Classification of personality disorders varies significantly, with the predominant models being either categorical or dimensional. As in the case of broader classification of mental disorders, personality disorders are mainly classified in accordance with two diagnostic frameworks: namely, the Diagnostic and Statistical Manual of Mental Disorders (DSM) and the International Classification of Diseases (ICD). , the latest editions of these are the DSM-5-TR and ICD-11, respectively. While the main system in the former classifies personality disorders as distinct categories; the latter classifies a single personality disorder dimensionally according to severity, with the option to additionally diagnose trait domains. A hybrid approach is implemented in the Alternative DSM-5 Model for Personality Disorders, with diagnoses being specific or trait specified; both of these are based on both severity and traits. The ICD-11 classifies schizotypal disorder among primary psychotic disorders rather than as a personality disorder as in the DSM-5.
Conceptual approaches
Personality disorder classification can generally be broken down into a categorical approach and a dimensional approach. The categorical approach views personality disorders as discrete entities that are distinct from each other as well as from normal personality. In contrast, the dimensional approach suggests that personality disorders exist on a continuum, with traits varying in degree rather than kind. There has been a sustained movement toward replacing categorical models of personality disorder classification with dimensional approaches. This dimensional perspective may allow for more nuanced understanding and flexible diagnostic practices.
Categorical approach Classical views of personality disorder as discrete categories have had benefits for understanding and communicating psychopathology throughout history, such as for: a contained organization of symptoms to facilitate standardized research, organizing public awareness and stigma reduction campaigns, allocating public health funding and appropriate treatment intensities, and normalizing clear labels for communicating patient formulations (a description of symptoms and their inter-relationships) to professionals and families.
Since its inception, the categorical system has steadily accumulated criticism. Attempts to reproduce the factor structure of the DSM-IV-TR's categorical model have been unsuccessful, suggesting that the categorical structure cannot robustly describe the architecture of personality psychopathology. Such issues are exacerbated by the substantial symptom overlap between disorders that facilitates their excessive and unwarranted comorbidity, with the majority of people with a PD being eligible for another PD diagnosis. As a result, individuals are substantially more likely to be diagnosed with several PDs than a singular one, contradicting the notion that categories provide neat constellations of inter-related symptoms.
Equally, this approach appears unable to accurately capture the full range of personality psychopathology. Estimates of patients who do not fit neatly into current categories range from 21 to 49%, accordingly given the general diagnosis of Personality Disorder – Not Otherwise Specified (PD-NOS). PD-NOS also appears to be in regular usage to describe mixed or complex presentations given the difficulties in classifying individuals within the current framework. It has been found that "many patients in clinical practice misleadingly receive multiple PD diagnoses, a 'not otherwise specified' PD diagnosis, or no PD diagnosis at all, even if a PD diagnosis is relevant to the presentation". Another issue is the heterogeneity within categories.
Setting standardized diagnostic thresholds (based upon polythetic symptoms) is difficult particularly when each symptom is given equal weighting. This means that individuals with the same number of symptoms can have substantially different levels of distress. Between each PD, diagnostic thresholds occur at different levels of pathology. Due to these issues, it is likely that many clinicians use their clinical judgment based upon an internalized representation of the disorder when making diagnoses. The current categorical approach falls short of fully representing personality psychopathology and providing a scientifically robust understanding of what personality is and what disorders of personality are.
Dimensional approach
In response to observed deficiencies in the categorical approach, dimensional models, which suggest that humans differ in degree not in kind, have been developed, assessing personality disorders in terms of severity of impairment and maladaptive personality traits. Within this perspective, PD occurs at maladaptive extremes of the standard personality traits all humans share and as specific combinations of these trait extremes. The degree of life impairment forms the basis for a PD diagnosis. This approach has gained substantial support, with broad calls and movements toward mainstream adoption.
The shift towards dimensional models is reflected in the inclusion of the AMPD in Section III of the DSM-5, and in the ICD-11's adoption of a dimensional system. These are believed to ameliorate several shortcomings of the categorical model, as well as improve clinical utility and potentially reduce stigma, although no research has so far specifically examined the effect on stigma. Emerging research indicates that dimensional models may also facilitate the personalization of psychotherapy by aligning treatment strategies with underlying trait dimensions rather than diagnostic categories.Zimmermann, J., Kerber, A., Rek, K., Hopwood, C. J., & Krueger, R. F. (2019). Integrating clinical and personality assessment: Toward the development of a shared framework. Journal of Personality Assessment, 101(3), 292–304. doi:10.1080/00223891.2018.1483373. Despite some important differences in the prevailing approaches, dimensional models of PD typically consider two key criteria: severity and style.
+Mapping of ICD-11 PD classification to the AMPDICD-11AMPDSeverity levelCriterion ANoneNo impairment (0)Personality difficultyMild impairment (1)Mild personality disorderModerate impairment (2)Moderate personality disorderSevere impairment (3)Severe personality disorderExtreme impairment (4)Traits and patternsCriterion BNegative affectivityNegative AffectivityDetachmentDetachmentDisinhibitionDisinhibitionDissocialityAntagonismAnankastia(Rigid perfectionism)(Schizotypal disorder)Psychoticism
Severity Severity captures the core distress that is common to all PDs, its impact on the individual's self-direction and identity (intrapersonal functioning), as well as their ability to form close relationships and empathize with others (interpersonal functioning). Indices of global severity are robust predictors of both the presence of a personality disorder and prognosis, and track with fluctuations in clinical functioning. According to the ICD-11, severity is the key and sole requirement for making a diagnosis of PD. The central placement of impairment is grounded in research that global severity ratings are sensitive and specific predictors of PD, and provide better estimates of clinician-rated psychosocial impairment than specific categorical diagnoses do. The severity of personality disorder (i.e., mild, moderate, severe) may be more indicative of dysfunction and outcomes than the specific typology of the disorder.
Style The second criterion describes the stylistic features of the presentation, largely in relation to some derivation of the Five-Factor Model (FFM) of personality. The DSM-5's Alternative Model of Personality Disorders (AMPD) Criterion B comprises the traits of negative affectivity (continua from emotional stability to neuroticism), detachment (introversion to extroversion), antagonism (agreeableness to antagonism), disinhibition (conscientiousness to impulsivity), and psychoticism (closed to experience to open to experience). The DSM-5's approach to diagnosing PD in the AMPD differs from the ICD-11 as it requires the presence of one or more elevated traits. Nevertheless, there is a growing interest in using only Criterion A for understanding, diagnosing, and managing PD. The FFM has the ability to explain all personality variation, with current dimensional PD models capturing dysfunctional versions or extremes of these traits.
DSM-5 (section II)
In the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders, a categorical classification was retained for personality disorders. Located in Section II (Diagnostic Criteria and Codes; where official diagnoses are listed), it contains ten specific personality disorders grouped into three clusters (A, B, and C), as well as three other diagnoses. Thus, it lists personality disorders in the same way as other mental disorders, rather than on a separate "axis", as previously. The clusters are based on descriptive similarity between the disorders they encompass, and it is not proven that they possess clinical utility.
The clusters, as well as definition of personality disorders being done through specific sets of criteria, have been part of the DSM since the DSM-III (1980). The classification system was retained from the DSM-IV (1994) due to the Board of Trustees of the American Psychiatric Association having decided to reject the AMPD. This system was carried forward in the more recent DSM-5-TR.
The DSM-5 and the more recent DSM-5-TR provide a definition and six criteria for general personality disorder. Any of its ten personality disorder diagnoses is subject to this definition, which requires that a differential diagnosis is performed in order to verify that the disturbance is not the result of other mental disorders, medical conditions or substances, and that the disturbance is stable over time and "inflexible and pervasive across a broad range of personal and social situations", having evident continuity since "at least to adolescence or early adulthood". Additionally, disturbance must be evident in regards to at least two of four specified aspects of functioning, namely: cognition, affectivity, interpersonal functioning and impulse control.
Cluster A People with these disorders can be paranoid and have difficulty being understood by others, as they often have odd or eccentric modes of speaking and an unwillingness and inability to form and maintain close relationships. Significant evidence suggests a small proportion of people with Cluster A personality disorders, especially schizotypal personality disorder, have the potential to develop schizophrenia and other psychotic disorders. These disorders also have a higher probability of occurring among individuals whose first-degree relatives have either schizophrenia or a Cluster A personality disorder. Importantly, contrary to common belief, a recent meta-analysis shows that diagnosis remission rates and treatment dropout rates in patients with Cluster A personality disorders are not significantly different from those of other personality disorders.
Cluster B Cluster B personality disorders are characterized by dramatic, impulsive, self-destructive, emotional behavior and sometimes incomprehensible interactions with others.
Cluster C Cluster C personality disorders are characterised by a consistent pattern of anxious thinking or behavior. Earlier, this cluster has included passive–aggressive personality disorder in the DSM-III-R.
Other personality disorders The DSM-5 chapter on personality disorders also contains three diagnoses for conditions not matching these ten disorders, which nevertheless exhibit characteristics of a personality disorder:
Personality change due to another medical conditionpersonality disturbance due to the direct effects of a medical condition