The Future of Psychiatric Treatment

Is it time to update diagnostic criteria for psychiatric disorders?

Posted May 19, 2021 

By: Eugene Rubin MD, PhD and Charles Zorumski MD


  • The current diagnostic framework is based on observable symptoms and not on physiologic mechanisms.
  • Recent work utilizing statistical modeling suggests that psychiatric disorders may be better categorized into broader groups.
  • Such broad groups of disorders may better align with illness mechanisms and treatment specificity.

Antidepressants, in addition to treating symptoms of depression, are also effective anti-anxiety agents. Drugs to treat schizophrenia are also helpful in treating individuals with bipolar disorder. Why might this be so?

First, some background: The current edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) provides the diagnostic criteria used by most mental health professionals. The diagnostic methods in the DSM-5 were introduced in the third edition, in 1980. The diagnostic approach outlined in DSM-III derived from a paper published in 1972 by John Feighner and colleagues from the Department of Psychiatry at Washington University. Definitions of illnesses were based on groups of symptoms that tend to occur together. Such illnesses often occur in family members, providing support for the idea that genetic influences underlie many of these disorders. A diagnosis tends to be stable over time. In the Feighner et al. diagnostic framework, illnesses were not categorized based on underlying physiological mechanisms, but rather by clinical phenotype, i.e., observable characteristics. Nevertheless, these definitions of psychiatric illnesses were important because they allowed researchers and clinicians to reliably define patient populations. In other words, different researchers and clinicians could agree on a particular patient’s diagnosis, a problem that had plagued prior diagnostic systems.

This phenotype-based approach to diagnosing mental disorders has been tremendously valuable not only in facilitating research but also in developing new treatments. Much has been learned about illnesses such as depressive disorders, anxiety disorders, schizophrenia, and bipolar disorder even though the definitions of these disorders have been based on phenotype (symptoms) and not on underlying disease mechanisms.

Over time, weaknesses in this diagnostic approach have become increasingly evident. For example, one person might fulfill clinical criteria for depression but have symptoms that overlap only slightly with the symptoms of another person who also fulfills criteria for depression. Utilizing DSM-5 criteria, many patients are diagnosed with multiple different psychiatric illnesses simultaneously (called co-morbidities). Genetic research has demonstrated that psychiatric illnesses like schizophrenia and bipolar disorder may share a large number of genes that predispose to both illnesses. In fact, some evidence indicates that there is genetic sharing across many if not most psychiatric disorders. Finally, as already mentioned, medications to treat one type of psychiatric illness are often helpful in the treatment of another category of illness. All of this suggests that the approach currently used in diagnosing psychiatric illnesses needs updating.

New research is examining ways of categorizing psychiatric symptoms that are compatible with recent biologic and genetic data. One such approach builds hierarchies of categories – the Hierarchical Taxonomy of Psychopathology (HiTOP). As comprehensively reviewed by Roman Kotov and colleagues in 2021, HiTOP is based on comprehensive statistical analyses to determine which psychiatric symptoms co-occur. These patterns then were used to construct broad categories of disorders that subsume several current diagnostic illnesses. For example, one broad category labelled “internalizing disorders” includes depressive disorders and several types of anxiety disorders. Another category, labelled “thought disorders,” includes symptoms of both schizophrenia and bipolar disorder.

The symptoms of disorders within a hierarchical category may share physiologic underpinnings, meaning that dysfunction in the same brain circuits may underlie these behaviors. Thus, it is possible that treatments such as antidepressants influence physiologic processes related to a range of internalizing disorders. If antidepressants help correct the physiological abnormality related to this broad category, it is predictable that currently defined disorders such as depression and anxiety disorders would respond to this treatment. Similarly, if antipsychotics influence physiologic processes involved in the thought disorder category of symptoms, then it is understandable that this class of drugs would help treat individuals suffering from either currently defined schizophrenia or bipolar disorder.

The Research Domain Criteria (RDoC) is another approach to investigating psychiatric disorders. It attempts to integrate biologic processes that underlie normal and abnormal human behavior. Thus, HiTOP targets diagnostic issues and RDoC biologic issues. Eventually, there will be a convergence of these two approaches.

Current neuroscientific research is helping redefine diagnostic categories into groups that have similar underlying physiological abnormalities, and these physiological abnormalities could help identify new treatments that cut across current categories of psychiatric illnesses (called trans-diagnostic approaches). As this research progresses, it is possible that treatments will become increasingly specific, but that specificity may not be defined by current diagnostic categories. The more we understand the physiologic bases of psychiatric disorders, the more likely it is that new and better treatments will be discovered.