Skip to main content

To Diagnose Mental Illness, Read the Brain

Rather than relying on symptoms, scientists are developing a “brain circuits first” approach to mental health. 

Although scientists have learned a lot about the brain in the last few decades, approaches to treating mental illnesses have not kept up. As neuroscientists learn more about brain circuits, Stanford psychiatrist Amit Etkin foresees a time when diagnoses will be based on brain scans rather than symptoms. Etkin, who will be speaking at the World Economic Forum’s Annual Meeting of the New Champions in Tianjin, China, from June 26 to 28, spoke with Scientific American about his research on the neurological basis of emotional disorders and the future of mental health treatment.

[An edited transcript of the interview follows.]

The high cost of treating mental illness doesn’t get talked about very much. Why is that? It’s a really interesting issue. The costs associated with mental illness are not just the care of people who have an illness, which often starts early in life and continues as a lifelong process, but also the cost to employers in decreased productivity and the cost to society in general. A report that came out recently in Health Affairs showed that spending within our health system in the U.S. is greater for mental illness than for any other area of medicine, and yet our understanding of these illnesses is incredibly backwards. Treatments are no different than they were 40 years ago, so that feels like a problem that is only getting bigger without an obvious solution.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


Why hasn’t there been much progress? It was really not until about 10 years ago that [mental health professionals] started realizing how little difference we have made. There are a few fundamental issues and mistakes we’ve made. One is that in the absence of knowing what the causes of the illnesses that we treat are, we focus on the symptoms, and that has already led us down the wrong path. If you go to another country and you ask somebody to tell you their symptoms, as a clinician you might have the sense that they have anxiety or depression. In Asian countries they express that in a somatic way: “I can’t sleep” or “I feel weak.” The biology cannot be that different, but the symptoms are different because they’re culturally bound. If you look at different parts of the U.S. you’ll see people expressing symptoms in different ways depending on their local culture. If that’s the case, then a symptom-based definition is problematic. The long and short of it is that people have named syndromes or disorders that they don’t actually know represent a valid entity that is distinct from another entity.

What do you see as the path forward? How do we rectify those mistakes? Realizing these errors has coincided with the era of imaging, and even more recently with the really exciting focus on individual subject analyses: is there  something about this particular person’s brain that allows me to predict something? I call this the circuits first approach. We understand behavior is essentially underpinned by brain circuits. That is, there are circuits in the brain that determine certain types of behaviors and certain types of thoughts and feelings. That’s probably the most useful way of organizing brain function. If you can start characterizing circuit disruptions for compensatory symptoms at an individual subject level and then link that to how you can provide interventions, then you can get away completely from diagnoses and can intervene with brain function in a directed way.

How close do you think we are to making diagnoses based on something more tangible than symptoms? In the selected pockets where we have the best data, within the next five to 10 years. And that’s really more a factor of how much we need to show to get FDA approval and get into a commercialized product that people can use. Getting it out of the lab, in other words.

In the lab you’re using transcranial magnetic stimulation (TMS), but is the consumer product ultimately a pill? That’s a really interesting question. There’s been this assumption that medications are either preferred or maybe the best way to go about treating things, and that comes a bit from the history of psychiatry but also from the rest of medicine. I’m not sure that a pill is necessarily going to be the best approach for psychiatry. Washing the brain in a drug that affects many parts of the brain, and also affects many parts of the body, is a pretty crude and nonspecific way to affect a very discrete part of the brain. In contrast, as our neurostimulation approaches have proved, we can have a lot more specificity for our target. And if we set as our own bar that the treatment has to work within days or a week, then we can actually achieve something powerful and quick to the point where I’m not sure there would be any reason to take medication in many contexts if you can achieve it with stimulation. We’re not there yet for sure, but I can definitely see it evolving that way and that’s what we in my lab are very excited about.

Have you gotten any feedback from people about potentially doing treatments where you’re directly stimulating parts of the brain? Are people comfortable with it? We had an individual with chronic PTSD who came to our study where our goal was to understand how psychotherapy works for people with PTSD. With PTSD, really the only effective treatment is psychotherapy. There are some medications but they don’t work all that well. Part of the study was simply mapping the brain with TMS while measuring brain activity with fMRI. Nothing therapeutic at all. Psychotherapy is a very effective but very challenging therapy. It involves talking about your trauma and regaining control over your experience. He said that if there were any way he could have TMS instead he would strongly prefer that. Having to go through a difficult therapy where you’re trying to gain control over a traumatic experience is not appealing to people even if you tell them that it’s a very effective intervention. So having something that is less emotionally laden and more rational in its approach I think is very appealing to people.

There’s been a backlash against taking medication and the scattershot way that we approach things, and that’s leading people to look at other interventions that don’t have much evidence of efficacy behind them. That’s mostly out of frustration and desperation. Patients wants to get better faster so they can go back to their lives and their families and if you offer them an effective alternative I don’t see why that wouldn’t be very popular. In psychiatry in particular it can be a real comfort because everything else can seem so subjective and having something that has a number, that has an objective test, helps to validate the experience of the person. The person suffering from a mental illness knows they’re suffering, but they can’t say, “my level is 17 of this thing in my blood,” and that hugely effects how society views mental illness and even how patients themselves view their illnesses.

This interview was produced in conjunction with the World Economic Forum.