Understanding & Healing the Mind | Dr. Karl Deisseroth

Andrew Huberman introduces Dr. Karl Deisseroth, a Stanford psychiatrist and research scientist, as the first guest of the Huberman Lab podcast. In Huberman's description, Deisseroth's clinical practice covers obsessive compulsive disorder, autism, attention deficit disorders, schizophrenia, mania, anxiety disorders and eating disorders. His laboratory builds tools for understanding how the nervous system works both in health and in disorders of the mind.

Huberman centers the introduction on what he calls "channelopsins": proteins from algae that can be introduced into the nervous systems of animals and humans so that light can precisely control neuron activity. He contrasts this with drug treatments, which can relieve some symptoms but often carry side effects and, in many individuals, simply do not work. He calls the technology transformative and says it was recently used to let a fully blind adult see light for the first time. These are Huberman's claims in his introduction; the excerpt does not go into detail on them.

He also previews the conversation's wider scope. It covers Deisseroth's book "Projections, a Story of Human Emotions," which uses stories of interactions with patients to explain how the brain works in health and disease. It also touches on psychedelic treatments, consciousness, and how Deisseroth organizes his thinking while running a clinical practice and a 40-plus-person laboratory and raising five children. Huberman notes that the podcast is separate from his Stanford teaching and research roles.

Asked how neurology differs from psychiatry, Deisseroth starts with a personal note: he is a psychiatrist married to a neurologist. He observes that many neuroscientists and clinicians think the two fields should eventually merge again, since they began together. The difference he draws is about measurement. Psychiatry deals with disorders where nothing visibly physical is wrong. No blood test or brain scan diagnoses schizophrenia or depression in an individual patient.

Psychiatry is much more mysterious, and the only tools we have are words.

Neurologists, by contrast, can see a stroke on a brain scan or seizure activity on an EEG, and they treat based on those measurables. Psychiatry does have rating scales for depression and autism, but Deisseroth stresses that these are "words still." He finds it odd that the most complex object in the universe has to be approached through language alone.

Huberman asks whether talkative patients are easier to diagnose. Deisseroth agrees that, in principle, quieter patients are harder to assess, but adds that reduced speech can itself be a symptom: in depression, in the negative symptoms of schizophrenia, and in autism. Some patients are so depressed they cannot speak. That makes it hard to tell depression apart from other reasons someone might be silent, which he describes as part of "the art and the science of psychiatry."

On comorbidity, his main example is Parkinson's disease. It impairs moving, walking and swallowing, and severe depression is extremely common in it. He argues the depression is not just a reaction to a life-threatening illness, because some other serious neurological disorders, such as ALS, do not show strong comorbid depression. In Parkinson's, dopamine neurons in the midbrain die, and he suggests this loss "probably" underlies both the movement disorder and the depression. In his clinic he has treated the depression while a neurologist treated the movement symptoms, working as a team.

Deisseroth expects quantitative tests for psychiatric conditions eventually. He points to current work using external EEG to look at brain rhythms and the ratios between frequencies. He is careful about how far this has come: it is "not as good as it could be" and does not yet give the confidence needed for an individual patient. His reasoning is that psychiatric disease is physical, arising from circuits, connections and projections that are not working typically, so measurables should be possible in principle.

He thinks such tests would be good overall, because drawing out symptoms precisely through interviews takes time and skill. He accepts they could be abused or misused, but says that is true of all of medicine.

He names stigma as the biggest challenge, while noting that progress is being made. Patients often feel they should handle things on their own and don't come in, which can delay treatment and let symptoms get worse. His example is that going a year or more with a serious untreated anxiety issue can convert into depression on top of the anxiety. He grants that people can sometimes master problems themselves, but says "usually some help is a good thing."

Huberman recalls a line Deisseroth used in a lecture years earlier: "We don't know how other people feel. Most of the time we don't even really know how we feel." He asks how a psychiatrist works with emotional vocabulary that offers no objective way of comparing experiences. Deisseroth describes a tension between clinical terms and how people use the same words in everyday speech. When a patient says "I'm depressed," the first job is to work out what they mean.

His method is to swap jargon for concrete questions. How much does the patient look forward to the future? How much hope do they have? How much are they planning ahead? A reply like "I don't see how I'm going to get to tomorrow" is, in his words, tragic but precise: it points to the hopelessness symptom of depression.

He is open about the limits. Whether he is studying fish, mice or rats in the lab, talking with a friend, or seeing a patient, he gets words, behaviors and actions as feedback, but no direct access. He suggests psychiatrists may be more introspective than average, which partly explains why they choose the specialty, but that this "doesn't mean we have answers."

I never really know what's going on inside the mind of the other person.

Huberman suggests that reports about physical behavior, such as poor sleep or not eating, may be more checkable than internal narratives, and that devices could in principle verify them. Deisseroth agrees. He notes that sleep and eating are already diagnostic criteria, the "vegetative signs" of depression.

The complication he raises is that these signs go in different directions for different people. Some depressed people sleep more and some sleep less. Some are physically agitated and some move less. So a single snapshot is not enough; you need a baseline to see change. Collecting baseline data from healthy people raises ethical issues he says have not been solved. Phones and accelerometers could gather a lot of it, but that would have to be handled very carefully for privacy reasons.

Huberman mentions efforts he heard about from Sam Golden at the University of Washington: devices that might let people notice they are heading toward a depressive episode, through signals such as voice inflection or how fast they get up from a chair, and then decide for themselves whether to tell a clinician. Deisseroth likes this because it gives patients agency. He notes that family members often notice changes before the patient does.

His example of a sign patients notice but may not link to depression is early morning awakening. As people slide into depression they wake earlier and earlier, perhaps at 5:00 a.m. and then 4:00 a.m., and cannot fall back asleep. Put together with other information, he says, this could be a useful warning that helps someone seek treatment.

Asked whether any psychiatric treatments reliably work, Deisseroth says yes. Despite the field's deep mysteries, he suggests many psychiatric treatments may do better than some other specialties at actually producing therapeutic benefit. He says both medication and talk therapy have been shown to be extremely effective in many cases.

His first example is cognitive behavioral therapy for panic disorder. It helps people recognize the early cognitions that lead toward a panic attack and trains them to derail that sequence. For a motivated, insightful patient, he says, a fairly standardized course of six to 12 sessions, or even fewer, can have a powerful effect. He emphasizes that this is done "just with words."

He also points to antipsychotic medications, which he says can clear up the positive symptoms of schizophrenia, such as auditory hallucinations and paranoia.

Antipsychotic medications, they have side effects, but boy do they work

At Huberman's prompting, he clarifies the terms. Positive symptoms are additions, like hallucinations or paranoia. Negative symptoms are subtractions: withdrawal, and "thought-blocking," where a person cannot move forward through a sequence of thoughts. Both can be part of schizophrenia. He says the positive symptoms are currently treated more effectively than the negative ones.