<< Back  

 Can't open the PDF? Click here for help.


Does the diagnosis of multiple Axis II disorders have clinical significance?

Mark Zimmerman, MD

Department of Psychiatry and Human Behavior, Brown Medical School, Department of Psychiatry, Rhode Island Hospital, Providence, RI, USA

Janine N. Galione, BS

Department of Psychiatry and Human Behavior, Brown Medical School, Department of Psychiatry, Rhode Island Hospital, Providence, RI, USA

Iwona Chelminski, PhD

Department of Psychiatry and Human Behavior, Brown Medical School, Department of Psychiatry, Rhode Island Hospital, Providence, RI, USA

Diane Young, PhD

Department of Psychiatry and Human Behavior, Brown Medical School, Department of Psychiatry, Rhode Island Hospital, Providence, RI, USA

Kristy Dalrymple, PhD

Department of Psychiatry and Human Behavior, Brown Medical School, Department of Psychiatry, Rhode Island Hospital, Providence, RI, USA

Theresa A. Morgan, MPhil

Department of Psychiatry and Human Behavior, Brown Medical School, Department of Psychiatry, Rhode Island Hospital, Providence, RI, USA

BACKGROUND: A stated goal of the DSM-5 Work Group on Personality and Personality Disorders (PDs) has been to reduce the high rate of comorbidity among PDs. Few studies have examined whether the diagnosis of multiple PDs has clinical significance. In the present report from the Rhode Island Methods to Improve Diagnostic Assessment and Services project, we tested the hypothesis that patients with >1 DSM-IV PD would have more severe forms of psychopathology than patients who were diagnosed with only 1 DSM-IV PD.

METHODS: A total of 2,150 psychiatric outpatients were evaluated with semi-structured diagnostic interviews for DSM-IV Axis I and Axis II disorders and measures of psychosocial morbidity.

RESULTS: For 8 of the 10 PDs, the majority of patients had at least 1 additional PD, although at least 20% of patients diagnosed with each PD were diagnosed with only 1 PD. Compared with patients with 1 PD, patients with ≥2 PDs had significantly more psychosocial morbidity.

CONCLUSIONS: The co-occurrence of PDs conveys clinically significant information. Moreover, despite high levels of comorbidity, each PD also existed as a stand-alone entity. These findings raise questions about the DSM-5 Work Group’s emphasis on reducing comorbidity in Axis II.

KEYWORDS: personality disorder, comorbidity, DSM-5



The publication of DSM-III was a watershed event in the study of personality disorders (PDs). The operational definition of 11 PDs in DSM-III was followed by the development of research diagnostic interviews that allowed for the reliable assessment of these disorders1 and continued growth of research on PDs.2

Along with a burgeoning empiric literature, a stream of critical commentary of the DSM’s section on PDs also has been steady since the publication of DSM-III.3-6 One consistent criticism of the PD section has been the high rate of comorbidity among PDs.5,6 The high rate of diagnostic comorbidity has raised questions of whether the PDs represent unique clinical entities, which has been cited as 1 of the reasons for changing the approach towards diagnosing PDs in DSM-5. Indeed, the empiric goal of decreasing “excessive comorbidity” was identified on the DSM-5 Web site (www.dsm5.org) as a core rationale for including only 5 PD types within a larger trait-dimensional framework.7 However, it is unclear what the Work Group views as the clinical benefit of reducing the number of PDs. Also unstated is whether there is an acceptable level of comorbidity, either clinically or empirically.

Although a large literature base has established high rates of diagnostic comorbidity among the Axis I disorders in DSM-III and DSM-IV, as well as the clinical significance of such comorbidity,8-18 few studies have examined the clinical significance of diagnosing multiple Axis II disorders. In a sample of psychiatric outpatients, Nakao et al19 found that the number of DSM-III-R PDs was significantly correlated with ratings on the Global Assessment of Functioning (GAF) scale. Soeteman et al20 found the number of DSM-IV PDs was significantly associated with quality of life. The results of these studies suggest that co-occurrence of PDs provides clinically useful information of either severity of illness or decrease in functioning. We are not aware of any studies examining whether having >1 PD is associated with clinical indices of illness severity such as psychiatric hospitalization, occupational functioning, and suicide attempts.

In this report from the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project, we tested the hypothesis that patients with >1 PD would have more severe forms of psychopathology than patients who were diagnosed with only 1 PD. As markers of illness severity, we examined the number of Axis I disorders, suicidal ideation at the time of the evaluation, history of suicide attempts, history of psychiatric hospitalizations, GAF ratings, and time missed from work due to psychiatric disorders.


A total of 2,150 psychiatric outpatients were evaluated with semi-structured diagnostic interviews for DSM-IV Axis I and Axis II disorders in the Rhode Island Hospital’s Department of Psychiatry outpatient practice. This private practice group mostly treats individuals with health insurance (including Medicare, but not Medicaid) on a fee-for-service basis, and is distinct from the hospital’s outpatient residency training clinic that predominantly serves lower income, uninsured, and medical assistance patients. TABLE 1 shows the demographic characteristics of the sample. The majority of the patients were white, female, married or single, and had some college education. The mean age was 38.5 years (standard deviation [SD] = 12.8). The most frequent current DSM-IV diagnoses were major depressive disorder (43.0%), social phobia (26.8%), generalized anxiety disorder (19.3%), and panic disorder (17.7%) (TABLE 2).


Demographic characteristics of 2,150 psychiatric outpatients

Characteristic n %
  Female 1,310 60.9%
  Male 840 39.1%
  Less than high school 178 8.3%
  Graduated high school 1,343 62.5%
  Graduated college or greater 629 29.3%
Marital status
  Married 869 40.4%
  Living with someone 127 5.9%
  Widowed 36 1.7%
  Separated 112 5.2%
  Divorced 325 15.1%
  Never married 681 31.7%
  White 1,952 90.8%
  Black 95 4.4%
  Hispanic 58 2.7%
  Asian 21 1.0%
  Other 24 1.1%
Age (years) Mean = 38.5 SD = 12.8
SD: standard deviation.


Current DSM-IV Axis I diagnoses of 2,150 psychiatric outpatientsa

DSM-IV diagnosis n %
Major depressive disorder 925 43.0%
Bipolar disorder 111 5.2%
Dysthymic disorder 179 8.3%
Generalized anxiety disorder 415 19.3%
Panic disorder 381 17.7%
Social phobia 576 26.8%
Specific phobia 225 10.5%
Obsessive-compulsive disorder 138 6.4%
Posttraumatic stress disorder 247 11.5%
Adjustment disorder 149 6.9%
Schizophrenia 8 0.4%
Eating disorder 143 6.7%
Alcohol abuse/dependence 207 9.6%
Drug abuse/dependence 103 4.8%
Somatoform disorder 167 7.8%
Attention-deficit/hyperactivity disorder 138 6.4%
Impulse control disorder 254 11.8%
aIndividuals could be given more than 1 diagnosis

Patients were interviewed by a trained diagnostic rater who administered the Structured Interview for DSM-IV Personality (SIDP-IV)21 and a modified version of the Structured Clinical Interview for DSM-IV (SCID).22 The Rhode Island Hospital institutional review committee approved the research protocol, and all patients provided informed, written consent. A minority of patients evaluated in the practice received the SCID and SIDP-IV because of a lack of available diagnostic raters or patients’ preference for a less time-consuming standard clinical evaluation. Patients who did and did not participate in the study were similar in sex, education, marital status, and scores on self-administered symptom questionnaires.23

We integrated into the SCID interview the item from the Schedule for Affective Disorders and Schizophrenia (SADS)24 on the amount of time missed from work due to psychiatric reasons during the past 5 years. The SCID/SADS interview also included assessments of prior psychiatric hospitalizations, current suicidal ideation (rated on a 0 to 6 scale on the SADS), and lifetime history of suicide attempts. Based on the results of the SCID/SADS and SIDP-IV interviews, the GAF was rated.

Diagnostic raters included PhD-level psychologists and research assistants with university degrees in the social or biologic sciences. Research assistants received 3 to 4 months of training during which they observed at least 20 interviews and were observed and supervised in their administration of >20 evaluations. Psychologists only observed 5 interviews and were observed and supervised in their administration of 15 to 20 evaluations. During the course of training, the senior author met with each rater to review the interpretation of every item on the SCID. Also during training every interview was reviewed on an item-by-item basis by the senior rater who observed the evaluation, and by the senior author who reviewed the case with the interviewer. At the end of the training period the raters were required to demonstrate exact or near exact agreement with a senior diagnostician on 5 consecutive evaluations. Throughout the MIDAS project, ongoing supervision of the raters consisted of weekly diagnostic case conferences involving all members of the team. In addition, every case was reviewed by the senior author.

Reliability of PD diagnoses was examined in 47 patients. A joint-interview design was used in which 1 rater observed another conducting the interview, and both raters independently made their ratings. The reliabilities of any PD (k = 0.90), any Cluster A (k = 0.79), B (k = 0.79), or C (k = 0.93) PD were good to excellent. The reliability of diagnosing ≥2 PDs (k = 0.81) also was high. Too few patients were diagnosed with individual PDs to calculate kappa coefficients.

Data analysis

We compared 3 groups of patients—those diagnosed with 0, 1, or ≥2 PDs. We faced a decision as to where to place patients diagnosed with PD not otherwise specified (PDNOS). We operationally defined PDNOS as falling 1 criterion below the DSM-IV diagnostic threshold on ≥2 PDs and not meeting criteria for any of the DSM-IV PDs. We hypothesized that patients with PDNOS would be less severely ill than patients who met full criteria for a PD, and therefore we were concerned placing patients with PDNOS in the group with 1 PD might bias the results by potentially inflating the difference between the groups with 1 PD vs ≥2 PDs. Therefore we included patients with PDNOS with patients with 0 diagnoses, although this biased the findings against finding a difference between the groups with 0 vs 1 PD.

We examined the relationship between the number of Axis II disorders and the following indicators of illness severity: number of current Axis I disorders, lifetime Axis I disorders, lifetime psychiatric hospitalizations, lifetime suicide attempts, suicidal ideation at the time of the evaluation, GAF ratings, and amount of time unemployed during the past 5 years due to psychiatric reasons. The outcome variables were examined categorically as well as continuously. We a priori defined as indicators of severe illness: ≥3 Axis I disorders, ≥3 psychiatric hospitalizations, ≥3 suicide attempts, a rating of ≥3 on the SADS suicidal ideation item, a GAF rating of ≤50, and unemployed due to psychiatric reasons for at least 2 years in the past 5 years. Analyses of variance were used to compare 3 groups (0, 1, or ≥2 Axis II disorders) on continuously distributed variables. Tukey follow-up tests were conducted for 2-group comparisons. Categorical variables were compared by the chi-square statistic. For all analyses pairwise comparisons were done only when the 3-group analysis was significant.


More than one-quarter of the patients were diagnosed with 1 of the 10 DSM-IV PDs (28.6%, n = 614). Among the 614 patients with a PD, the mean number of disorders was 1.5 (SD = 1.1). TABLE 3 shows that avoidant PD, borderline PD, and obsessive-compulsive PD were the most frequent disorders in the sample. For 8 of the 10 PDs, the majority of patients had another PD. The mean number of PDs was highest in patients with schizotypal PD (TABLE 3).


Frequency of comorbidity in 2,150 psychiatric outpatients with specific DSM-IV personality disorders

  Number with an Axis I disorder Number with another personality Axis II disorder Number with ≥2 additional personality Axis II disorders Total number of Axis I disorders Total number of additional Axis II disorders
Personality disorder n n % n % n % Mean SD Mean SD
Paranoid 69 66 95.6% 51 73.9% 24 34.8% 3.3 1.9 1.3 1.2
Schizoid 18 17 94.4% 12 66.7% 5 27.8% 2.5 1.5 1.1 1.1
Schizotypal 10 10 100% 6 60.0% 5 50.0% 3.0 2.0 1.5 1.6
Histrionic 17 16 94.1% 8 47.1% 6 35.3% 2.8 2.1 1.1 1.3
Borderline 204 199 97.5% 109 53.4% 41 20.1% 3.3 1.7 0.8 1.0
Antisocial 41 36 87.8% 24 58.5% 13 31.7% 3.1 2.2 1.1 1.1
Narcissistic 40 34 85.0% 26 65.0% 14 35.0% 2.4 1.9 1.2 1.1
Dependent 39 39 100% 28 71.8% 15 38.5% 3.6 1.4 1.2 0.9
Obsessive-compulsive 147 143 97.3% 67 45.6% 23 15.6% 2.6 1.6 0.7 0.9
Avoidant 285 284 99.6% 108 37.9% 33 11.6% 3.4 1.6 0.5 0.8

Compared with patients without a PD, the patients with 1 PD had more current and lifetime Axis I disorders, attempted suicide more frequently, reported more suicidal ideation at the time of the evaluation, were hospitalized more frequently, had poorer social functioning, and missed more time from work due to psychiatric illness (TABLE 4). Patients with a PD were rated significantly lower on the GAF. Likewise, TABLE 4 shows that compared with patients with 1 PD, patients with ≥2 PDs had significantly more current and lifetime Axis I disorders, history of suicide attempts, suicidal ideation at the time of the evaluation, number of psychiatric hospitalizations, time missed from work due to psychiatric illness, and lower ratings on the GAF. Importantly, these results were the same when the number of Axis I disorders was entered as a covariate. All indicators of severe illness were significantly more frequent in patients with 1 PD compared with patients with 0 PDs, and patients with ≥2 PDs compared with patients with 1 PD (TABLE 5).


Association between number of DSM-IV personality disorders and indicators of illness severity in 2,150 psychiatric outpatients

  Number of personality Axis II disorders  
Indicator of illness severity (mean, SD) 0 (n = 1,536) 1 (n = 431) ≥2 (n = 183) 3-group testa 3-group significanceb 0 vs 1 1 vs ≥2
No. of current Axis I disorders 1.6 (1.2) 2.7 (1.6) 3.5 (1.7) F(2, 398.1) = 180.3 <.001 <.001 <.001
No. of lifetime Axis I disorders 2.7 (1.6) 3.9 (1.9) 4.9 (1.8) F(2, 417.9) = 187.7 <.001 <.001 <.001
Global Assessment of Functioning 55.8 (9.1) 50.2 (8.4) 46.1 (8.4) F(2, 443.9) = 151.3 <.001b <.001b <.01b
Suicidal ideation 0.6 (1.0) 1.1 (1.3) 1.7 (1.5) F(2, 399.6) = 62.7 <.001b <.0b <.001b
No. of suicide attempts 0.3 (1.5) 0.8 (4.9) 1.5 (4.4) F(2, 353.4) = 8.4 <.001b <.01b <.05b
No. of psychiatric hospitalizations 0.4 (1.0) 0.6 (1.3) 0.9 (1.5) F(2, 393.9) = 12.9 <.001b <.01b <.05b
Time unemployed in past 5 yearsc 2.2 (1.7) 2.8 (2.1) 3.7 (2.4) F(2, 364.4) = 38.8 <.001b <.001b <.001b
aThe assumption of homogeneity was not met, therefore the Welch F ratio is reported and the degrees of freedom are adjusted.
bAnalyses controlled for number of Axis I disorders.
cPatients who were not expected to work (eg, student, retired) were excluded leaving a final sample of 1,362 with 0 disorders, 400 with 1 disorder, and 165 with ≥2 disorders


Association between number of DSM-IV personality disorders and indicators of severe illness in 2,150 psychiatric outpatients

  Number of personality Axis II disorders  
Indicators of severe illness (%, n) 0 (n = 1,536) 1 (n = 431) ≥2 (n = 183) 3-group test 3-group significance 0 vs 1 1 vs ≥2
≥3 current Axis I disorders 20.6% (317) 49.9% (215) 69.9% (128) χ2 = 280.2 <.001 <.001 <.001
≥3 lifetime Axis I disorders 47.9% (736) 76.3% (329) 91.3% (167) χ2 = 205.3 <.001 <.001 <.001
Global Assessment of Functioning <50 29.4% (452) 55.3% (238) 74.2% (135) χ2 = 203.0 <.001 <.001 <.001
Serious suicidal ideation 7.4% (113) 15.8% (68) 30.1% (55) χ2 = 98.9 <.001 <.001 <.001
History of ≥3 attempts 2.9% (44) 7.0% (30) 14.8% (27) χ2 = 57.8 <.001 <.001 <.01
History of ≥3 psychiatric hospitalizations 5.4% (83) 8.1% (35) 14.8% (27) χ2 = 24.4 <.001 <.05 <.05
Unemployed ≥2 years in past 5 yearsa 7.5% (102) 14.5% (58) 30.3% (50) χ2 = 85.6 <.001 <.001 <.001
aPatients who were not expected to work (eg, student, retired) were excluded leaving a final sample of 1,362 with 0 disorders, 400 with 1 disorder, and 165 with ≥2 disorders


Since the publication of DSM-III, critics of the PD section have commented that high rates of comorbidity are problematic. Because of the “comorbidity problem,” the PD Work Group for DSM-5 recommended reducing the number of specific PDs. The initial recommendation was to delete 5 PDs.25 More recently this has been modified to recommend deleting 4 PDs.26 The Work Group also proposes the inclusion of a dimensional index of functional impairment meant to capture severity of PD pathology.26

What information would be lost if the “comorbidity problem” were fully addressed? As noted in the introduction, few studies have examined the clinical significance of multiple Axis II diagnoses. Most studies of PDs have examined their frequency and correlates in patients with a single or limited number of Axis I disorders.27-33 In these studies, the typical comparison is between patients with and without any PD, or between patients with a diagnosis in 1 of the 3 PD clusters. Few studies have assessed the range of PDs in heterogeneous samples of psychiatric patients, and, of those the largest studies have been based on unstructured clinical evaluations.34-39 A problem with using unstructured diagnostic interviews is under-recognition of psychopathology in routine clinical practice.23,40,41 Such studies suggest that private practitioners typically diagnose only 1 PD even when criterion-level information suggests patients meet criteria for multiple PDs.42 Finally, the studies that have used semi-structured interviews to assess PDs have had relatively small sample sizes,43-50 only 1 of which included >400 patients.46 In order to have an adequate sample size of patients with >1 PD it is necessary to evaluate a large sample. It would be less costly to ascertain a large sample by using a self-report measure of PDs; however, these scales have high false-positive rates and the percentage of patients with multiple Axis II disorders would be overestimated.51,52

Was any single PD overrepresented in the group with multiple Axis II disorders? The patients with multiple PDs were more frequently hospitalized, attempted suicide, and diagnosed with more Axis I disorders—a profile that seems characteristic of borderline PD.53,54 Although borderline PD was the most frequent PD among patients with multiple Axis II disorders, less than one-quarter of the patients with multiple PDs were diagnosed with borderline PD. Therefore, as TABLE 3 shows, multiple Axis II disorders should not be equated with borderline PD or any other single PD.

Critics of PD classification have indicated that high rates of comorbidity raise questions about whether each PD is a unique entity.6 The data in TABLE 3 indicates that each PD was diagnosed as a stand-alone disorder without another PD at least 20% of the time. How frequently must a disorder occur alone for it to be considered a unique entity? We are not aware of any specific recommendations regarding the minimum percentage necessary to suggest sufficient diagnostic uniqueness, although our opinion is that a stand-alone frequency of >20% is more than sufficient to qualify as a unique disorder. More importantly, it appears that removing 5 of the PD categories (as originally proposed on the DSM-5 Web site) does not “solve” the issue of high comorbidity: the remaining disorders continue to co-occur with similar frequency, between 20% to 30%.55 Although the inclusion of a “Personality Disorder Trait Specified” diagnosis may allow for some representation of comorbidity using traits rather than categories, it is currently unclear to what extent the proposed traits would provide commensurate clinical information. Therefore, it would be key to have evidence demonstrating the clinical and research utility of the proposed system for DSM-5 as compared with the current system.

The current proposal for the DSM-5 also details an index of “self and interpersonal functioning,” essentially a likert-type rating of severity of the disorder and resultant functional impairment. Recent research demonstrates that “severity” is a strong predictor of PD treatment outcome.56,57 It is possible that an index of severity/dysfunction could capture some of the variance accounted for by comorbidity, but a global severity index is unlikely to capture specific dysfunction unique to each PD or even PD trait.57 Even more importantly, we are not aware of any research that compares the clinical or predictive utility of the proposed severity system as compared to identification of ≥1 specific PDs. Clearly, more research needs be done on the potential loss or gain of information in the changes to Axis II as they are currently proposed.

Before concluding, the limitations of the study should be recognized. The study was conducted in a single clinical practice in which the majority of patients were white, female, and had health insurance. Replication of the results in other clinical samples with different demographic characteristics is warranted. The same interviewers completed the Axis I and Axis II interviews and assessed all variables. Although it is possible that the interviewers were biased because the ratings were not blind, the study was not specifically designed to examine the question that was the primary focus of this article, and therefore we are skeptical that rater bias was responsible for the findings. Strengths of the study are the large, comprehensive sample size, and the use of highly trained diagnostic interviewers to reliably administer a semi-structured diagnostic interview.


The co-occurrence of PDs conveys clinically significant information. Patients with ≥1 PD have more severe levels of psychopathology characterized by more Axis I pathology, psychiatric hospitalizations, suicide attempts, and missed time from work. Moreover, despite high levels of comorbidity, each PD existed as a stand-alone entity. These findings raise questions about the need to radically modify the DSM-IV PD section in order to reduce levels of comorbidity.

DISCLOSURE: The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.


  1. Zimmerman M. Diagnosing personality disorders: a review of issues and research methods. Arch Gen Psychiatry. 1994;51:225–245.
  2. Blashfield RK, Intoccia V. Growth of the literature on the topic of personality disorders. Am J Psychiatry. 2000;157:472–473.
  3. Tyrer P. What’s wrong with DSM-III personality disorders? J Pers Disord. 1988;2:281–291.
  4. Tyrer P, Ferguson B. Problems in the classification of personality disorder. Psychol Med. 1987;17:15–20.
  5. Clark LA. Assessment and diagnosis of personality disorder: perennial issues and an emerging reconceptualization. Annu Rev Psychol. 2007;58:227–257.
  6. Trull TJ, Durrett CA. Categorical and dimensional models of personality disorder. Annu Rev Clin Psychology. 2005;1:355–380.
  7. Skodol AE, Bender DS, Morey LC, et al. Personality disorder types proposed for DSM-5. J Pers Disord. 2011;25:136–169.
  8. Henry C, Van den Bulke D, Bellivier F, et al. Anxiety disorders in 318 bipolar patients: prevalence and impact on illness severity and response to mood stabilizer. J Clin Psychiatry. 2003;64:331–335.
  9. Vieta E, Colom F, Corbella B, et al. Clinical correlates of psychiatric comorbidity in bipolar I patients. Bipolar Disord. 2001;3:253–258.
  10. Otto MW, Simon NM, Wisniewski SR, et al. Prospective 12-month course of bipolar disorder in out-patients with and without comorbid anxiety disorders. Br J Psychiatry. 2006;189:20–25.
  11. Uher R, Mors O, Hauser J, et al. Body weight as a predictor of antidepressant efficacy in the GENDEP project. J Affect Disord. 2009;118:147–154.
  12. Gunstad J, Phillips KA. Axis I comorbidity in body dysmorphic disorder. Compr Psychiatry. 2003;44:270–276.
  13. Carpentier PJ, Krabbe PF, van Gogh MT, et al. Psychiatric comorbidity reduces quality of life in chronic methadone maintained patients. Am J Addict. 2009;18:470–480.
  14. Howland RH, Rush AJ, Wisniewski SR, et al. Concurrent anxiety and substance use disorders among outpatients with major depression: clinical features and effect on treatment outcome. Drug Alcohol Depend. 2009;99:248–260.
  15. Grant BF, Hasin DS, Stinson FS, et al. Prevalence, correlates, co-morbidity, and comparative disability of DSM-IV generalized anxiety disorder in the USA: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Psychol Med. 2005;35:1747–1759.
  16. Angst J, Gamma A, Endrass J, et al. Obsessive-compulsive syndromes and disorders: significance of comorbidity with bipolar and anxiety syndromes. Eur Arch Psychiatry Clin Neurosci. 2005;255:65–71.
  17. Roy-Byrne PP, Stang P, Wittchen HU, et al. Lifetime panic-depression comorbidity in the National Comorbidity Survey. Association with symptoms, impairment, course and help-seeking. Br J Psychiatry. 2000;176:229–235.
  18. Simon NM, Otto MW, Wisniewski SR, et al. Anxiety disorder comorbidity in bipolar disorder patients: data from the first 500 participants in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). Am J Psychiatry. 2004;161:2222–2229.
  19. Nakao K, Gunderson JG, Phillips KA, et al. Functional impairment in personality disorders. J Pers Disord. 1992;6:24–33.
  20. Soeteman DI, Verheul R, Busschbach JJ. The burden of disease in personality disorders: diagnosis-specific quality of life. J Pers Disord. 2008;22:259–268.
  21. Pfohl B, Blum N, Zimmerman M. Structured Interview for DSM-IV Personality. Washington DC: American Psychiatric Press, Inc.; 1997.
  22. First MB, Spitzer RL, Gibbon M, et al. Structured Clinical Interview for DSM-IV Axis I Disorders - Patient edition (SCID-I/P, version 2.0). New York, NY: Biometrics Research Department, New York State Psychiatric Institute; 1995.
  23. Zimmerman M, Mattia JI. Psychiatric diagnosis in clinical practice: is comorbidity being missed? Compr Psychiatry. 1999;40:182–191.
  24. Endicott J, Spitzer RL. A diagnostic interview: the Schedule for Affective Disorders and Schizophrenia. Arch Gen Psychiatry. 1978;35:837–844.
  25. American Psychiatric Association. Personality and personality disorders. http://www.dsm5.org/ProposedRevisions/Pages/PersonalityandPersonalityDisorders.aspx. Accessed July 26 2010.
  26.  American Psychiatric Association. Changes to the reformulation of personality disorders for DSM-5 (Updated June 21 2011). 2011. http://www.dsm5.org/ProposedRevision/Pages/PersonalityDisorders.aspx. Accessed June 28, 2011.
  27. Bowden-Jones O, Iqbal MZ, Tyrer P, et al. Prevalence of personality disorder in alcohol and drug services and associated comorbidity. Addiction. 2004;99:1306–1314.
  28. Tenney NH, Schotte CK, Denys DA, et al. Assessment of DSM-IV personality disorders in obsessive-compulsive disorder: comparison of clinical diagnosis, self-report questionnaire, and semi-structured interview. J Pers Disord. 2003;17:550–561.
  29. Ross S, Dermatis H, Levounis P, et al. A comparison between dually diagnosed inpatients with and without Axis II comorbidity and the relationship to treatment outcome. Am J Drug Alcohol Abuse. 2003;29:263–279.
  30. Zikos E, Gill KJ, Charney DA. Personality disorders among alcoholic outpatients: prevalence and course in treatment. Can J Psychiatry. 2010;55:65–73.
  31. Sakai R, Nestoriuc Y, Nolido NV, et al. The prevalence of personality disorders in hypochondriasis. J Clin Psychiatry. 2010;71:41–47.
  32. Courjaret J, Schotte CK, Wijnants H, et al. Chronic fatigue syndrome and DSM-IV personality disorders. J Psychosom Res. 2009;66:13–20.
  33. Bagby RM, Vachon DD, Bulmash E, et al. Personality disorders and pathological gambling: a review and re-examination of prevalence rates. J Pers Disord. 2008;22:191–207.
  34. Koenigsberg HW, Kaplan RD, Gilmore MM, et al. The relationship between syndrome and personality disorder in DSM-III: experience with 2,462 patients. Am J Psychiatry. 1985;142:207–212.
  35. Kass F, Skoldol AE, Charles E, et al. Scaled ratings of DSM-III personality disorders. Am J Psychiatry. 1985;142:627–630.
  36. Oldham JM, Skodol AE. Personality disorders in the public sector. Hosp Community Psychiatry. 1991;42:481–487.
  37. Mezzich J, Coffman G, Goodpastor S. A format for DSM-III diagnostic formulation: experience with 1,111 consecutive patients. Am J Psychiatry. 1982;139:591–596.
  38. Kenan MM, Kendjelic EM, Molinari VA, et al. Age-related differences in the frequency of personality disorders among inpatient veterans. Int J Geriatr Psychiatry. 2000;15:831–837.
  39. Molinari V, Ames A, Essa M. Prevalence of personality disorders in two geropsychiatric inpatient units. J Geriatr Psychiatry Neurol. 1994;7:209–215.
  40. Basco MR, Bostic JQ, Davies D, et al. Methods to improve diagnostic accuracy in a community mental health setting. Am J Psychiatry. 2000;157:1599–1605.
  41. Zimmerman M, Mattia JI. Differences between clinical and research practice in diagnosing borderline personality disorder. Am J Psychiatry. 1999;156:1570–1574.
  42. Herkov MJ, Blashfield RK. Clinician diagnosis of personality disorders: evidence of a hierarchical structure. J Pers Assess. 1995;65:313–321.
  43. Thuo J, Ndetei DM, Maru H, et al. The prevalence of personality disorders in a Kenyan inpatient sample. J Pers Disord. 2008;22:217–220.
  44. Grilo C, McGlashan T, Quinlan D, et al. Frequency of personality disorders in two age cohorts of psychiatric inpatients. Am J Psychiatry. 1998;155:140–142.
  45. Dahl AA. Some aspects of DSM-III personality disorders illustrated by a consecutive sample of hospitalized patients. Acta Psychiatr Scand. 1986;328:61–67.
  46. Fossati A, Maffei C, Bagnato M, et al. Patterns of covariation of DSM-IV personality disorders in a mixed psychiatric sample. Compr Psychiatry. 2000;41:206–215.
  47. Marinangeli M, Butti G, Scinto A, et al. Patterns of comorbidity among DSM-III-R personality disorders. Psychopathology. 2000;33:69–74.
  48. Oldham JM, Skodol AE, Kellman HD, et al. Comorbidity of axis I and axis II disorders. Am J Psychiatry. 1995;152:571–578.
  49. Ottosson H, Bodlund O, Ekselius L, et al. DSM-IV and ICD-10 personality disorders: a comparison of a self-report questionnaire (DIP-Q) with a structured interview. Eur Psychiatry. 1998;13:246–253.
  50. Stangl D, Pfohl G, Zimmerman M, et al. A structured interview for the DSM-III personality disorders. A preliminary report. Arch Gen Psychiatry. 1985;42:591–596.
  51. Widiger TA, Samuel DB. Evidence-based assessment of personality disorders. Psychol Assess. 2005;17:278–287.
  52. Zimmerman M, Coryell WH. Diagnosing personality disorders in the community. A comparison of self-report and interview measures. Arch Gen Psychiatry. 1990;47:527–531.
  53. Zanarini MC, Frankenburg FR, Dubo ED, et al. Axis I comorbidity of borderline personality disorder. Am J Psychiatry. 1998;155:1733–1739.
  54. Zimmerman M, Mattia JI. Axis I diagnostic comorbidity and borderline personality disorder. Compr Psychiatry. 1999;40:245–252.
  55. 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. 2012;73:202–207.
  56. Hopwood CJ, Malone JC, Ansell EB, et al. Personality assessment in DSM-5: empirical support for rating severity, style, and traits. J Pers Disord. 2011;25:305–320.
  57. Morey LC, Berghuis H, Bender DS, et al. Toward a model for assessing level of personality functioning in DSM-5, Part II: Empirical articulation of a core dimension of personality pathology. J Pers Assess. 2011;93:347–353.

CORRESPONDENCE: Mark Zimmerman, MD Bayside Medical Center 235 Plain Street Providence, RI 02905 USA E-MAIL: mzimmerman@lifespan.org