Tweet

Will Deleting Personality Disorders in DSM-5 Improve Diagnosis and Reduce Comorbidity?

Associate Professor of Psychiatry and Human Behavior, Brown University, RI; Director of Outpatient Psychiatry, Rhode Island Hospital

First published in Psychiatry Weekly, Volume 7, Issue 7, April 9, 2012

This interview was conducted on March 13, 2012 by Lonnie Stoltzfoos

Introduction

The Personality and Personality Disorders Work Group to revise DSM-5 has said that the rate of comorbidity in personality disorders is too high. To reduce comorbidity in the future, the Work Group proposed, in 2010, eliminating 5 of the 10 personality disorders from the DSM-5, specifically dependent, histrionic, narcissistic, paranoid, and schizoid personality disorders (narcissistic personality disorder has since been taken off the list for potential exclusion).

There are several problems with this plan, according to Dr. Mark Zimmerman.

“First of all, the studies the Work Group cites demonstrating a so-called problem with comorbidity all came from patient populations,” he explains. “There is, in fact, a significant rate of comorbidity among the disorders in patient populations, which aligns with the well-recognized phenomenon that comorbidity may be associated with treatment-seeking. If you look in general population samples, however, the comorbidity rate is much lower than in patient samples.

“Secondly, the proposal to delete these disorders is offered in the absence of any data,” continues Dr. Zimmerman. “We felt it was appropriate to examine the impact of such a change and whether it would have the desired effects or whether there might be unintended effects.”

Comorbidity in the General Population

Prompted by the first proposed revision to DSM-5, in which 5 disorders were recommended for deletion, Zimmerman and colleagues¹ analyzed that proposition’s effect on comorbidity, using data from a large sample of psychiatric outpatients, from the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project. This study assessed 2,150 psychiatric outpatients with the Structured Interview for DSM-IV personality disorders (SIDP-IV), administered by highly trained raters, over a period of 10 years. Most subjects were white, female, married or single, with some college education. The most common DSM-IV Axis I psychiatric diagnoses were major depressive disorder, generalized anxiety disorder, and social phobia.

Among those diagnosed with one of the 10 DSM-IV personality disorders (28.6%, n=614), approximately 30% met the diagnostic criteria for ≥2 personality disorders. When the 5 originally proposed deletions were excluded from the analysis, the incidence rate fell by 3% and the comorbidity rate dropped to 21%. “Whether that is an acceptable level of comorbidity is uncertain,” says Dr. Zimmerman, “because the Work Group proposing these changes never indicated an acceptable level of comorbidity.”

Those who met the criteria for a “deleted disorder” had considerably greater psychosocial morbidity, compared to those with no personality disorder, which suggests that deleting disorders would raise the incidence of false-negative diagnoses. Measures of psychosocial morbidity in this study included lifetime incidence of psychiatric hospitalizations, number of suicide attempts, number of axis I disorders, level of social functioning, time missed from work due to psychopathology, and Global Assessment of Functioning ratings.

“Aside from the empirical findings, there is a qualitative aspect to all this,” says Dr. Zimmerman. “How, for example, do you determine which disorders you retain and which to exclude? The Work Group certainly did not recommend for exclusion those disorders most likely to co-occur with other disorders. And if the goal was to retain those disorders that are only the most psychosocially impairing, once again that didn’t seem to be the case.”

Conclusion

Personality disorders can be challenging to treat, and evidence suggests that clinicians can harbor negative attitudes toward some patients with these disorders, such as those with borderline personality disorder.² Dr. Zimmerman does not, however, believe that a potential field bias against personality disorders in general is responsible for the Work Group’s apparent value judgments on which disorders to exclude and why.

“I think this is a Work Group that has received a charge and is doing the best it can with that charge,” he says. “There is a determined effort to incorporate dimension ratings into DSM-5, and, because there is so much research in abnormal and normal psychology in the nature of personality, the personality disorders may be the most logical place to begin.

“However, before changes are incorporated into new editions of the diagnostic manual there should be solid, replicated, research demonstrating that the new approach is superior in some way to the prior approach,” he continues. “The DSM-5 personality disorders Work Group has not demonstrated that they have developed a superior approach toward diagnosing personality disorders, and it is my understanding that no large-scale studies are planned prior to the publication of DSM-5 that will compare DSM-IV to DSM-5. Thus, changes may be incorporated into DSM-5 without an adequate scientific basis.”

Disclosure: Dr. Zimmerman reports no affiliations with, or financial interests in, any organization that may pose a conflict of interest.

References:

1. Zimmerman M, Chelminski I, Young D, Dalrymple K, Martinez J. Impact of deleting 5 DSM-IV personality disorders on prevalence, comorbidity, and the association between personality disorder pathology and psychosocial morbidity. J Clin Psychiatry. 2012;73:202-207.

2. Black DW, Pfohl B, Blum N, et al. Attitudes toward borderline personality disorder: a survey of 706 mental health clinicians. CNS Spectr. 2011 Mar 1. [Epub ahead of print]

 Tweet