This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.


Commentary: The Problematic DSM-5 Personality Disorders Proposal: Options for Plan B

Michael B. First, MD

Published: October 24, 2011

See related article by Zimmerman, et al, and commentary by Paris.

The Problematic DSM-5 Personality Disorders Proposal: Options for Plan B

Widespread dissatisfaction with the DSM-IV classification of personality disorders has led to calls for a significant overhaul in the classification of personality disorders for DSM-5. Problems with the DSM-IV approach include (1) the inherent unsuitability of a categorical diagnostic model for diagnosing personality disorders; (2) the high rates of diagnostic comorbidity, especially evident in severely ill patients; (3) high rates of personality disorder not otherwise specified (NOS), especially in outpatient populations with milder illness; (4) arbitrary and non-empirically based diagnostic thresholds for making a personality disorder diagnosis; (5) questionable clinical utility; and (6) the fact that the personality disorder categories emanate from clinical tradition rather than reflecting a solid empirical base.1

To address these problems, the DSM-5 Personality and Personality Disorders Work Group has proposed a radically new approach. On February 10, 2010, the group’s first version of the proposal, a hybrid prototype/dimensional trait approach, was posted on the DSM-5 Web site.2 It involved the removal of half of the DSM-IV diagnoses, the replacement of diagnostic criteria sets with prototype matching, the inclusion of a rating for the level of social and interpersonal functioning, and the provision of a 6-domain, 37-trait dimensional model. This initial proposal was met with considerable criticism regarding its complexity and unfamiliarity,3,4 problematic clinical utility,4-6 weak empirical justification,4,7,8 deletion of certain personality disorders,7-11 and replacement of diagnostic criteria with prototype matching.8,11,12 In response to these critiques and public comments submitted to the DSM-5 Web site, a revised "simplified" version of essentially the same proposal was posted in January 201113 and was followed in July 2011 by a more substantially revised version that added back a narcissistic personality disorder type, replaced the prototype matching approach with types defined in terms of completely rewritten diagnostic criteria using trait terminology, and reduced the number of dimensional traits to 25.

Although no criticisms of this third revision have yet been published, a July 11, 2011, letter to the DSM-5 Task Force from a group of 31 of the top experts in personality disorders (including Dr Zimmerman and myself) raises a number of serious objections to the proposal, including that

it is too complicated, it is unfamiliar to the clinicians who will be expected to use it, it will aggravate (not ameliorate) the problems with clinical utility, it lacks a scientific rationale, it is an amalgam of trait psychology and the existing typology lacking both a conceptual and empirical base, the efforts to capture existing types are disconnected from what is known about these disorders, and there is no effort to integrate the evidentiary base of DSM-IV or subsequent research (or if there is, it is undocumented).

The letter concludes with the concern that this proposal "undermines psychiatry’s professional and public integrity and worse, it undermines our credibility with the patients we are dedicated to serve" (J. G. Gunderson, MD, e-mail communication, July 11, 2011).

According to the "Guidelines for Making Changes to DSM-5," "A broad consensus of expert clinical opinion would generally be expected for all proposed changes or additions."14(p5) Thus, regardless of what one thinks of the merits of each of the concerns raised in the letter, in view of the fact that 31 of the most well-respected and influential researchers, scholars, and clinicians believe that this third go-round is untenable, and, with roughly a year to go until the final deadline for approval by the APA Board of Trustees, it seems quite likely that this proposal will not be approved for inclusion in the main body of DSM-5 and instead will have to be placed in the DSM-5 research appendix to facilitate the collection of additional data to establish its superiority over the DSM-IV system in terms of validity and clinical utility.

So where does that leave the personality disorders section in DSM-5? In contrast to the DSM-IV revision process, which framed its proposals in terms of various options to be considered,15 no alternative proposals have been offered by the DSM-5 Personality and Personality Disorders Work Group, ie, there is no "Plan B." The most obvious option for Plan B would be to leave the DSM-IV personality disorder classification essentially unchanged except that the personality disorders would be coded along with the other disorders in DSM-5 (ie, they would no longer be coded on a separate axis). The article by Zimmerman et al16 in this issue of the Journal, as well as other recent publications from the same group,17,18 provides persuasive evidence that many of the criticisms of the DSM-IV approach are unfounded. The current article argues, for example, that the DSM-IV provision allowing the clinician to indicate subthreshold traits (eg, "narcissistic personality disorder with borderline and dependent features") renders it a quasi-dimensional approach with 3 levels (ie, absent, subthreshold, and threshold) and that this approach is as strongly associated with indicators of psychosocial morbidity as more finely grained (ie, 5-level) approaches, all of which are superior to a strictly categorical approach. Zimmerman and colleagues conclude that, instead of changing the current DSM-IV diagnostic approach, attempts should be made to increase clinicians’ recognition that DSM-IV already includes a valid dimensional rating.

A significant drawback of keeping the status quo is that it would fail to move the field forward in terms of addressing the many valid criticisms of the DSM-IV approach. There is general consensus in the field that a dimensional approach to the diagnosis of personality disorders is the better way to go. However, the implementation of a dimensional approach that makes the DSM too complicated and unfamiliar and that lacks perceivable clinical utility will not be acceptable to clinicians and thus not used, making any purported benefits moot.19 Therefore, another option for Plan B is to make incremental changes to the personality classification to ease the transition from the current familiar categorical approach to a more unfamiliar but superior dimensional framework. Three such changes that might pave the way for an eventual fully empirically based dimensional approach and that could be implemented in DSM-5 are (1) basing the categorical diagnosis of personality disorder on meeting the general criteria for a personality disorder rather than exceeding an arbitrary symptom count threshold; (2) providing concise definitions for mild, moderate, and severe levels of personality disorder; and (3) implementing a dimensional profile based on dimensionalized versions of the current familiar categorical constructs.

Although DSM-IV includes general criteria for the diagnosis of personality disorder based on the general diagnostic guidelines for personality disorder included in ICD-10,20 unlike ICD-10, which requires that these guidelines be met for the diagnosis of a personality disorder, the relationship between the DSM-IV general criteria and the specific diagnostic categories was never explicated. Therefore, the general criteria for personality disorder included in the DSM-5 proposal and that are explicitly required in order to make a diagnosis of personality disorder could be viewed as an incremental change from both the DSM-IV and ICD-10 general criteria and thus most likely be found acceptable by clinicians. Similarly, given that the DSM-IV classification currently provides severity specifiers (ie, mild, moderate, severe) for every disorder, a simplified version of the DSM-5 "levels of personality functioning continuum" (currently depicted in a 5-row-by-4-column table) that would provide criteria for mild, moderate, and severe subtypes of personality disorder would also most likely be found by clinicians to be both acceptable and clinically useful.

Finally, in order to help familiarize clinicians with the method of characterizing a patient’s personality traits in terms of a profile of continuous dimensions, dimensionalized versions of the DSM-IV categories could be implemented, an approach first suggested by Kass and colleagues21 in the 1980s using a 4-point scale for each dimension and further elaborated by Oldham and Skodol22 using a 6-point scale. Thus, on the basis of the item counts, the clinician would be able to note which dimensions are the most salient by indicating their relative intensity using standardized scales. For example, using a 4-point scale of severity (absent, low, medium, and high), a severely ill patient who would have been diagnosed with borderline personality disorder and narcissistic personality disorder with clinically significant but subthreshold dependent and avoidant traits in DSM-IV would be diagnosed with a single disorder, ie, severe personality disorder, with high borderline, high narcissistic, medium avoidant, and medium dependent features.

Advantages of this approach include addressing the comorbidity problem (ie, making a single diagnosis of personality disorder with a dimensional profile instead of multiple comorbid diagnoses), the NOS problem (ie, formerly "subthreshold" or "mixed" cases would be indicated by making a diagnosis of personality disorder with the mixed features indicated by the profile), and the arbitrary threshold problem (ie, basing the diagnosis of personality disorder on a clinical judgment of the presence or absence of the general criteria for personality disorder rather than by counting symptoms to determine if they exceed an arbitrary threshold). It would also have the advantage of allowing clinicians to become familiar with the concept and practice of using dimensions for the diagnosis of personality disorder while still retaining the familiarity of the current categorical personality disorder constructs. Finally, as noted by Skodol,23 the research utility of dimensionalizing the personality disorder categories has been established in a number of studies over the past 20 years, including the Children in the Community Study24 and the Collaborative Longitudinal Personality Disorders Study.25

Although this option falls far short of the ultimate goal of replacing the DSM-IV personality disorder categories, which are mostly based on clinical tradition, with a dimensional approach based on empirically derived traits, the difficulties involved in developing a dimensional system and establishing its clinical utility within the constraints of the DSM-5 development process indicate that such a radical shift in personality disorder classification requires a more extended timeline. The first step toward actualizing this goal would be to add an empirically derived trait model (or ideally, several competing trait models, one based on extreme variants of normal traits versus a model restricted to pathological traits) to the DSM-5 research appendix and conduct an intensive research program to demonstrate its clinical utility. To address the more challenging issue of making such approaches more familiar and acceptable, these dimensional trait models of personality should be introduced into graduate and professional training programs in psychiatry, psychology, and clinical social work at an early stage of the student’s training. Given that plans are afoot to make DSM-5 a "living document,"26 once the validity, clinical utility, and user acceptability of a dimensional trait approach are well established, its implementation would not have to wait until DSM-6.

Author affiliations: Department of Psychiatry, Columbia University; and Department of Clinical Phenomenology, New York State Psychiatric Institute, New York.

Potential conflicts of interest: Dr First’s department, Biometrics Research, receives royalties from American Psychiatric Publishing, Inc, for the Structured Clinical Interview for Personality Disorders (SCID-II).

Funding/support: None reported.


1. First M, Bell C, Cuthbert B, et al. Personality disorders and relational disorders: a research agenda for addressing crucial gaps in DSM. In: Kupfer D, First M, Regier D, eds. A Research Agenda for DSM-V. Washington, DC: American Psychiatric Publishing, Inc; 2002:123-200.

2. American Psychiatric Association. DSM-5 Development Web site.

3. Shedler J, Beck A, Fonagy P, et al. Personality disorders in DSM-5. Am J Psychiatry. 2010;167(9):1026-1028. PubMed doi:10.1176/appi.ajp.2010.10050746

4. Gunderson J. Commentary on "Personality traits and the classification of mental disorders: towards a more complete integration in DSM-5 and an empirical model of psychopathology." Personal Disord 2010;1(2):119-122. doi:10.1037/a0019974

5. Pilkonis PA, Hallquist MN, Morse JQ, et al. Striking the (im)proper balance between scientific advances and clinical utility: commentary on the DSM-5 proposal for personality disorders. Personal Disord. 2011;2(1):68-82. PubMed doi:10.1037/a0022226

6. Clarkin JF, Huprich SK. Do DSM-5 personality disorder proposals meet criteria for clinical utility? J Pers Disord. 2011;25(2):192-205. PubMed doi:10.1521/pedi.2011.25.2.192

7. Widiger TA. The DSM-5 dimensional model of personality disorder: rationale and empirical support. J Pers Disord. 2011;25(2):222-234. PubMed doi:10.1521/pedi.2011.25.2.222

8. Widiger TA. A shaky future for personality disorders. Personal Disord. 2011;2(1):54-67. doi:10.1037/a0021855

9. Bornstein RF. Reconceptualizing personality pathology in DSM-5: limitations in evidence for eliminating dependent personality disorder and other DSM-IV syndromes. J Pers Disord. 2011;25(2):235-247. PubMed doi:10.1521/pedi.2011.25.2.235

10. Ronningstam E. Narcissistic personality disorder in DSM-V—in support of retaining a significant diagnosis. J Pers Disord. 2011;25(2):248-259. PubMed doi:10.1521/pedi.2011.25.2.248

11. Pincus AL. Some comments on nomology, diagnostic process, and narcissistic personality disorder in the DSM-5 proposal for personality and personality disorders. Personal Disord. 2011;2(1):41-53. doi:10.1037/a0021191

12. Zimmerman M. A critique of the proposed prototype rating system for personality disorders in DSM-5. J Pers Disord. 2011;25(2):206-221. PubMed doi:10.1521/pedi.2011.25.2.206

13. Skodol AE. Revision of the personality disorder model for DSM-5 [letter]. Am J Psychiatry. 2011;168(1):97. PubMed doi:10.1176/appi.ajp.2010.10101466

14. Kendler K, Kupfer D, Narrow W, et al. Guidelines for Making Changes to DSM-5. Revised October 21, 2009. Accessibility verified August 19, 2011.

15. Task Force on DSM-IV. DSM-IV Options Book: Work in Progress 9/1/99. Washington, DC: American Psychiatric Association; 1991.

16. Zimmerman M, Chelminski I, Young D, et al. Does DSM-IV already capture the dimensional nature of personality disorders? J Clin Psychiatry. 2011;72(10):1333-1339.

17. Zimmerman M. Is there adequate empirical justification for radically revising the personality disorders section for DSM-5? [published online ahead of print April 18, 2011] Personal Disord. doi:10.1037/a0022108

18. Zimmerman M, Chelminski I, Young D, et al. Impact of deleting 5 DSM-IV personality disorders on prevalence, comorbidity, and the association between personality disorder pathology and psychosocial morbidity. J Clin Psychiatry. In press.

19. First MB. Clinical utility in the revision of the Diagnostic and Statistical Manual of Mental Disorders (DSM). Prof Psychol Res Pr. 2010;41(6):465-473. doi:10.1037/a0021511

20. World Health Organization. The ICD-10 Classification of Mental and Behavioral Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva, Switzerland: World Health Organization; 1992.

21. Kass F, Skodol AE, Charles E, et al. Scaled ratings of DSM-III personality disorders. Am J Psychiatry. 1985;142(5):627-630. PubMed

22. Oldham JM, Skodol AE. Charting the future of axis II. J Pers Disord. 2000;14:17-29. PubMed doi:10.1521/pedi.2000.14.1.17

23. Skodol AE. Dimensionalizing existing personality disorder categories. In: Millon T, Krueger RF, Simonsen E, eds. Contemporary Directions in Psychopathology: Scientific Foundations of the DSM-5 and ICD-11. New York, NY: Guilford Press; 2010:362-373.

24. Cohen P, Crawford TN, Johnson JG, et al. The Children in the Community Study of developmental course of personality disorder. J Pers Disord. 2005;19(5):466-486. PubMed doi:10.1521/pedi.2005.19.5.466

25. Skodol AE, Gunderson JG, Shea MT, et al. The Collaborative Longitudinal Personality Disorders Study (CLPS): overview and implications. J Pers Disord. 2005;19(5):487-504. PubMed doi:10.1521/pedi.2005.19.5.487

26. Kupfer DJ, Regier DA. Neuroscience, clinical evidence, and the future of psychiatric classification in DSM-5. Am J Psychiatry. 2011;168(7):672-674. PubMed doi:10.1176/appi.ajp.2011.11020219

Submitted: August 16, 2011; accepted August 16, 2011.

Corresponding author: Michael B. First, MD, Department of Psychiatry, Columbia University and Department of Clinical Phenomenology, New York State Psychiatric Institute, 1051 Riverside Dr-Unit 60, New York, NY 10032 (

J Clin Psychiatry 2011;72(10):1341-1343 (doi:10.4088/JCP.11com07337).

© Copyright 2011 Physicians Postgraduate Press, Inc.

Related Articles

Volume: 72

Quick Links: