Jason Schnittker Ph.D.

The Diagnostic System

Controversies in the Classification of Psychiatric Disorders

To write better criteria, we should first ask, what is mental illness?

Posted Nov 17, 2017

After significant debate, the fifth edition of The Diagnostic and Statistical Manual of the American Psychiatric Association (DSM-5) was published in 2013. Even before it was published it was met with considerable controversy and now many regard it as a missed opportunity. Although specific diagnostic criteria were updated – and even small changes can be quite consequential – the manual was not the revolution many were expecting. At the very least the form of DSM-5 looks very much like the form of DSM-IV-TR, published more than 10 years earlier. 

Why does the classification of psychiatric disorders remain so controversial? Why is there so much inertia to the diagnostic framework psychiatry put in place with DSM-III? There is no doubt that our understanding of psychiatric disorders has improved enormously over the last thirty years, perhaps especially in the last 10 years. For this reason, the authors of DSM-5 were in a much better position to employ detailed and cutting-edge science than were those who wrote previous editions. We know much more, for instance, about the genetic determinants of psychiatric disorders.  We also know more about how psychiatric disorders appear in the brain. In short, we have much more information at our disposal. Yet the DSM continues to rely more or less on the model it adopted in 1980. It still contains lists of symptoms and rules on how to combine them in order to determine whether someone does or does not have a disorder.

I think we need to take a step back. Rather than advancing an entirely new approach or even endlessly critiquing the DSM, I think we need to understand why it might be so difficult to craft diagnostic criteria that are fully satisfying to everyone. 

Understanding these controversies requires an appreciation for how many different audiences are served by the DSM and, furthermore, how different their needs are. The DSM is used by clinicians, certainly, but it is also used by scientists and the public too. These different audiences are not pursuing exactly the same thing when they flip through the pages of the manual. Crafting diagnostic criteria that might better serve one group often diminishes the value of those criteria for another. 

The DSM is fundamentally useful, as has been pointed out many times and, indeed, forms the core argument for those who defend the DSM. It’s not a bad argument. But the idea of usefulness begs additional questions that are rarely asked: if the DSM is useful, for whom is it useful, why, and in what ways? It is here, I think, that the different users of the DSM begin to depart. Scientists, for instance, now appreciate that many psychiatric disorders reflect dimensions of psychological functioning rather than clear categories of disorder. It is very difficult, however, for clinicians to use spectrums rather than categories. Patients, too, want to know whether they are sick or well, not the degree to which they are more or less anxious than someone else. Similarly, there is evidence that clinicians in their everyday practice reduce the DSM to a handful of diagnoses. 

No matter how well this sort of shorthand serves clinicians, it doesn’t serve science especially well. Science benefits from “more data” and from more granular assessments of individuals. In addition, science can, with great insight, pursue aspects of behavior that are not directly related to illness per se. Scientists have effectively studied fear circuits in the brain, for instance. Yet these circuits are not necessarily related to the conscious experience of fear, even though such experiences are probably what most concerns patients.        

There is growing recognition of these differences, which has been made clear by the efforts to provide different audiences with their own special tools. Science, for instance, is moving in the direction of studying elementary dimensions of psychological and behavioral functioning, led by the development of Research Domain Criteria (RDoC) under the auspices of the National Institute of Mental Health. RDoC is explicitly regarded as a scientific tool and not as a tool for diagnosis (least not yet). There is much to like about the RDoC, though it does serve to separate, at least for the moment, what clinicians diagnose from what scientists study.    

At this point, it is possible to write diagnostic criteria that maximize certain aims. If we want psychiatric disorders that correspond directly to dysfunctions in the brain, we can do that. If we want to focus on psychiatric disorders that have especially strong and well-established genetic components, we can do that too. But I argue that we also need to take a step back and think at a conceptual level about mental illness, in ways that touch upon matters philosophy as much as science. What, exactly, do we think mental illness is? What aims do we hope to achieve, independent of how much science we might have to support them? 

The authors of DSM-5 recognized this issue and made a serious effort to think about the conceptual foundation of psychiatric disorders. But discussions of this sort are uncomfortable, especially at a time when we are optimistic that science might soon be able to provide the answers for us and reveal what mental illness really is. Yet if the recent history of psychiatric classification teaches us anything it’s that we’re always wrangling over more elementary concerns, even if we’re not able to say so explicitly. It is important to think not only about what mental illness is but also what we think it ought to be. Without addressing these questions, as uncomfortable as they may be, controversy is inevitable.