6b
Christopher C. Conway
The categorical model of personality disorder (PD) diagnosis, despite its familiarity, may not be the optimal way to represent personality pathology. Weinberg (this volume) enumerates its principal defects: rampant comorbidity, within-diagnosis heterogeneity, poor coverage of personality problems encountered in the clinic, and arbitrary boundaries between disorders and between disorder and wellness.
For many researchers, these problems signal a pressing need to develop evidence-based rubrics that can supplement, and eventually supplant, DSM and ICD as alternate routes to research, assessment, and treatment of PD (e.g., Hopwood et al., 2018). One dimensional system is now gaining traction in the research literature as a viable alternative to the categorical approach: the Hierarchical Taxonomy of Psychopathology (HiTOP; Kotov et al., 2017). In this chapter, I will present evidence that supports HiTOP as a PD research framework.
Sketching the HiTOP Model
HiTOP is based on quantitative modeling of the co-occurrence of signs and symptoms of mental illness. This research tradition dates back to factor analytic studies of youth psychopathology symptoms. Achenbach (1966) and others showed that anxiety, depressive, and somatic complaints clustered together to form an internalizing spectrum, whereas disruptive and inattentive behaviors cohered around an externalizing spectrum. Decades later, Krueger and coworkers (e.g., Krueger, Caspi, Moffitt, & Silva, 1998) followed this same quantitative strategy to uncover latent internalizing and externalizing dimensions accounting for diagnostic comorbidity in adult samples.
These latent dimensions subsequently replicated across diverse demographic groups, developmental stages, and regions of the world. Antisocial and borderline PD features often were involved in these modeling efforts, but clinical disorder symptoms predominated in the early stages. More recently, this sort of quantitative modeling research routinely has incorporated a wider array of mental health problems, including the full complement of PD conditions (e.g., Markon, 2010).
The dimensions emerging from these structural models coalesced into the HiTOP framework depicted in Figure 6.b.1. One key property is that the model is hierarchically oriented. Thus, PD can be conceptualized and assessed at various levels of abstraction, depending on the research objective. At the apex, a general factor of psychopathology is presumed to cut across all expressions of mental illness. At the spectrum level, the familiar internalizing and externalizing dimensions are joined by somatoform, thought disorder (reflecting liability to schizoid, schizotypal, and paranoid PDs), and detachment (reflecting avoidant, dependent, histrionic [inversely], and schizoid PDs). At a lower level still, spectra divide into subfactors, such as antisocial behavior, which is a marker of antisocial PD. The DSM PD entities are located at the syndrome level, which, in turn, subsumes symptom components and maladaptive traits at the most granular level of the hierarchy.
Figure 6.b.1
Hierarchical Taxonomy of Psychopathology (HiTOP) consortium working model. Constructs higher in the figure are broader and more general, whereas constructs lower in the figure are narrower and more specific. Dashed lines denote provisional elements requiring further study. At the lowest level of the hierarchy (i.e., traits and symptom components), for heuristic purposes, conceptually related signs and symptoms (e.g., Phobia) are indicated in bold, with specific manifestations indicated in parentheses. Abbreviations–ADHD: attention-deficit/hyperactivity disorder; GAD: generalized anxiety disorder; IED: intermittent explosive disorder; MDD: major depressive disorder; OCD: obsessive-compulsive disorder; ODD: oppositional defiant disorder; SAD: separation anxiety disorder; PD: personality disorder; PTSD: posttraumatic stress disorder.
The superspectrum, spectrum, and subfactor levels account for the comorbidity commonly observed across PDs. For instance, the well-known overlap of borderline PD with depressive and trauma-related disorders is explained by shared dependence on the distress subfactor (Figure 6.b.1). Meanwhile, the comorbidity between borderline and other Cluster B PDs is reflected in a common placement in the antagonistic externalizing spectrum. In latent variable modeling terms, borderline PD cross-loads on (or is saturated by) both internalizing and externalizing dimensions (e.g., Eaton et al., 2011). The heterogeneous expression of borderline PD is captured by various maladaptive traits represented at the base of the hierarchy. The borderline-relevant traits are scattered across several spectra (e.g., identity problems and affective lability under internalizing; impatient urgency and aggression under disinhibited externalizing). In sum, the HiTOP model allows researchers to conceptualize borderline and other PDs in terms of basic units (e.g., maladaptive traits and symptom components); broader, cross-cutting liabilities (e.g., internalizing and antagonistic externalizing spectra); or both.
HiTOP as a Research Framework for Personality Problems
For most PD researchers, HiTOP is a new and potentially foreign taxonomy of personality problems. This unfamiliarity could render it inaccessible to most mental health professionals, unless investigators are able to clearly show its practical utility. And I argue that HiTOP does indeed confer benefits, relative to DSM, that extend beyond its quantitative, empirical approach to nosology. Specifically, the dimensions composing HiTOP are useful research targets (Conway et al., 2019). They can illuminate the precise pathways that connect personality problems with putative causes and consequences. They can also make PD research more efficient.
To illustrate obstacles (to utility and efficiency) the categorical system creates for researchers, let’s take the case of borderline PD. Borderline PD generally has poor discriminant validity. While it is related to plenty of negative outcomes (e.g., hospitalization, romantic dysfunction), so are many other PDs. This observation is problematic because borderline PD rarely presents in isolation, and instead often co-occurs with other PDs including antisocial and schizotypal PDs.
Researchers are therefore faced with the dilemma of trying to disentangle the putative effects of borderline versus other PDs to uncover its unique causes and correlates. One solution is to recruit a “pure” sample of patients with only a borderline diagnosis, but this decision implies collecting an extremely unrepresentative sample, given the laundry list of conditions typically accompanying borderline in clinically referred samples. Another approach is to collect a representative sample of people with borderline PD who have comorbid diagnoses. The collection of representative data has its merits. At the same time, this strategy also is limited by the fact that statistical associations between borderline PD and outcomes potentially are confounded by co-occurring conditions and thereby undercuts inferences about the unique role of borderline PD. This state of affairs has led, in my view, to borderline PD failing several of Robins and Guze’s (1970) standards for disorder validation. That is, there are no laboratory tests (or other criteria) that reliably distinguish borderline from all other PDs and clinical disorders. While Weinberg’s (this volume) Table 6.4 summarizes some abnormalities observed in borderline PD (e.g., amygdala hyperreactivity, hypothalamic-pituitary-adrenal axis dysfunction), these characteristics are not specific (i.e., pathognomonic) to that condition. This is the trend for the full gamut of PD entities.
Research based on the HiTOP framework bypasses this dilemma. There is no expectation that syndromes have specific connections with psychopathology causes, correlates, or treatments. In fact, the bulk of the evidence to date suggests that the most potent presumed causes of PD (e.g., child maltreatment, molecular genetic risks) operate at higher-order levels (e.g., superspectrum, spectrum; cf. Conway et al., 2019). For instance, in the National Epidemiological Study of Alcohol and Related Conditions, child abuse was not related directly to antisocial PD, but rather to an overarching externalizing spectrum theorized to represent the commonality of several PDs (e.g., antisocial, borderline, histrionic) and clinical disorders (e.g., alcohol use disorder, drug use disorder; Keyes et al., 2012).
Another way to describe this research framework is that the supposed connection between a syndrome of interest and some covariate can be examined across multiple levels of the HiTOP hierarchy. Take the example of borderline PD and psychosocial functioning. Weinberg (this volume) points out that role functioning in social, occupational, and leisure arenas is disrupted across many PDs, but possibly borderline PD most of all. With the benefit of the HiTOP lens, researchers can empirically examine – not just assume – what part (or parts) of borderline PD account for this close relationship with social dysfunction (cf. Hopwood et al., 2011).
The borderline syndrome itself, comprised of the nine diagnostic criteria codified in DSM-IV and DSM-5 Section II, may be at the heart of functioning deficits. Alternately, social dysfunction may be a fully transdiagnostic characteristic, in which case its strongest connection would be with the general factor of psychopathology. Incidentally, borderline patients have been found to have especially high standing on this general factor (Sharp et al., 2015). Another possibility is that social dysfunction is best explained at the spectrum level, perhaps by variation in disinhibited externalizing. Finally, symptom components or maladaptive traits could be the most potent predictors of dysfunction. Most clinicians would attest that traits like risk-taking, hostility, and withdrawal often contribute to diminished performance in work and social spheres. This research process, which acknowledges the hierarchical organization of personality problems, can more precisely identify the etiological and clinical pathways to disorder, impairment, and treatment response.
Summary and Conclusion
The categorical model of PD diagnosis continues to guide most clinical and research training worldwide. Weinberg (this volume) illustrates that categorical PDs are the basis for most research activity – contributing to the appearance of strong validity – and can be useful summary constructs in clinical communication. Yet a voluminous research literature over the past 50 years has shed light on undeniable problems with this approach for PD research, assessment, and treatment across mental health disciplines. An international consortium of mental health professionals has assembled to study the HiTOP as an evidence-based alternative to the DSM and ICD rubrics (for more detail, see Kotov et al., 2017; Krueger et al., 2018).
I underlined here that HiTOP potentially has value not just for PD nosology, but also for PD research. There are clear motivations for a new perspective on PD research. Few diagnostic categories are routinely studied, discriminant validity evidence is scarce, and categories are so heterogeneous that the exact mechanisms linking PDs to their causes and consequences are difficult to establish.
The HiTOP system offers an alternate approach that could revitalize PD research. Under this model, mental health correlates are examined not just in relation to syndromal PD, but also to dimensions at higher and lower levels of breadth in an empirical hierarchy of psychopathology processes. Through this sort of reformulation of research questions, etiological and clinical models of PD stand to become more precise, accurate, and efficient, but research on this framework in the PD literature is just beginning.
References
Achenbach, T. M. (1966). The classification of children’s psychiatric symptoms: A factor-analytic study. Psychological Monographs: General and Applied, 80, 1–37.
Conway, C. C., Forbes, M. K., Forbush, K. T., Fried, E. I., Hallquist, M. N., Kotov, R., … Eaton, N. R. (2019). A hierarchical taxonomy of psychopathology can transform mental health research. Perspectives on Psychological Science, 14, 419–436.
Eaton, N. R., Krueger, R. F., Keyes, K. M., Skodol, A. E., Markon, K. E., Grant, B. F., & Hasin, D. S. (2011). Borderline personality disorder co-morbidity: Relationship to the internalizing–externalizing structure of common mental disorders. Psychological Medicine, 41, 1041–1050.
Hopwood, C. J., Kotov, R., Krueger, R. F., Watson, D., Widiger, T. A., Althoff, R. R., … Bornovalova, M. A. (2018). The time has come for dimensional personality disorder diagnosis. Personality and Mental Health, 12, 82–86.
Hopwood, C. J., Malone, J. C., Ansell, E. B., Sanislow, C. A., Grilo, C. M., McGlashan, T. H., … Gunderson, J. G. (2011). Personality assessment in DSM-5: Empirical support for rating severity, style, and traits. Journal of Personality Disorders, 25, 305–320.
Keyes, K. M., Eaton, N. R., Krueger, R. F., McLaughlin, K. A., Wall, M. M., Grant, B. F., & Hasin, D. S. (2012). Childhood maltreatment and the structure of common psychiatric disorders. British Journal of Psychiatry, 200, 107–115.
Kotov, R., Krueger, R. F., Watson, D., Achenbach, T. M., Althoff, R. R., Bagby, R. M., … Zimmerman, M. (2017). The Hierarchical Taxonomy of Psychopathology (HiTOP): A dimensional alternative to traditional nosologies. Journal of Abnormal Psychology, 126, 454–477.
Krueger, R. F., Caspi, A., Moffitt, T. E., & Silva, P. A. (1998). The structure and stability of common mental disorders (DSM-III-R): A longitudinal-epidemiological study. Journal of Abnormal Psychology, 107, 216–227.
Krueger, R. F., Kotov, R., Watson, D., Forbes, M. K., Eaton, N. R., Ruggero, C. J., … Zimmermann, J. (2018). Progress in achieving quantitative classification of psychopathology. World Psychiatry, 17, 282–293.
Markon, K. E. (2010). Modeling psychopathology structure: A symptom-level analysis of Axis I and II disorders. Psychological Medicine, 40, 273–288.
Robins, E., & Guze, S. B. (1970). Establishment of diagnostic validity in psychiatric illness: Its application to schizophrenia. American Journal of Psychiatry, 126, 983–987.
Sharp, C., Wright, A. G., Fowler, J. C., Frueh, B. C., Allen, J. G., Oldham, J., & Clark, L. A. (2015). The structure of personality pathology: Both general (‘g’) and specific (‘s’) factors? Journal of Abnormal Psychology, 124, 387–398.