7c
Joshua D. Miller and Thomas A. Widiger
We appreciate the thoughtful commentaries provided by De Fruyt and De Clercq (this volume) as well as Sellbom (this volume) on our chapter on trait models of personality disorder (PD). We use this rejoinder to address a few of the comments articulated in the two commentaries, starting with De Fruyt and De Clercq’s.
Drs. De Fruyt and De Clercq
De Fruyt and De Clercq are quite supportive of the overall trait model approach albeit they appear to be more pessimistic about the ease and amount of time before trait-based models of PDs are instantiated in diagnostic taxonomies in a more central and singular manner. For instance, they state that there is a “long way to go before trait-based models will be formally accepted in established diagnostic taxonomies.” However, it now appears that the World Health Organization will approve in 2019 for ICD-11 a five-domain trait model that will replace (with one exception) the diagnostic PD categories (Reed et al., 2019). This trait model overlaps substantially with the alternative model of PD (AMPD) included in DSM-5 and will consist of negative affectivity, detachment, disinhibition, dissocial, and anankastic, along with a borderline pattern specifier; diagnoses will also include a single severity dimension ranging from mild to severe personality disorder. We believe clinicians will adapt and adopt these new models with relative ease given their relatively straightforward application and their prior judgments of the substantial clinical utility of trait models (e.g., see Widiger, 2019, for a review).
We largely agree with De Fruyt and De Clercq’s suggestions regarding the measurement of impairment, particularly the notion that its assessment should follow rather than precede the assessment of personality disorder traits and their call for a more explicit focus on concrete domains of functioning (e.g., love and work; see also Pilkonis, Hallquist, Morse, & Stepp, 2011). There is even a degree of specificity of the five-factor model personality domains with regard to the impairment identified by De Fruyt and De Clercq, with the domains of antagonism and introversion concerning interpersonal impairments, low conscientiousness concerning work and school, and neuroticism concerning level of distress (Mullins-Sweatt & Widiger, 2010). We note, however, that it is remarkably difficult to separate the assessment of personality traits, especially personality disorder traits, from impairment given that the vast majority of traits inherently include information with respect to impairment (e.g., Miller, Sleep, & Lynam, 2018; Sleep, Lynam, Widiger, Crowe, & Miller, 2019). That is, elevations on trait domains such as negative affectivity, antagonism, or disinhibition already provide information about impairment.
Dr. Sellbom
Sellbom’s commentary focuses on additional challenges to the adoption of these models including which level of the trait hierarchy should be used, broader domains or narrower facets. It is noteworthy that the DSM-5 and ICD-11 are at odds in this respect such that the ICD-11 model operates only at the domain level, whereas the DSM-5 AMPD focuses on 25 narrower traits that covary in such a way as to yield five higher order factors. Sellbom raises the important point that more work is needed to determine the optimal level of abstraction needed that balances coverage and parsimony. Although facet level descriptions are more complex than domain level descriptions, we expect that clinicians will embrace the more nuanced description that facets allow for once the clinicians become more adept with the application of the five domains
Sellbom noted a number of other issues in adopting these models, including the lack of formal test manuals, normative data, and measures of non-credible responding for many but not all existing PD trait measures. Sellbom comes at these issues, in part, from the perspective of a scholar working primarily on issues related to the Minnesota Multiphasic Personality Inventory (MMPI), in which these criteria have long been met. For instance, Sellbom notes that the MMPI-2 Restructured Form uses “standardized scores of 1.5 to 2 SDs above a normative mean to indicate a clinical elevation.” We agree that normative data are informative, but are worried about the notion of identifying “clinical” elevations simply on the degree of elevation, as if all forms of psychopathology have an equal prevalence rate. Cut scores on maladaptive trait scales should be based (in part) on the degree of social and/or occupational impairment, which is the case for DSM-5 Section III and ICD-11. We agree normative data would be helpful but believe that a field-wide discussion of the type of normative data needed is necessary, as well as the solicitation of funds necessary to collect samples of this nature. Such funding will be necessary given the costs of such endeavors. The American Psychiatric Association might well be able to fund such work given that it owns the copyright to the most commonly used measure of the DSM-5 AMPD. Of course, it should perhaps be noted that the profits from the clinical and research applications of the MMPI help to fund its validation and construction of supportive material, whereas the DSM-5 AMPD is assessed by freely available measures. Finally, we agree with Sellbom that the creation of scales to measure non-credible responding (e.g., inconsistent responding; over- and under-reporting) will be important, particularly if these scales are to be used in high stake settings. Such scales have been created for the family of FFM PD scales discussed in our chapter but these scales require further validation as do the newer post-hoc scales being created for the measures aligned with the DSM-5 AMPD.
Conclusions
In conclusion, while some challenges to the adoption of trait-based models of PD remain, including those highlighted by Drs. De Fruyt, De Clercq, and Sellbom, we are optimistic that they are surmountable given the immense motivation within the field to move to a more empirically valid and useful model of personality disorders. The instantiation of such models in both the DSM-5 and the soon to be released ICD-11 suggest that such long anticipated change is finally here (e.g., Frances, 1993).
References
Frances, A. (1993). Dimensional diagnosis of personality: Not whether, but when and which. Psychological Inquiry, 4, 110–111.
Miller, J. D., Sleep, C. E., & Lynam, D. R. (2018). DSM-5 alternative model of personality disorder: Testing the trait perspective captured in criterion B. Current Opinion in Psychology, 21, 50–54.
Mullins-Sweatt, S. N., & Widiger, T. A. (2010). Personality-related problems in living: An empirical approach. Personality Disorders: Theory, Research, and Treatment, 1, 230–238.
Pilkonis, P. A., Hallquist, M. N., Morse, J. Q., & Stepp, S. D. (2011). Striking the (im)proper balance between scientific advances and clinical utility: Commentary on the DSM-5 proposal for personality disorders. Personality Disorders: Theory, Research, and Treatment, 2, 68–82.
Reed, G. M., First, M. B., Kogan, C. S., Hyman, S. E., Gureje, O., Gaebel, W., … Saxena, S. (2019). Innovations and changes in the ICD-11 classification of mental, behavioural and neurodevelopmental disorders. World Psychiatry, 18, 3–19.
Sleep, C. E., Lynam, D. R., Widiger, T. A., Crowe, M., & Miller, J. D. (2019). An evaluation of DSM-5 Section III Personality Disorder Criterion A (Impairment) in accounting for psychopathology. Psychological Assessment, 31(10), 1181–1191
Widiger, T. A. (2019). Considering the research: Commentary on “The trait-type dialectic: construct validity, clinical utility, and the diagnostic process.” Personality Disorders: Theory, Research, and Treatment, 10, 215–219.