Nature | Editorial

Across the divide

Diagnostic boundaries separating mental disorders hamper effective treatments.

Article tools

Scientists who attended the 2009 Winter Workshop in Psychoses in Barcelona, Spain, may not have realized it at the time, but they were part of a revolution. In previous years, organizers named the event the Winter Workshop on Schizophrenia and Bipolar Disorders. It was one of the few conferences at which those who studied schizophrenia and those who worked on bipolar illnesses would meet.

As Nick Craddock, a psychiatrist who studies both conditions at Cardiff University, UK, says in a News Feature on page 416, a merger of these two distinct groups — even in semantic terms — would have been unthinkable until very recently. Psychiatrists diagnose schizophrenia and bipolar disorder as two separate conditions. This separation is respected by drug companies, regulators, research funders, journals and bench researchers. Add that lot up, and you get a fundamental problem with psychiatry.

Next month, the American Psychiatric Association will release the long-awaited fifth version of its Diagnostic and Statistical Manual of Mental Disorders (DSM-5), which lists mental illnesses and their symptoms. Work on preparing the DSM-5 has been clouded in controversy, and the arguments over which conditions should have been included and which left out will rumble on for some time.

“Patients’ illnesses cannot be broken down into discrete groups in the way that is taught at medical school.”

The more fundamental problem, as the News Feature explores, is growing doubt about the way the DSM-5 classifies mental disorders. Psychiatrists have long known that the illnesses of patients they see in the clinic cannot be broken down into discrete groups in the way that is taught at medical school. Symptoms overlap and flow across diagnostic boundaries. Patients can show the signs of two or three disorders at the same time. Treatments are inconsistent. Outcomes are unpredictable.

Science was supposed to come to the rescue. Genetics and neuro­imaging studies would, all involved hoped, reveal biological signatures unique to each disorder, which could be used to provide consistent and reliable diagnoses. Instead, it seems the opposite is true. The more scientists look for biomarkers for specific mental disorders, the harder the task becomes. Scans of the DNA and brain function of patients show the same stubborn refusal to group by disease type. Genetic risk factors and dysfunction in brain regions are shared across disorders. Psychiatrists joke that their patients have not read the textbooks. The reality is serious and more troubling — the textbook is wrong.

The American Psychiatric Association routinely points out that its DSM disease categories are intended only as diagnostic tools. It does not claim that they mark genuine biological boundaries. But the system is set up as if they do. That might explain why biomarkers and new drugs for mental illness remain elusive. The system should change. Funders and journals must encourage work that cuts across the boundaries. Researchers should be encouraged to investigate the causes of mental illness from the bottom up, as the US National Institute of Mental Health is doing. The brain is complicated enough. Why investigate its problems with one hand tied behind our backs?

Journal name:
Date published:

For the best commenting experience, please login or register as a user and agree to our Community Guidelines. You will be re-directed back to this page where you will see comments updating in real-time and have the ability to recommend comments to other users.

Comments for this thread are now closed.


Comments Subscribe to comments

There are currently no comments.

sign up to Nature briefing

What matters in science — and why — free in your inbox every weekday.

Sign up



Nature Podcast

Our award-winning show features highlights from the week's edition of Nature, interviews with the people behind the science, and in-depth commentary and analysis from journalists around the world.