AUDIE CORNISH, HOST:
From NPR News, this is ALL THINGS CONSIDERED. I'm Audie Cornish.
MELISSA BLOCK, HOST:
And I'm Melissa Block.
In San Francisco this weekend, the American Psychiatric Association is scheduled to release an updated version of the DSM, that's the Diagnostic and Statistical Manual of Mental Disorders. It's a huge tome that helps mental health professionals diagnose patients with problems such as depression, anxiety and schizophrenia. The last DSM revision came 20 years ago, and in that time, scientists have learned a lot about the brain. Still, the new manual is remarkably similar to the old one.
NPR's Jon Hamilton wanted to know why.
JON HAMILTON, BYLINE: If you go to a hospital with chest pain, chances are you'll get a diagnosis based on tests. Stuff like electrical activity in your heart or enzymes in your blood, blood flow through your arteries. If your problem is emotional pain, though, your diagnosis will be based on a conversation.
And Michael First, a psychiatrist at Columbia University, says the new version of the DSM won't change that.
DR. MICHAEL FIRST: There's lots of changes throughout the manual that reflect the research in the last 20 years. But because that kind of research hasn't allowed for a paradigm shift, the DSM is not a paradigm shift either.
HAMILTON: I'm talking to Michael First because I want to understand why psychiatry and psychology aren't more like cardiology. First ought to know. He's been a central figure in updating the DSM for decades. He tells me that when he and his colleagues began work on the DSM, they didn't expect it to turn out like this.
FIRST: We were hoping and imagining that research would advance at a pace that laboratory tests would have come out. And here we are, 20 years later, and we still, unfortunately, rely primarily on symptoms to make our diagnoses.
HAMILTON: And First says that's not ideal - not for doctors, not for patients, not for scientists. The problem is that the new DSM is still classifying mental disorders based mostly on their surface appearance, not their underlying biology. And the history of science shows that appearances can be deceiving.
Take hippos. At first, scientists thought hippos must be related to pigs. After all, they look a bit alike and have similar teeth. But fossils and genetic studies showed that the closest living relatives of hippos are dolphins and whales. So to avoid the hippo problem, many areas of medicine have begun to look beyond the superficial appearance of an illness. I wanted to understand how you do that.
HAMILTON: So I talked to Sue Desmond-Hellman. She's the chancellor of the University of California, San Francisco. She's also an oncologist.
DR. SUE DESMOND-HELLMAN: For literally centuries, doctors have looked at diseases using signs and symptoms: Is your blood sugar high? Do you have a lump somewhere? And that signs and symptom-driven disease definition or taxonomy could be seen as holding us back.
HAMILTON: The hippo problem is still around, and that's one reason Desmond-Hellman has become an advocate for something called "precision medicine," which tries to classify diseases in a way that reflects their underlying cause. She tells me one good example is breast cancer.
DESMOND-HELLMAN: I'm a cancer doctor and had the incredible opportunity to work on revolutionizing how we treat breast cancer, based on what gene - what's in your DNA that's driving, that's signaling the breast cancer to grow. And it's a wonderful opportunity because if a patient has a breast cancer that's driven by something in the genome, we can turn it off.
HAMILTON: Desmond-Hellman says this revolution happened in part because cancer researchers looked beyond the old way of classifying tumors, simply by what part of the body they appeared in. And she says something similar needs to happen in mental health.
DESMOND-HELLMAN: Everything from autism to Alzheimer's is classified in ways that clearly don't work today and clearly don't give the folks who are caring for these people - the families or clinicians or researchers - the information we need to improve mental health.
HAMILTON: Desmond-Hellman adds that she's not criticizing psychiatrists or psychologists. She says they have no choice but to rely on the DSM because there still aren't genetic tests or brain scans that offer a better way to classify patients.
But why isn't there a better way? I directed that question to Tom Insel, who directs the National Institute of Mental Health. We chatted about the future in a place that embodies the past.
DR. THOMAS INSEL: This is the original building on the NIH campus from the 1930s, a place where Franklin Roosevelt would come on weekends.
HAMILTON: Insel tells me that mental health researchers have spent decades searching for the sort of lab tests that have become standard in other areas of medicine.
INSEL: We've tried. You know, we've actually looked to using brain imaging, using various endocrine tests, looking at a range of other kinds of biomarkers. So far, that has been found wanting.
HAMILTON: For example, they have tried to find genes or a brain scan that will identify patients who have depression. Insel tells me that might be possible if all the people with depression have the same underlying problem. But what if they don't?
INSEL: It may be that what we're seeing when we've looked at somebody, for instance, with depression, is a little bit like looking at somebody who has a fever. Some people who have a fever have a bacterial infection, some a viral infection, some an endocrine problem - a whole range of reasons why that would be your presenting symptom, and a whole range of different treatments that you need for each of those causes.
HAMILTON: So a few years ago, the NIMH began moving away from research based on categories defined only by signs and symptoms. Instead, it's pushing an approach that emphasizes basic functions in the brain. Researchers might get money to study circuits involved in fear, or so-called working memory.
Insel says this sort of research has the potential to uncover problems in the brain that will change the way mental disorders are classified.
INSEL: A biological thing that presents with depression in some people might present with psychosis or anxiety in others. And so, maybe what you'll find is a problem that cuts across the current diagnostic categories.
HAMILTON: But Insel says findings like that are years off. Maybe they'll get there in time for the next version of the DSM, maybe.
Michael First, the psychiatrist who has spent much of his career working on the DSM, tells me he's learned to be cautious about anticipating major advances in any problem that affects the brain - like Alzheimer's. First tells me that back in the 1990s, he helped write a sort of guidebook to DSM-IV.
FIRST: And in that book, I stuck my neck out. We said that by the time DSM-5 comes out, Alzheimer's - we're certain - will be the first diagnosis that has a laboratory test. Actually, you know, sort of laid it out there. And I'm wrong.
HAMILTON: Even though scientists have learned a huge amount about the plaques and tangles associated with Alzheimer's, even though high-tech scans can reveal ever more subtle changes in the brains of people with the disease, there's still no lab test that's good enough to diagnose Alzheimer's. So doctors still rely on old-fashioned tests of short-term memory and thinking.
First says migraine headaches are another brain problem still waiting for a lab test.
FIRST: If you have a headache and you go to a neurologist, even though they might run some blood tests, the actual diagnosis of a migraine versus a cluster headache depends upon the description of the symptoms. So psychiatry is not unique.
HAMILTON: In fact, First says, psychiatrists have hit many of the same frustrating roadblocks encountered by other specialists who deal with the brain.
FIRST: A lot of it really has to do with the brain is very, very complicated, and it really hasn't yielded its secrets yet.
HAMILTON: So there it is. I wanted to know why mental health isn't more like cardiology or oncology. And the answer is something scientists tell me all the time: The brain is the most complicated thing in the universe; nearly 100 billion neurons, trillions and trillions of connections - even the human genome doesn't come close.
But First says all this isn't a reason to give up on the DSM or the ability of mental health professionals to treat mental disorders.
FIRST: The fact that we don't know exactly how they work doesn't make it so that we can't actually help a lot of people. When people walk into our offices for help, they don't walk into our offices for some explanation of the neurobiology of what's going on. They want some relief of suffering, and the DSM remains the most valuable tool for psychiatrists to be able to do that.
HAMILTON: First says he still believes that lab tests will eventually be a part of mental health care. But he's now less certain that it will happen in his lifetime.
Jon Hamilton, NPR News.