Changes in Psychiatric Diagnosis

November 1, 2008

DSM-IV-TR, our current diagnostic classification system of psychiatric disorders, follows the diagnostic paradigm first established by DSM-III in 1980.

DSM-IV-TR, our current diagnostic classification system of psychiatric disorders, follows the diagnostic paradigm first established by DSM-III in 1980. Acknowledging the field’s ignorance about the underlying etiology and pathophysiology of mental disorders, DSM-III eschewed basing the structure of the classification and definition of disorders on unproven hypotheses about the causes of psychiatric disorders in favor of an “atheoretical” approach in which disorders were defined by their presenting symptoms.1 This reliance on purely descriptive operationalized definitions resulted in the establishment of a widely accepted reliable diagnostic system that has enhanced communication among mental health care professionals and greatly facilitated psychiatric research.

Current problems

Critics of DSM-IV have justifiably been frustrated with the superficial nature of this descriptive approach. Paul McHugh,2 in a 2005 JAMA editorial, complained that “because the manual fails to identify what underlies the symptomatic expression of a condition, it cannot suggest intelligible principles relating one disorder to another or illuminate why certain of them bunch together. For these reasons, faith in the criterion method has gradually faltered.” It has always been understood that the descriptive approach adopted by DSM will eventually be replaced by a classification informed by an understanding of etiology, which would put psychiatry in line with much of the rest of medicine. Given that DSM-V is slated for publication in 2012, has the time finally arrived for an etiologically based classification, or, falling short of those aspirations, at least one that includes objective laboratory tests as part of its diagnostic definitions?

Given the enormous advances in imaging technology and molecular genetics over the past 15 years, one might assume that such an approach is finally within our reach. In fact, when Allen Frances and I wrote the DSM-IV Guidebook, given the explosion in new technologies available in the early 1990s, we were very optimistic that by the time DSM-V came out (which we envisioned would be sometime around 2007), it would include at least some biologically based diagnostic criteria. For example, in our explanation of the diagnostic criteria for dementia of the Alzheimer type, we wrote3:

"[O]f all of the criteria sets in DSM-IV, this is the one most likely tobecome quickly outdated. As of 1995, the definite diagnosis of Alzheimer disease can be made only by autopsy or brain biopsy. . . . Fortunately, research in this area is uncovering fascinating leads concerning the genetics and pathogenesis of this form of dementia. It seems likely that before DSM-V is published, diagnostic tests for Alzheimer disease will evolve that will supersede the current clinically based criteria set. This event should be a marvelous moment in the history of psychiatry, hopefully to be followed by . . . similar advances in the understanding and diagnosis of other psychiatric conditions (eg, schizophrenia, bipolar disorder)."

Unfortunately, our hopes for the development of diagnostically specific tests that could be included in the DSM-V criteria sets for Alzheimer disease have not been borne out. One of the 12 research planning conferences, convened with the goal of reviewing the scientific literature in advance of starting formal work on DSM-V, focused on diagnostic issues in dementia.4 Biomarkers (eg, cerebrospinal fluid t) have been identified that have sufficient diagnostic sensitivity and specificity to distinguish between patients with Alzheimer disease and normal healthy controls; however, these markers have not been demonstrated to be useful in differentiating Alzheimer dementia from other types of dementia. This, of course, is precisely the diagnostic scenario most likely to be faced by the psychiatrist in actual clinical settings.5

The lack of diagnostic sensitivity and specificity is even more the case for putative biomarkers in other diagnostic categories.6 The essential problem is that despite recent scientific advances, the study of the causes of mental disorders remains remarkably complex: for example, while it initially appeared possible to discover the “gene” behind a particular mental disorder, it is now clear that disorders are caused by a complex interaction of genes and the environment.7

Likely changes

It therefore seems likely that although the text of DSM-V will be enriched by the fruits of neurobiological research from the past 20 years, neurobiology and genetics will not inform the development of diagnostic criteria in DSM-V.8 Diagnostic criteria will therefore continue to be based on clinical findings. Cumulative research into the underlying pathophysiology of disorders may have an impact on the organization of the diagnostic groupings in DSM-V, however.9

Disorder groupings in DSM-IV, like diagnostic criteria, are based entirely on superficial symptoms. For example, the only thing that the disorders in the somatoform section of DSM-IV have in common is that affected patients are most likely to present in general medical settings. Several of the DSM-V research conferences focused on exploring possible opportunities to reorganize DSM-V that would reflect commonalities between disorders based on underlying pathophysiological and genetic factors rather than presenting symptoms.

For example, obsessive-compulsive disorder (OCD) is currently classified among the anxiety disorders in DSM-IV. This classification is based on the observation that patients with OCD typically present with anxiety resulting from their obsessions and compulsions. Besides this superficial similarity, however, OCD appears to have little in common with the other anxiety disorders in terms of treatment response (eg, OCD responds only to medications that affect the serotonin system whereas other anxiety disorders respond to a broader range of medications), family history, neurocircuitry, and comorbidity patterns. These dissimilarities suggest that OCD belongs in a separate diagnostic grouping.

One of the DSM-V research planning conferences, co-chaired by Drs Eric Hollander and Joseph Zohar, examined which disorders in DSM might be considered for inclusion in an obsessive-compulsive spectrum grouping, based on phenomenology, comorbidity, course of illness, treatment response, genetics, neuroimaging, and other validators.10 They proposed the following candidates for inclusion in the OC spectrum:

• Obsessive-compulsive personality disorder (OCPD)
• Hoarding
• Tic disorders
• Sydenham and other pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS)
• Trichotillomania
• Body dysmorphic disorder
• Hypochondriasis
• Autism
• Eating disorders
• Pathological gambling and other behavior addictions
• Substance dependence

Conference participants concluded that the strongest evidence for inclusion in an OC spectrum exists for tics, body dysmorphic disorder, and hypochrondriasis, with less support for hoarding, trichotillomania, PANDAS, and OCPD.

Other areas of change

Another area of likely change in DSM-V concerns current dissatisfaction with its categorical nature: that is, in DSM-IV a person either has or does not have a particular diagnosis. Categorical systems or typologies are the predominant approach used in medical classifications for several reasons. Classifying the world into categories is a fundamental characteristic of human thinking embodied in the nouns of everyday speech (eg, animals, plants, planets, chemical elements).

The yes/no nature of categorical diagnosis also serves to facilitate clinical decision making, which is typically characterized by a number of yes/no decisions (eg, whether to treat, whether to hospitalize). Moreover, it has traditionally been assumed that most medical diseases are discrete entities. However, while this assumption might be true for a few conditions (eg, Down syndrome, fragile X syndrome, Alzheimer disease, Creutzfeldt-Jakob disease), the vast majority of psychiatric disorders exist on a continuum with no discrete boundaries between disorders (eg, schizophrenia and bipolar disorder) and between disorders and normality (eg, depression).

The imposition of arbitrary diagnostic thresholds, as in DSM, leads to high rates of subthreshold and not otherwise specified diagnoses, and high rates of diagnostic comorbidity. For this reason, there has been great interest in incorporating dimensional approaches into DSM-V.11 Such approaches could improve measurement-based care and allow clinicians to provide a dimensional indication of cross-cutting symptoms, such as impulsivity or lack of insight. The challenge for DSM-V will be to figure out how to include dimensions that clinicians will find useful rather than burdensome.12-14


When DSM-V is published in 2012, it will probably be the last of these diagnostic manuals as we know them. The current model for revising DSM is to convene workgroups at certain intervals that will review the scientific literature and propose changes to the definitions of every disorder in the manual based on new findings. The obvious problem with this model is that scientific advances are made all the time-and not at the same rate for each class of disorders.

If a new finding concerning the pathophysiology of schizophrenia renders the definition obsolete, why should a new definition have to wait until the American Psychiatric Association (APA) revises the entire manual? Similarly, for disorders in which there has been little scientific research, the convening of workgroups devoted to proposing changes to the criteria would increase the likelihood that changes will be made unnecessarily.

Although this problem has long been recognized, technical constraints on how changes could be disseminated have kept the APA locked into a traditional model in which (like textbooks) the entire manual is periodically revised. Advances in electronic publishing are on the verge of freeing the APA from this model and may allow changes to be made piecemeal to DSM in response to scientific advances. It is therefore likely that future editions of DSM will be distributed as an electronic document that can be updated as needed.


1. Spitzer RL, Williams JBW, Skodol A. DSM-III: the major achievements and an overview. Am J Psychiatry. 1980;137:151-164.
2. McHugh PR. Striving for coherence: psychiatry’s efforts over classification. JAMA. 2005;293:2526-2528.
3. Frances A, First MB, Pincus H. DSM-IV Guidebook. Washington, DC: American Psychiatric Press; 1995.
4. Sunderland T, Jeste D, Baiyewu O, et al, eds. Diagnostic Issues in Dementia: Advancing the Research Agenda for DSM-V. Arlington, VA: American Psychiatric Association; 2007.
5. Sunderland T, Hampel H, Takeda M, et al. Biomarkers in the diagnosis of Alzheimer’s disease: are we ready? J Geriatr Psychiatry Neurol. 2006;19:172-179.
6. Charney D, Barlow D, Botteron K, et al. Neuroscience research agenda to guide development of a pathophysiologically based classification system. In: Kupfer D, First MB, Regier DA, eds. A Research Agenda for DSM-V. Washington, DC: American Psychiatric Association; 2002:31-84.
7. Kendler KS. Reflections on the relationship between psychiatric genetics and psychiatric nosology. Am J Psychiatry. 2006;163:1138-1146.
8. Hyman SE. Can neuroscience be integrated into the DSM-V? Nat Rev Neurosci. 2007;8:725-732.
9. Phillips KA, Price L, Greenberg B, Rasmussen S. Should the DSM diagnostic groupings be changed? In: Phillips KA, First MB, Pincus HA, eds. Advancing DSM: Dilemmas in Psychiatric Diagnosis. Washington, DC: American Psychiatric Association; 2003:57-84.
10. Hollander E, Kim S, Zohar J. OCSDs in the forthcoming DSM-V. CNS Spect. 2007;12:320-323.
11. Helzer JE, Kraemer HC, Krueger RF. The feasibility and need for dimensional psychiatric diagnoses. Psychol Med. 2006;36:1671-1680.
12. Sussman N. The developmental process for the Diagnostic and Statistical Manual of Mental Disorders, 5th ed. Prim Psychiatry. 2007;14:44-47.
13. First MB, Westen D. Classification for clinical practice: how to make ICD and DSM better able to serve clinicians. Int Rev Psychiatry. 2007;19:473-481.
14. First MB. Clinical utility: a prerequisite for the adoption of a dimensional approach in DSM. J Abnorm Psychol. 2005;114:560-564.