May 2013 may be remembered as a watershed (or maybe a Waterloo) in the history of psychiatry. Two major events have set the stage for a fundamental debate about how we should think about the nature of mental illness. The American Psychiatric Association (APA) is about to publish the fifth edition of its diagnostic system of classification: DSM-5. And, three weeks before the publication, Thomas Insel, director of the National Institute of Mental Health (NIMH), announced that his agency will be moving away from funding studies based on the DSM categories. The goal will be to build a new system of classifying psychiatric disorders based on a basic understanding of how genetics, neurobiology and cognitive functions shape the brain and mind. As Insel put it, "patients with mental disorders deserve better" and "we cannot succeed if we use DSM categories as the 'gold standard.'" Days later, the chair of the DSM-5 Task Force wrote that a new system based on neuroscience is so far "a promissory note" and that the DSM remains essential for the diagnosis of individuals who are suffering in the here and now. "Our patients deserve no less, " he concluded.
This collision of visions didn't happen overnight. The modern system of psychiatric classification emerged with the publication of DSM-III in 1980 -- an attempt to address the Wild West nature of psychiatric diagnosis. In the 1970s, influential studies documented that diagnosis was highly operator-dependent: The same patient might be diagnosed as manic depressive by one clinician, schizophrenic by another, and borderline by a third. By providing specific criteria, DSM-III gave the field a common language, and the DSM remains a useful resource for diagnosis. DSM-III and its successors aimed to be "atheoretical" -- that is, they described the signs and symptoms of psychiatric disorder without saying anything about their causes, whether "biochemical imbalances" or repressed Oedipal conflicts. That was a conservative and laudable approach given the lack of any clear account of how these disorders occur. But it also meant that psychiatric diagnosis lacked any grounding in a science of how the mind and the brain work. DSM-III was a giant leap forward for reliability (consistency), but the question of validity (how accurately a diagnosis captures something real) remained wide open. Over the following decades, successive iterations of the DSM tweaked these criteria, but they still represent a consensus of experts.
Meanwhile, advances in neuroscience, genetics, and psychology were beginning to fill in details about the workings of the mind and brain. The tool kit for studying these phenomena -- including brain imaging, genetic and epigenetic analysis, and experimental psychology -- has reached a level that was unimaginable in 1980.
The dialogue between the APA (which publishes and owns the DSM) and the NIMH (which is the largest funder of mental health research) is about where we go from here. Do we focus on completing the pointillist portrait that scientists have been painting, or do we continue to simply color within the lines drawn by the DSM? Insel's announcement has drawn a line in the sand to say that it's time we build an understanding of the mind and brain from the bottom up.
Indeed, a growing body of research is challenging the boundaries drawn between disorders by the DSM. In the last five years, genetic researchers have identified specific genetic variants that confer risk to psychopathology. And the results have shown that, at least at a genetic level, the lines between DSM disorders are fuzzy. One indication of this has come from studies of rare copy number variants (CNVs), a form of genetic variation in which chunks of DNA are deleted or duplicated. It turns out that the same CNV can be a cause of multiple neuropsychiatric conditions including autism, schizophrenia, intellectual disability, ADHD, and mood disorders. In March of this year, an international group of scientists found overlap in the genetic component of disorders as different as autism, ADHD, depression, bipolar disorder, and schizophrenia. Our DNA has not read the DSM.
It is also becoming clear that there are few if any bright lines between disorder and the range of normal. That's not really a surprise. Many of the conditions we recognize as psychiatric disorders are variations of mental and neural systems that evolved to adapt to the challenges our evolutionary ancestors faced: avoiding harm, forming attachments, understanding other people's intentions and feelings, choosing a mate. We have brain circuitry dedicated to these basic domains of mental functioning. But when these systems go awry, the result can be real suffering. It's easy to see how anxiety disorders can emerge from an exaggeration of our harm avoidance systems. But there are many examples. Impairments in social cognition are obvious in autism, but to lesser degrees, autism-related traits occur across the population and recent twin studies show that genes contributing to autism also shape the normal spectrum of social functioning.
The most contentious debates circling the DSM-5 process are all tied to the fundamental question of how we define the boundaries of disorder and the lines between normal and abnormal. When does the pain of grief become depression? How broad is the spectrum of autism? At what point can we say that the child with problems concentrating and sitting still has ADHD or that the child with excessive mood swings and irritability has bipolar disorder? When does preoccupation with physical symptoms cross the line into an illness ("somatic symptom disorder")? How do we avoid pathologizing the normal range of adaptations to life? The only way to really answer such questions is to take a step back and map out the basic architecture of the brain and the mind. We first have to answer questions like: What is the brain designed to do? How is it organized? How does it develop? We are not going to solve this problem by consensus. We need the evidence.
Of course, all fields of medicine have relied on clinical observation until biology gave up its secrets and allowed us to move from classification based on symptoms to one based on causes. For psychiatry, this transition has been particularly difficult because the organ of interest -- the brain -- is more complex and inaccessible than most.
The NIH is appropriately taking an active role in addressing this because it's a problem with immense public health implications. The reality is that taking on this project in a serious way will require a systematic effort to fund the research. Like it or not, science follows the money. The NIMH RDoC project (aimed at laying the scientific groundwork for a new system of diagnosis) and the NIH BRAIN Initiative are major investments in transforming brain research. Both will take a decade or more to complete and neither offers an alternative to our current DSM system today. But if we don't start mobilizing the scientific community to turn the new tools of neuroscience, genomics, and psychology to clinical use, a better alternative will remain a distant hope. Those who decry psychiatry for "pathologizing normal" should applaud the new effort to build a more accurate map of the landscape of normal and abnormal. As both a scientist and practicing psychiatrist, I see the current debate as a healthy sign: It's really about how we move the field forward while not losing sight of what we can offer our patients today. And one thing we do have consensus on is that those who suffer deserve our best.
The author is professor of psychiatry at Harvard Medical School and author of The Other Side of Normal (William Morrow, 2012)
For more by Jordan Smoller, click here.
For more on mental health, click here.