which would help explain why ketamine, working through NMDA antagonism, can be intensely psychedelic in a different register. Beyond that he is explicit that this is an area of active exploration without great answers, and that the field knows the receptor-level pharmacology far better than the post-receptor signaling that connects it to a changed self.

The clinical procedure Johnson describes is deliberately unglamorous. Screening runs over a couple of days and consists of structured psychiatric interviews covering past history and symptoms across the DSM, looking for disqualifying conditions. The main exclusions are the psychotic disorders, schizophrenia specifically, and bipolar disorder on the manic side, along with cardiovascular screening for heart disease. Only after that does preparation begin, and preparation has two jobs: building therapeutic rapport with the people who will be in the room, and explaining didactically what the drug might do.

That explanation is necessarily a laundry list, because psychedelics are known for their variability rather than their consistency. Johnson contrasts them with cocaine, which reliably stimulates, and alcohol, which reliably relaxes, and calls his own shorthand uppers, downers and all-rounders. The point of the list is to remove surprise by naming the whole range in advance, including the possibility that the session will be the best or the worst experience of the person's life.

you could have the most beautiful experience of your life or the most terrifying experience of your life

What is administered is not the mushroom. Participants receive pure psilocybin, a single molecular entity, a white powder, delivered in a small pill. Johnson notes that the mushrooms people typically encounter, most likely Psilocybe cubensis, are easy to grow and require almost nothing in the way of equipment, which is part of why they circulate so widely outside research. Inside the study, dose is measured in milligrams and the session lasts most of a day.

The dose range comes with a methodological detail worth keeping. Early work in healthy volunteers used a true placebo plus four active doses of five, ten, twenty and thirty milligrams, adjusted for body weight and expressed per seventy kilograms. Newer trials have dropped the weight adjustment, because across hundreds of volunteers the team concluded it was unnecessary. Most studies aiming at a psychedelic effect now sit in the twenty to thirty milligram range, though participants heavier than the seventy kilogram reference have effectively received more, in many cases closer to forty or forty-five milligrams. Onset takes anywhere from fifteen minutes to an hour, averaging about half an hour.

The session day itself is not filled with tasks, and this is a deliberate trade-off rather than an oversight. Research has found that the experience is less meaningful inside an fMRI scanner or while performing a battery of cognitive tests. Johnson has run that kind of study and says colleagues have too, but in a therapeutic protocol the priority is the quality of the experience, which means accepting how little can be measured while it is happening.

The container is built from small physical choices. Participants wear eye shades, which the team refuses to call blindfolds because of the connotation of being kidnapped; Huberman suggests inner sight shades instead. Closing the eyes does not silence the retina, it substitutes spontaneous activity for input, and Johnson describes the mind's eye as being on rocket boosters. He is also precise about what people do not typically see: pink elephants belong to the anticholinergics, drugs like atropine and scopolamine, which produce true hallucinations of people who are not there.

The instruction that does the most work is permissive rather than directive. Johnson tells people that any emotional response is welcome, including sobbing, and he notes that experienced users can be harder to train in this model because the ordinary social rule is the opposite. Someone who has taken mushrooms at a party has learned to hold it together for the sake of the group. The clinic has to undo that training before it can proceed.

in the real world, people with psychedelic experience, a lot of times the rule is, hold your shit.

Safety is handled as a monitored physiological event, not only a psychological one. Blood pressure and pulse are taken roughly every half hour, with the intervals spreading out over the time course. If pressure crosses a threshold the protocol brings in a physician, who administers nitroglycerin under the tongue; this has been needed only a few times, and it lowers blood pressure without altering the experience. The reassurance matters because the body can produce convincing evidence of catastrophe. Johnson remembers a woman on the couch asking him to tell her again that she could not die, certain that her heart was about to tear through her chest.

Matt, tell me again I can't die. I feel like my heart is going to rip through my chest.

Ketamine gives Johnson a natural experiment in framing. Spravato, the intranasal form marketed by Janssen, is esketamine, a prescription product approved for treatment-resistant depression. It is not delivered as psychedelic therapy. By that term Johnson means a specific package: telling the person they will have an altered experience, telling them to attend to it because they might learn from it, and discussing it afterwards. With Spravato the altered state is classified as an adverse effect to be disregarded.

It's not treated as psychedelic therapy.

The prescribed dose is low, in the mild psychedelic range but not strong, and the antidepressant effect arrives immediately while lasting about a week. Johnson treats a week as genuinely impressive by psychiatric standards, since most drugs in this domain are taken daily, but the comparison to psilocybin is where the interesting question sits: with the classic psychedelics the team is seeing effects a year or more later. He does not resolve the gap. It could be pharmacological, or it could be that ketamine delivers far more if the psychological experience is treated as material rather than noise.

That could be a pharmacological difference.

The historical precedent he cites points in the second direction. Krupitsky, working in Russia, ran extensive high-dose ketamine work in the 1990s and early 2000s and treated it exactly like psychedelic therapy: preparing people for the experience, helping them through it, discussing it afterwards. Johnson reports that this produced very high rates of success in some well-controlled trials for heroin addiction and for alcohol addiction. He uses this not as proof but as the reason a properly randomized comparison of framed versus unframed ketamine is worth running.