Review articles

By Dr. Simon B Thompson , Ms. Tanyth Macdonald
Corresponding Author Dr. Simon B Thompson
Psychology Research Centre , Bournemouth University, - United Kingdom BH12 5BB
Submitting Author Dr. Simon B Thompson
Other Authors Ms. Tanyth Macdonald
Psychology Research Centre, Bournemouth University, - United Kingdom BH12 5BB


Anxiety, Assessment, Categorisation, Depression, Diagnosis, Dimensional labelling, DSM manual, Labelling, Personality disorder, Prognosis

Thompson SB, Macdonald T. Categorising Personality Disorder: Entering a New dimension?. WebmedCentral PSYCHIATRY 2011;2(12):WMC002669
doi: 10.9754/journal.wmc.2011.002669

This is an open-access article distributed under the terms of the Creative Commons Attribution License(CC-BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Click here
Submitted on: 17 Dec 2011 12:26:19 PM GMT
Published on: 18 Dec 2011 07:14:02 AM GMT


Defining personality disorder is complicated and has been long debated in the past in terms of category versus dimension. This trend towards making an entity a part of other entities and a part of a larger continuum or dimension is growing in popularity and examples can be drawn from other diagnostic criteria such as that of anorexia nervosa (Thompson, 1993). There is current debate on whether personality disorders should be labelled by category versus dimension. There is a need to reconcile the debate not only for diagnosing and treating patients but also for establishing accurate and agreed criteria for the next generations of the diagnostic and statistical manuals of disorders that are widely cited and used. A balanced view is presented together with proposals for change. This theoretical framework adds value to the diagnostic process.


Personality disorders (PDs) are reported to be particularly prevalent in clinical settings, both in psychiatry and in mainstream medicine (Tyrer, 2008), and are estimated to affect between 9% - 13% of the adult population (Comer, 2011). PDs are currently defined in the Diagnostic and Statistical Manual of Disorders (DSM) (Fourth Edition) as an enduring pattern of behaviour that is markedly different from the individual's culture and has an onset in adolescence or early adulthood. It is also stable over time and leads to distress or impairment. (APA, 2000). PDs first appeared in the Third Edition of DSM and were described as constellations of personality traits that cause either significant functional impairment or subjective distress. (APA, 1980). The DSM-IV-TR (APA, 2000) introduced a major shift for the classification of PDs by placing them on Axis-II; this separated them from the more acute and episodic illnesses described on Axis-I. This helped to distinguish PDs in terms of pervasiveness, chronicity, age of onset and resistance to treatment (Trull & Durrett, 2005).
The categorical approach is based on assumptions that mental illnesses are considered to be medical conditions and that they have boundaries between normality and illness (Trull & Durrett, 2005). The DSM-IV incorporates this medical model and defines PDs within a categorical, hierarchical, taxonomic (categories) system, where the 10 PDs are polythetic (ie only a subset of criteria must be met to be eligible for a diagnosis (Cooper, Balis & Zimmerman, 2010). Therefore, approximately 7 - 9 items of each subset must be fulfilled to meet the diagnostic threshold for a PD (Trull & Durrett, 2005). The DSM-IV includes a hierarchical system identified by three clusters: Cluster A - odd/eccentric; Cluster B – dramatic or erratic; Cluster C – anxious or inhibited (APA, 2000). The rationale for this hierarchical approach was to help clinicians identify the PDs more readily by grouping those with similar features (Trull & Durrett, 2005).
Cluster A includes clients who may be described as withdrawn, distant, cold or suspicious (Comer, 2011). This cluster includes the PDs: paranoid, schizoid and schizotypal. Paranoid PD includes a pattern of pervasive distrust and suspiciousness to others and clients view others’ motives as malevolent (APA, 2000). Schizoid PD is described as a pattern of detachment from social relationships as well as a restricted range of emotional affect (APA, 2000). The last PD is schizotypal which is described as a pattern of acute discomfort in close relationships, odd ways of thinking as well as behavioural eccentricities (APA, 2000).
Cluster B PDs, are described as dramatic, emotional and attention seeking (Comer, 2011). These disorders include anti-social, borderline, histrionic and narcissistic personality disorder. Anti-social personality disorder is described as persistent disregard and the violation of others rights (APA, 2000). Often, they display patterns of misbehaviour from as young as 15 years old which include cruelty, running away, truancy as well as destroying property (Comer, 2011). Borderline personality disorder (BPD) is a pattern of instability in interpersonal relationships, self- image and marked impulsivity (APA, 2000). BPD patients tend to engage in reckless behaviours such as unsafe sex, substance and alcohol abuse as well as self-harming (Comer, 2011). Histrionic PD is described as a pattern of excessive emotionality and attention seeking. These clients tend to have exaggerated moods and use grandiose gestures and mannerisms to gain attention (Comer, 2011). Lastly, clients with narcissistic PD believe they are grandiose and have no empathy for others (APA, 2000).
The last cluster, Cluster C, is classified as clients who are described as tense, anxious or uptight (Comer, 2011). These disorders include avoidant, dependent and obsessive compulsive PD. People with avoidant PD are uncomfortable and inhibited in social situations; they feel overwhelmed by feelings of inadequacy and are sensitive to criticisms (APA, 2000) Dependent PD is described as a pattern of submissive and clinging behaviour related to the excessive need to looked after by others. These clients are clingy, fearing separation and incapable of making decisions (Comer, 2011). Obsessive-compulsive PD is described as being pre-occupied with orderliness, perfection and control. Their constant concern for being perfect is said to impair their productivity (Comer, 2011).
A key concern in PD literature is whether the structure of personality dysfunction is represented by ‘qualitatively distinct categories’ or ‘quantitatively varying dimensions’ (Blackburn, 2000, p. 5). An extensive amount of evidence seems to point towards a dimensional approach (Blackburn, 2000). One reason, may be that dimensional models portray a more complete and accurate quantitative description (Sprock, 2003). Perhaps, this is because scores are given for a series of dimensions, and the issues regarding poor fit, borderline cases and co-morbidity are no longer a concern as they would be with regards to a categorical approach. Many researchers over the years have debated whether the categorical approach to PD diagnosis is clinically useful, and if it adequately reflects the true nature of these disorders (Trull & Durret, 2005). Tyrer and colleagues (2006 p. S51) stated “the assessment of personality disorder is currently inaccurate, largely unreliable, frequently wrong, and in need of improvement.” The DSM-IV-TR has been derived out of only a small number of empirical studies, and this is said to be because the DSM-IV-TR lacks an appropriate structure (Widiger, Livesley & Clark, 2009). The categorical approach was rationally derived rather than empirically, and though there is little empirical evidence to support the structure of Axis-II clusters, this approach has been the choice of PD classification in all editions of the DSM thus far (Sheets & Craighead, 2007).
Currently, there is still no agreement between researchers and clinicians as to which approach is superior. Although more evidence demonstrates validity for a dimensional approach, there are a number of clinicians who favour the traditional categorical approach; for example, Verheul (2005) argued that most of the advantages of the categorical approach are minimal and that dimensional models that are intelligent can accommodate the purposes of diagnostic systems. The DSM-IV explicitly states that the “clustering system has serious limitations and has not been consistently validated” (APA, 1994, p. 630) and with the immense amount of criticisms gathered in research against the categorical approach it may be that the next generation of the manual, DSM-5, is to adopt a dimensional model. The question about which dimensional model to implement still remains a topic of debate and controversy.


To date, there have been 18 proposed models. The Five-Factor-Model (FFM) and Shedlers and Westen’s Assessment Procedure Approach (SWAP) will be reviewed here. The rationale for reviewing these two models is that they are proposed to be incorporated into the DSM-5. Furthermore, the FFM has received an extensive amount of evidence over the years.
Evidence for the categorical approach
Although there is little published evidence for the categorical approach, it can still be argued that there are a few advantages to this approach. For instance, it has been noted that the categorical approach neatly maps onto the decision of whether to provide treatment or not to provide treatment (Clark, 2007). With a single categorical diagnosis it does simplify the choice of clinical focus and appropriate treatment for that PD. A commonly cited advantage is that the categorical approach is easier to use for purposes of communication and conceptualisation. Burgess (2008) discusses a study in which over 80% of experts reported that the current categorical system is helpful with communication, and at least two thirds expressed that Axis-II is easy to use. Using discrete categories, it is said to be cognitively efficient; rather than describing and remembering all the features and characteristics of each person, one can identify them simply as having a PD, eg BPD (Blackburn, 2000; Rottman, et al., 2010).
Some clinicians agree that they may be resistant towards a dimensional approach as categories are familiar, consistent with tradition and are rated as easier to use (Sprock, 2003). Furthermore, clinicians have received extensive training in this approach and it may be difficult for them to move towards an entirely new one (Verheul, 2005). Some argue that it would be a shift in practice to move towards the dimensional approach (Widiger, 1993). It is suggested that simplification or revision of a dimensional approach that has clinical utility and provides empirical evidence should be incorporated into the DSM-5.
Sprock (2003) investigated the clinical utility of categorical, hybrid (existing categories re-conceptualised as dimensions) and dimensional approaches to diagnosing prototypic (when a person meets all of the diagnostic criteria for one disorder and does not have significant features of other disorders and non-prototypic PD cases). The study compared inter-rater reliability and clinician’s perceptions of the clinical utility of the current categorical approach, several proposed hybrid models and dimensional models (including the FFM) in the diagnosis of PD. Six vignettes were employed in this study, one prototypic and one non-prototypic case for each of the three PD clusters. The sample size was 500 practicing psychologists who were matched on age, sex and years of experience. The three measures of clinical utility included usefulness for professional communication, case conceptualization, and treatment planning (Sprock, 2003).
The results suggested that psychologists assigning categorical diagnoses to the prototypic cases were high in reliability. One shortcoming, noted by the author, is that most patients are not prototypic, as they do tend to meet some criteria for other PDs. This limits the categorical applicability (Sprock, 2003). With regards to the FFM, Sprock (2003) reported that clinicians rated the DSM-IV significantly higher to use than the FFM on all measures of clinical utility and concluded that overall clinicians did not find the dimensional models useful. Additionally, the psychologists were not confident in the usage and application of a dimensional approach as opposed to a categorical and hybrid approach.
Sprock (2003) did recognise that a major limitation of her study was that the vignettes were not based on real-life individuals with personality problems. Samuel and Widiger (2006) have examined real-life case vignettes including the case of Ted Bundy, a notorious serial killer, Ernest, a patient who suffered extreme loneliness and led a fantasy life based on storybook characters, and Madeline, a patient who suffered extreme abuse and engaged in self- destructive behaviours throughout her life. From this study the authors concluded that a dimensional model, specifically, the FFM, has greater clinical utility than the current categorical approach, and from the evidence provided they support a move towards a dimensional approach (Samuel & Widiger, 2006).
Evidence against the categorical approach
It is rare to come across a published paper that does not explicitly state all the limitations of the categorical approach. There seems to be a theme that runs through every paper which includes: co-morbidity, heterogeneity, inadequate scientific base, diagnostic overlap, the lack of clear boundary between normality and abnormality, and the lack of diagnostic stability over time (Clark, 2007; Trull & Durret, 2005; Widiger & Trull, 2007; Zimmerman, 2011).
Co-morbidity is an extensively cited problem in the current diagnostic system and there is widespread literature to date that supports this. Co-morbidity is recognised as two or more mental disorders co-occurring (Clark, 2007). The problem of co-morbidity is between the 10 PDs and among PDs across other axes (Axis I). Co-morbidity can indicate a poorer prognosis of treatment as well as unmet service needs (Coid, et al., 2009). It has been agreed by many clinicians, that the problems with co-morbidity may reflect the flaws of the psychiatric classification in the DSM-IV (Coid, et al., 2009).
An extensive amount of research has demonstrated that the prevalence of PDs among Axis-I, anxiety or mood disorders is quite high (Grant, et al., 2005). In a sample cof 43,093 individuals who were 18 years and older, the co-occurrence between PDs and anxiety and mood disorders is pervasive (Grant, et al., 2005). Further, results showed that the associations between mood and anxiety disorders as well as PDs are statistically significantly. Specifically, the PDs, AVPD and dependent PD were strongly related to mood and anxiety disorders as compared to the other seven disorders. Paranoid and Schizoid PD were most associated with panic disorder with agoraphobia, social phobia as well as generalised anxiety disorder. This study was beneficial in that it used a large sample and demonstrated that co-occurrence between PDs and Axis-I disorders were quite high. One limitation may be that it this study only looked at a sample within the US population and did not take into account other populations. Therefore, the question that arises is whether these results can be generalised across other cultures.
Coid and colleagues (2009) looked at PD co-morbidity in offenders in the United Kingdom. The sample for this study included 61,944 prisoners: 46,872 were male sentenced prisoners, 12,303 were remand prisoners, and 2,770 were women prisoners. The SCID-II was used to assess Axis-II disorders; Axis-I disorders were assessed using the Schedule for Clinical Assessment in Neuropsychiatry. The results indicated that anti-social PD was the most prevalent PD in this sample; this may have been because they were offenders who were given a prison sentence which supports the criteria for this PD, such as law-breaking behaviour and violation of others (Comer, 2011). Anti-social and BPD were significantly co-morbid and were strongly associated with substance abuse. There was a significant association between anti-social and paranoid PD. Additionally, co-morbidity between paranoid PD and BPD has been supported by other studies such as Skodal and colleagues (2005). The results also demonstrated that BPD was considerably co-morbid with Axis-II disorders, co-occurring from clusters A, B and C (Coid, et al., 2009).
Although this study had various strengths, such as being the first study to examine PD co-morbidity with regards to a national sample of prisoners, the authors did cite a few limitations, such as there were no measures of inter-rater reliability, which is important when conducting a study using these assessments. A second limitation could be that only 70% of eligible prisoners participated in this survey which could be considered a selection bias.
Accumulated evidence demonstrates that PDs are not stable over time and therefore the DSM definitions are in doubt. One influential study was The Collaborative Longitudinal Personality Disorders Study Overview and Implications (CLPS) by Skodal and collegues (2005). This study compared four specific DSM-IV PDs with each other and to major depressive disorder. The assessments were completed at baseline, 6 months, at 1 year and annually until 2005. The four PDs that were selected for this study represented the different clusters: schizotypal from Cluster A, BPD from Cluster B, avoidant PD from cluster C. Obsessive-compulsive PD was selected because it has been shown by factor analyses that it is separable from the other PDs despite its assignment to cluster C. The ages of the sample ranged between 18 to 45 years and there was no history of schizophrenia, schizoaffective disorder or current substance intoxication.
One of the main aims of this study was to assess the stability of PDs. It has been shown that fewer than half of PD patients remain at or above full criteria every month over intervals as short as 1 or 2 years (Skodal, et al., 2005). The results demonstrated that 10% of BPD patients remitted in the first 6 months, and this remission was associated with situational changes, such as leaving stressful relationships. From these findings the question that arises is whether certain PDs are more temporally fluctuating than previously assumed (Skodal, et al., 2005). These results suggest that viewed as dimensions, PDs showed considerably more stability compared to categories.
This study was useful in that it provided empirical evidence over a long period of time about the reliability and stability of PD in both a categorical as well as dimensional approach. Although the study supports a move towards a dimensional approach there was a limitation: the assessment for this study only used Structured Clinical Interviews and did not explore other measurements such as self-report instruments.
Samuel and colleagues (2011) looked at the temporal stability with regards to self-report and interview assessed PDs. The motivation for this was that the CLPS only used semi-structured interviews when assessing PD stability. This study had a sample size of 668 participants; the participants were allocated to one of the four PDs (BPD, AVPD, schizotypal and OCPD). The assessments that were used included the SNAP-2, DIPD-IV as well as the Structured Clinical Interview (Samuel, et al., 2011). One interesting finding that emerged from this study was that differential stability does depend on whether a dimensional or categorical scoring approach is adopted. For instance, the dimensional PD scores demonstrated higher ‘rank-order’ and ‘mean-level stability’ than interviews; however, this finding was not replicated with regards to categorical diagnoses (Samuel, et al., 2011). This study did find that PDs do appear less stable then the DSM-IV indicates.
Zimmerman (2011) argues that diagnostic instability that is addressed in most papers as a major limitation of the current DSM-IV is not a fault of the diagnostic system. Rather, diagnostic instability should be expected because of imperfect reliability, assessment methods that do not establish adolescent onset, state effects, regression to the mean, measurement error due to repeated assessments and real life change. One reason as to why stability may change is due to the simple fact that people change. People learn from their experiences and change or adapt their beliefs and attitudes and this happens to most people on a daily basis throughout their life.
Blashfield and Intoccia (2000) conducted a computer search of all the PD literature until 2000. The authors concluded that the only PD that is ‘alive and well’ is BPD. Anti-social and schizotypy are growing and the remaining seven PDs are ‘dead’ or ‘dying’ (Blashfield & Intoccia, 2000, p. 473). There has not been very much evidence presented to date on the different PDs with the exception to BPD and Anti-social PD.
A further problem with the current categorical system is the arbitrary and unstable boundaries in PDs. According to Widiger and Trull (2007), the DSM-IV presents specific rules for distinguishing between the presence and absence of each of the individual diagnostic categories. The PDs schizotypal and BPD diagnosis are the only two PDs where a published rationale has been provided. There has not been a rationale or supportive discussion provided for AVPD, DPD, OCPD, schizoid, and paranoid PDs (Widiger & Trull, 2007). With regards to BPD and schizotypal diagnosis, the thresholds no longer apply; this is because their cut-off points were selected on the basis of agreement between clinicians and since then there have been revisions, deletions and additions to their criterion (Widiger & Trull, 2007).
Heterogeneity is one of the most commonly cited disadvantages of the DSM-IV which is where two people can be given the same diagnosis (eg BPD) but may present with completely different behaviours and symptoms. For example, it has been reported that there are 256 symptom combinations that can result in the diagnosis of BPD (Eaton, 2010). Several empirical observations regarding psychopathological variation challenge the categorical model expressed in the DSM-IV. Beseiro-Gonzalez, Lemos-Giraldez and Muniz (2004) conducted a study to examine the construct validity of the three PD clusters in the DSM-IV. The study assessed whether possible deficits or particular characteristics involved in pre-frontal executive functions, inhibition or activation of psycho-physiological processes or personality big traits could be of any use to explain and support the proposed PDs main constructs. All participants in this study met the criteria for a PD in the DSM-IV. The sample consisted of 138 subjects, 66 were males and 72 were females, exclusion criteria included subjects under the age of 17 and the presence of chronic substance abuse or a chronic organic condition. Furthermore, out of the sample, 42 subjects were diagnosed as belonging to Cluster A, 85 to cluster B and 21 to Cluster C. Different assessments were used such as the MCMIII, the BFQ and a variety of neuropsychological tests (Bestiero-Gonzalez, Lemos-Giraldez & Muniz, et al., 2004).
The results from this study demonstrated that whilst comparing the three DSM cluster groups on neuropsychological functions the results yielded no support for Cluster B and Cluster C distinction. However, they did in some respect confirm Cluster A differential characteristics based mainly on attention and concept formation deficits (Beseiro-Gonzalez, Lemos-Giraldez & Muniz, et al., 2004). With regards to the FMM there were no significant differences found among clusters in neuroticism and openness to experience, as well no significant differences were observed between DSM-IV cluster groups on conscientiousness and agreeableness. The findings from this study indicate that the clusters may be more heterogeneous than the DSM suggests and multidimensional in nature rather than un-dimensional and provide evidence that the division of the DSM PDs into three discrete clusters may not be empirically justified, as they provide weak evidence for its construct validity, with the exception of cluster A.
Every class of mental disorders includes a diagnosis of not otherwise specified (NOS) (APA, 2000). Clinicians are expected to use this diagnosis when they determine that a person has a particular class of a mental disorder but that the person’s symptoms are not adequately represented by any one of the individual diagnostic categories . With regards to PD (PDNOS) studies have reported that PDNOS is frequently used in Axis-II disorders, one reason for this may be that the DSM-IV-TR lacks specific diagnostic criteria for the individual PDs. Studies have cited that many clinicians are using this diagnosis frequently (Widiger & Trull, 2007). Verheul and Widiger (2004) conducted a meta-analysis of 51 studies and reported that PDNOS were the most frequently used Axis-II diagnosis with prevalence rates ranging from 21% to 49%. In studies that implemented structured interviews, PDNOS was the single most frequently used diagnosis, where in non-structured interviews; PDNOS was the third most frequently used PD diagnosis (Verheul & Widiger, 2004). Mullins-Sweatt and colleagues (2009) state that the reason for this may be because the current categories do not provide adequate coverage. These studies do suggest that clinicians are not finding the existing diagnostic categories to be adequate in their coverage of PD symptomology.
Evidence for the dimensional approach
The DSM-IV-TR acknowledges that an “alternative perspective to the categorical approach is the dimensional perspective, that personality disorders represent maladaptive variants of personality traits that merge imperceptibly into normality and into one another” (APA 2000, p. 689). In 2004, a research planning workshop sponsored by the APA presented papers towards adopting a dimensional model in the DSM-5. The papers examined dimensional models from multiple perspectives that included behaviour, genetics, neurobiology, childhood antecedent and cultural factors as well as alternative dimensional structures and clinical utility (Widiger, et al., 2008).
Widiger and Simonsen (2005) described 18 candidate dimensional approaches and stated that it would be unjustified to just pick a model and utilise it. Rather, a joint consensus needs to be reached about which model has the most evidence for it, clinical utility and will successfully transition into the DSM-5 (Clark, 2007). According to the evidence provided thus far, it suggests replacing the PDs with an empirically derived model of dimensional personality traits (Widiger & Simonesen, 2005). Krueger and Eaton (2010) suggest that the DSM-5 should incorporate personality and psychopathology to a greater degree than what the DSM-IV currently does, as most experts agree, that PDs are abnormal variants of normal personality.
Widiger and Trull (2007) establish that a dimensional approach of PD will address most weaknesses of the categorical model. For instance, it will get rid of heterogeneity through the use of multi-factorial descriptions of an individual’s psychopathology profile. These descriptions will not place persons into single diagnostic categories that do not take into account the unique symptoms, features and traits that characterise an individual person. A dimensional approach will rather describe an individual in terms of their own distinctive constellation of maladaptive personality traits. This will essentially remove the multiple co-morbid and overlapping PD diagnoses. Additionally, the dimensional approach will eliminate the not otherwise specified diagnosis choice (PDNOS).
Before any model is implemented into the DSM-5, it must provide evidence that it has clinical utility. Numerous studies suggest that the complexity of the DSM-IV classification of PDs interferes with user accuracy and evidence suggests that a dimensional model may not necessarily be as complicated, which is a major advantage to the ease of use for clinicians. Verheul (2005) concluded that categorical models have the least clinical utility, specifically with respect to coverage, subtlety clinical decision making and reliability. The purely dimensional models represent the most evidence for clinical utility. With regards to the hybrid models, they have advantages over the categorical model with respect to coverage, consistency with change, and subtlety but with regards to dimensional model have disadvantages in coverage, consistency with aetiology and clinical decision making (Verheul, 2005). If personality disorders are simply maladaptive variants of normal traits, and all persons have some degree of a personality disorder, then a diagnosis of personality disorder might lose credibility as a public health issue. Therefore, a dimensional model should aim to utilise the current categories in the upcoming DSM-5.
Five Factor model (FFM)
Normal personality was thought to be governed by traits, and though traits are unobservable, they are to some extent responsible for individual differences in overt behaviour (Livesley, et al., 1994). Highly similar traits govern both normal and disordered personality. Furthermore, rather than viewing the domain of PD as consisting of separate diagnosis, it can be seen as descriptions of traits in which individuals differ. Researchers agree that PDs are abnormal variants of normal personality traits that become intensified with development. A personality trait model would characterise current personality constructs and redefine PDs by using personality trait models that were developed separately from the DSM classification perspective. Presently, the DSM-IV does not have a specific model that accounts for personality traits, but does allow a clinician to record them on Axis-II (Krueger & Eaton, 2010). Trull and Durrett (2005) point out the advantages to this approach such as a greater understanding of PD constructs is possible, additional aetiological hypotheses concerning the development of PDs may be produced because much is already known about correlates of and factors underlying development of major personality traits and lastly, these models of personality may help identify and define varieties of personality pathology that are not represented currently in the DSM. The Five Factor Model (FFM) is one such model that has received the most clinical utility and empirical evidence thus far (Lynam & Widiger, 2001).
The FFM was originally developed using non-clinical samples and the aim was to provide a complete account of personality traits and dimensions (Trull & Durrett, 2005). Evidence for this model suggested that it could be applied to issues regarding psychopathology. Researchers found a relationship between the FFM and personality pathology. Furthermore, theorists have stated that psychopathology is best described dimensionally as extreme variations of normal personality traits and these traits are based on the FFM. Additionally, the FFM accounts for both normal and abnormal personality (Trull and Durret, 2005). The five domains: extraversion, agreeableness, conscientiousness, emotional instability (or neuroticism) and openness (or intellect, imagination or unconventiality) have been identified. With regards to generalisabity, lexical studies have been conducted in many other languages (including Turkish, Russian and Italian) and have provided further evidence for the existence of the five broad domains. Widiger and Trull (2007) provide evidence for the FFM from heritability, universality, childhood antecedents as well as temporal stability and use this evidence to justify the FFM as a potential dimensional candidate for the DSM-5.
The issue of temporal stability is a major issue concerning PDs, as the definition states explicitly that a PD is ‘enduring’ as well as ‘stable over time’. Traits have provided more evidence in that they do remain stable over time. Roberts and DelVecchio (2000) used 152 longitudinal studies where 3,217 test-retest correlation coefficients were compiled and the results demonstrated that trait consistency increased from .31 in childhood to .54 during the college years to, .64 at age 30, and then appeared to plateaux at .74 between the ages of 50 and 70 years old. From these results they concluded that traits do appear to be stable throughout one’s life.
Skodal (2005, p. 495) comments: “traits of general personality functioning (eg five factor traits) tend to be stable, with stability estimates in .70 to .80 range over two years”. The evidence to date does demonstrate that traits are found to be more stable than disorders (Skodal, 2010).
Shedler and Westen’s procedural assessment model (SWAP)
This entails quantifying each PD construct and a given score to indicate the degree to which the symptoms for each disorder are present. Scores represent the actual number of criteria present for each PD, or a rating which indicates the degree to which criteria are present for the disorder. Clinicians rate the overall similarity or “match” between a patient and the prototype using a 5-point scale (Westen, et al., 2010). Prototypes consist of paragraph descriptions of each disorder rather than a list of criteria, as in the DSM-IV. This system allows inclusion of richer diagnostic criteria and allows a clinician to organise the criteria in a way that would facilitate memory (Westen, et al., 2010). According to Westen and Shedler (2007), the SWAP has been used to refine and dimensionalise existing DSM-IV diagnostic categories and criteria, identify factors or trait dimensions relevant to describing personality pathology and empirically identify diagnostic groupings without de-valuing the existing DSM typology of PDs. This is in line with the suggestions that a dimensional approach cannot entirely replace the categorical approach and the SWAP generates both dimensions as well as categories (Fist, 2005; Verhel, 2005; Westen, et al., 2010).
Prototype ratings have been shown to have high inter-rater reliability with a median of .72 in non-psychotic patients seeking outpatient treatment. Rottman and colleagues (2010) found that clinicians made more correct diagnosis when using the SWAP as opposed to using the 30 FFM traits alone. It was suggested that traits alone are too ambiguous, but when given with a descriptive narrative it was much easier to make correct diagnoses. Further support was provided by Spitzer and colleagues (2008) who assessed the clinical utility of five dimensional models which included: the SWAP, FFM, Cloninger’s Psychobiological Model, criteria counting model based on current DSM-IV diagnostic criteria, and a prototype matching model based on current DSM-IV diagnostic criteria.
This study recruited a random sample of psychologists and psychiatrists who were asked to apply all five approaches to a patient under their care and rate the clinical utility of each model. The results demonstrated that the two prototype matching models (SWAP and prototype matching on current DSM-IV criteria) were more clinically useful and relevant as judged by the clinicians. The FFM was judged the least clinically useful. It was concluded by the authors that the prototype matching systems most “faithfully capture personality syndromes seen in practice and allow for rich descriptions without proportionate increase in time or effort” (Spitzer, et al., 2008 p. 356).
Westen, Shedler and Bradley (2006) state that when using this model, a clinician would match his or her perception of the patient’s maladaptive personality traits with a narrative description consisting of 15-20 sentences describing a prototypic case. Westen, Shedler and Bradley (2006) reported that this approach reduced co-morbidity among Cluster B PDs, predicted external validiators and was rated higher on measures of clinical utility than the corresponding DSM-IV PDs. Wood and colleagues (2007) stated that this co-morbidity reduction may be because the SWAP PD categories are more stringent than DSM PD categories and therefore fail to give any diagnosis to ambiguous cases. The only published study to examine this issue was by Westen and colleagues (2006) where it was found that co-morbidity and coverage were substantially lower for SWAP PD diagnosis than for corresponding DSM PD categories. It could be possible that the SWAP categories fail to identify a large number of PD cases and this should be thoroughly considered. With regards to inter-rater reliability, no studies thus far have been presented concerning this important issue. Additionally, no studies have examined the temporal stability of SWAP PD scores or diagnoses (Wood, et al., 2007).
A note on the proposed changes for DSM-5
The DSM-5 Work Group published their first paper which is the description as well as the rationale for the proposed changes (Skodal, et al., 2011, p. 5). The proposed changes identify core impairments in personality functioning, pathological personality traits and prominent pathological personality types. The assessment for PD may constitute four components:
1.five identified severity levels of personality functioning based on degrees of impairment in core self and interpersonal capacities;
2.five specific personality disorder types, each defined by impairments in core capacities and a set of pathological, personality traits and one trait specific type;
3.six broad higher order personality trait domains, with 4-10 lower order, more specific trait facets within each domain, for a total of 37 specific trait facets;
4.New general criteria for PD based on severe or extreme deficits in core capacities of personality functioning and extreme levels of pathological personality traits.
Skodal and colleagues (2011, p. 5) states that the “primary purpose of this projected assessment is to identify personality related problems and their severity; and to distinguish these problems according to broad, clinically salient types in association with patient-specific personality trait profiles.” The DSM-5 Work Group has proposed a measure of severity of impairment in core capacities that is central to personality functioning.
The proposal contains five specific PD types that are to be rated on a dimension of graded membership, these include: antisocial or psychopathic, avoidant, borderline, obsessive-compulsive and schizotypal. Each type is to be identified by core impairments in personality functioning and is associated with a trait list specifying its component pathological personality traits (Skodal, et al., 2011). The remaining PDs as well as the PDNOS will be given a diagnosis of PD trait-specified. This will be done by characterising core impairments combined with specification by individuals of a unique set of personality traits based on the most prominent descriptive features. The rationale for these changes are based on reducing excessive co-morbidity, arbitrary diagnostic thresholds, instability of the current DSM-IV-TR criteria sets as well as the limited validity for some of the existing PDs. According to Skodal and colleagues (2011), a reduction of five PDs is expected to reduce co-morbid PD diagnoses, and with the implementation of a dimensional rating of five types this will mean that personality pathology occurs along a continuum.
Anti-social, borderline and schizotypal PD have the most clinical utility as well as empirical evidence. The PDs that remain in the DSM-5 are retained based on the extent to which each PD has in terms of impairment in social functioning. For instance, schizotypal and BPD have been found to cause significantly more impairment in social situations, work and at leisure compared to patients with less severe types, such as OCPD (Skodal, et al., 2011). Additionally, OCPD and BPD are associated with the highest total economic burden in terms of medical costs. The five PDs will be conceptualised as a dimensional representation and will incorporate a prototype matching approach. This prototype dimensional model, SWAP, has been empirically derived by Westen and Shedler (2004). The Work Group propose a hybrid approach, consisting of both disorder and trait constructs, and a few studies have supported this approach (Skodal, et al., 2011).
With regards to the domains, the Work Group propose six broad, higher order personality trait domains which include negative emotionality, detachment, antagonism, disinhibition, compulsivity and schizotypy and each domain will consist of lower order more specific trait facets (Skodal, et al., 2011). The reasoning behind this is that a trait based system is intended to reduce the limitations in the current DSM-IV. A PD diagnostic system that is trait based will also provide a means to describe normal and abnormal personality of every patient. Furthermore, this will also limit the amount of PDNOS diagnoses as all patients will have a specific personality profile based on different traits. With regards to excessive within diagnosis heterogeneity, a trait based system will reflect the degree of similarity or difference between individuals, therefore, each person will not have such inconsistent behaviours for one diagnosis (Skodal, et al., 2011).
With regards to the schizotypal domain there has been controversy as to whether it should be included in the schizophrenic spectrum. However, research has outlined that it might belong in the PD domain as well. Rsearch has found that the schizotypal domain forms an important additional factor in analyses of both normal and abnormal personality (Skodal, et al., 2011). Therefore, it was decided by the Work Group to add this domain to the other four domains. With regards to the 6th domain, it was based on research that OCPD does not fit well onto any of the FFM domains, and that compulsivity would better represent this PD. Therefore, compulsivity was added to form the 6th domain. These six domains will be represented in the DSM-5.
The trait facets for each domain were based on a thorough review of existing measures of normal and abnormal personality as well as from recommendations by experts in this area. Before anything is published, the structural validity of the trait-based model will be tested in field trials.
Pincus (2011) states that the DSM-5 cites only one unpublished study to support a hybrid model which includes both types and traits. Though hybrid models are gaining credibility they lack empirical evidence, as well as clinical utility. Furthermore, Verheul (2005) found that purely dimensional models, such as SNAP, FFM and DAPP outperform the hybrid models in regard to coverage, consistency with aetiology and clinical decision making. Verheul (2005) does state that hybrid models would be easy to implement into the DSM-5 as little revision would be needed to the existing criteria, but the question is would these hybrid models even change anything. If dimensional models have gained the most empirical evidence to date and 18 models have been validated, the question remains as to why the Work Group would utilise a hybrid approach especially if it has only received a small amount of research to date.
As the current categorical approach presents with many limitations, there is clearly a need for a revision in the upcoming DSM. A dimensional model has accumulated an extensive array of evidence and is noted that it may eradicate the limitations of the categorical approach. More work on the prototype matching model needs to be done because there is some evidence this approach does not provide enough validity for the justification of implementation into the DSM-5 (Wood, et al., 2007).
Complexity in diagnosis is already well documented, for example, the differential diagnosis of Alzheimer’s disease and dementias (Thompson, 2003) and in post-traumatic stress disorder (Thompson, 2011). Whatever happens with the DSM-5, whether the Work Group implement a hybrid model or chose a different dimensional model, until future research can test it, clinicians will not know for certain if this has made an improvement on diagnosing patients with PD.


1. APA - American Psychiatric Association., 1980. Diagnostic and statistical manual of mental disorders. 3rd ed. Washington, DC: American Psychiatric Association.
2. APA - American Psychiatric Association, 1994. Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC: American Psychiatric Association.
3. APA - American Psychiatric Association, 2000. Diagnostic and statistical manual of mental disorders. 4th ed. Text revised. Washington, DC: American Psychiatric Association.
4. Besteiro-Gonzalez, J. L., Lemos-Giraldez, S., & Muniz, J., 2004.  Neuropsychological, Psycho physiological, and Personality Assessment of DSM-IV clusters of Personality disorders. European Journal of Psychological Assessment, 20(2), 99-105.
5. Blackburn, R., 2000. Treatment or incapacitation? Implications of research on personality disorders for the management of dangerous offenders. Legal & Criminal Psychology, 5, 1-21.
6. Blashfield, R. K., Intoccia, V., 2000. Growth of the literature on the topic of personality disorders. American Journal of Psychiatry, 157, 472-73.
7. Burgess, D. R., 2008. The conceptualization of personality disorder categories versus dimensions. Doctor of Philosophy Dissertation. Auburn: Auburn University.
8. Clark, L. A., 2007. Assessment and diagnosis of personality disorder: Perennial issues and an emerging re-conceptualization. Annual Review of Psychology, 58, 227-257.
9. Coid, J., Moran, P., Bebbington, P., Traolach, B., Jenkins, R., Farrell, M., Singleton, N., & Ullrich, S., 2009. The co-morbidity of personality disorder and clinical syndromes in prisoners. Criminal Behaviour and Mental Health, 19, 321-333.
10. Comer, R.J. 2011. Abnormal Psychology. 6th ed. New York: Worth Publishers.
11. Cooper, L. D., Balsis, S., & Zimmerman, M., 2010. Challenges associated with a polythetic system: criteria considerations in the personality disorders. Journal of Abnormal Psychology, 199(4), 886-895.
12. Eaton, N. R., 2010. Personality pathology in DSM-5. Psychological science is changing the way we think about mental disorder. publications/observer/2010/november-10/personality-pathology-in-dsm-5.html
13. First, M. B., 2005. Clinical utility: A prerequisite for the adoption of a dimensional approach in DSM. Journal of Abnormal Psychology, 114, 560–564.
14. Grant, B. F., Hasin, D. S., Fredrick, S. S., Dawson, A. D., Chou, P. S., Ruan, J. W., & Huang, B., 2005. Co-occurrence of 12 month mood and anxiety disorders and personality disorder in the US: results from the national epidemiologic survey on alcohol and related conditions. Journal of Psychiatric Research, 39, 1-9.
15. Krueger, R. F., & Eaton, N. R., 2010. Personality traits and the classification of mental disorders: toward a more complete integration in DSM-5 and an empirical model of psychopathology. American Psychiatric Association, 1(2), 97-118.
16. Livesley, W. J., Schroeder, L. M., Jackson, D. N., & Lang, K. L., 1994. Categorical Distinctions in the study of personality disorder: implications for classification. Journal of Abnormal Psychology, 103(1), 6-17.
17. Lynam, D. R., & Widiger, T. A., 2001. Using the five factor model to represent the DSM-IV personality disorders: an expert consensus approach. Journal of Abnormal Psychology, 110, 401-412.
18. Mullins-Sweatt, S, V, Smitt, R, Verheul, J. Oldham, T.A., Widiger., 2009. Dimensions of personality; clinicians perspectives. The Canadian Journal of Psychiatry, 54(4), 247-259.
19. Pincus, A. L., 2011. Some comments on nomology, diagnostic process, and narcissistic personality disorder in the DSM–5 proposal for personality and personality disorders. Personality Disorders: Theory, Research & Treatment, 2, 41–53.
20. Roberts, B. W., & DelVecchio, W. F., 2000. The rank-order consistency of personality traits from childhood to old age: A quantitative review of longitudinal studies, Psychological Bulletin, 126(1), 3-25.
21. Rottman, B. M., Kim, N. S., Ahn, W.K., Sanislow, C. A., 2010. Can Personality experts recognise DSM-IV personality disorders from five factor model descriptions of patient cases? Journal of Clinical Psychiatry, e1-e10.
22. Samuel, D. B., Hopwood, C. J., Ansell, E. B., Morey, L. C., Sanislow, C. A., Markowitz, J. C., Yen, S., Shea, T. M., Skodal, A. E., & Grilo, C.M., 2011. Comparing temporal stability of self-report interview assessed personality disorder. Journal of Abnormal Psychology, 1-11.
23. Samuel D. B., & Widiger T. A., 2006. Clinicians' judgments of clinical utility: A comparison of the DSM-IV and five factor models. Journal of Abnormal Psychology, 115, 298–308.
24. Sheets, E. & Craighead, E., 2007. Toward an empirically based classification of personality pathology. Clinical Psychology Science & Practice, 14, 77-93.
25. Skodal, A. E., 2010. Rationale for proposing five specific personality disorder types. http://               &nb sp;  [Accessed 18 Jun 2011].               &nbs p;                &n bsp;       26. Skodal, A, E.,  W.,Gunderson, J. G., Shea, T., McGlashan, T. H., Morey, L. C., Sanislow, C. A., Bender, D. S., Grilo, C. M., Zanarini, M. C., Yen, S., Pagano, M. E., & Stout, R. L., 2005 The collaborative longitudinal personality disorders (CLPS) overview and implications. Journal of Personality Disorder, 19(5), 487-504.
27. Spitzer, R. L., First, M. B., Shedler, J., Westen, D., & Skodal, A. E., 2008. Clinical utility of five dimensional systems for personality diagnosis: a “consumer preference” study. Journal of Nervous & Mental Disease, 196(5), 356-374.
28. Sprock, J., 2003. Dimensional versus categorical classification of prototypic and no prototypic cases of personality disorder. Journal of Clinical Psychology, 59, 991-1014.
29. Thompson, S.B.N., 1993. Eating disorders: a guide for health professionals. London: Chapman & Hall.
30. Thompson, S.B.N., 2003. Dementia and memory: a handbook for students and professionals. Aldershot: Ashgate.
31. Thompson, S.B.N., 2011. Psychology of trauma: clinical reviews, case histories, research. Portsmouth: Blackwell-Harvard-Academic.
32. Trull, T. J., & Durrett, C. A., 2005. Categorical and dimensional models of personality disorder. Annual Review of Clinical Psychology, 1, 355-380.
33. Tyrer, P., 2008. Personality disorder and public mental health. Clinical Medicine, 8, 423-427.
34. Tyrer, P., Coomb, N., Ibrahimi, f., Mathilakath, A., Bajaj, P., 2006. Critical developments in the assessment of personality disorders. British Journal of Psychiatry, s51-s59.
35. Verheul, R., 2005. Clinical utility for dimensional models of personality pathology. Journal of Personality Disorders, 19, 283–302.
36. Verheul, R., & Widiger, T. A., 2004. A meta-analysis of the prevalence and usage of the personality disorder not otherwise specified (PDNOS) diagnosis. Journal of Personality Disorders, 18(4), 309-319.
37. Westen, D., Defiffe, J. A., Bradley, B., & Hilsenroth, M. J., 2010. Prototype personality diagnosis in clinical practice: a viable alternative for DSM-5 and ICD-11. Professional Psychology: Research and Practice, 41(6), 482-487.
38. Westen, D., & Shedler, J., 2007. Personality diagnosis with the Shedler-Westen assessment procedure (SWAP): integrating and statistical measurement prediction. Journal of Abnormal Psychology, 116(4), 810-822.
39. Westen, D., Shedler, J., Bradley, R., 2006. A prototype approach to personality disorder diagnosis. American Journal of Psychiatry, 163, 846-856.
40. Widiger, T. A., 1993. The DSM-III-TR Categorical personality disorder diagnosis: a critique and an alternative. Psychological Inquiry, 4(2), 75-90.
41. Widiger, T. A., Livesley, J. W., & Clark, L. A., 2009. An integrative dimensional classification of personality disorder. Psychological Assessment, 21(3), 243-255.
42. Widiger T. A., Simonsen E., 2005. Alternative dimensional models of personality disorder: finding a common ground. Journal of Personality Disorders, 19, 110-130.
43. Widiger, T. A., Simonsen, E., Krueger, R., Livesley, J., W., 2008. Personality disorder research agenda for the DSM-V. Journal of Personality Disorder, 19(3), 315-338.
44. Widiger, T. A., & Trull, T. J., 2007. Plate tectonics in the classification of personality disorder: Shifting to a dimensional model. American Psychologist, 62, 71-83.
45. Wood, J. M., Garb, H. Koren, D., & Nezworski, M. T., 2007. The Shedler-Westen Assessment Procedure-200 as a basis for modifying DSM personality disorder categories. Journal of Abnormal Psychology, 116(4). 823-836.
46. Zimmerman, M. (2011). Is There Adequate Empirical Justification for Radically Revising the Personality Disorders Section for DSM-5? Personality Disorders: Theory, Research & Treatment, 1-14.

Source(s) of Funding


Competing Interests



This article has been downloaded from WebmedCentral. With our unique author driven post publication peer review, contents posted on this web portal do not undergo any prepublication peer or editorial review. It is completely the responsibility of the authors to ensure not only scientific and ethical standards of the manuscript but also its grammatical accuracy. Authors must ensure that they obtain all the necessary permissions before submitting any information that requires obtaining a consent or approval from a third party. Authors should also ensure not to submit any information which they do not have the copyright of or of which they have transferred the copyrights to a third party.
Contents on WebmedCentral are purely for biomedical researchers and scientists. They are not meant to cater to the needs of an individual patient. The web portal or any content(s) therein is neither designed to support, nor replace, the relationship that exists between a patient/site visitor and his/her physician. Your use of the WebmedCentral site and its contents is entirely at your own risk. We do not take any responsibility for any harm that you may suffer or inflict on a third person by following the contents of this website.

5 reviews posted so far

Categorizing Personality Disorders.
Posted by Ms. Holly Chinnery on 08 Jan 2012 04:44:09 PM GMT

Categorising personality disorder - review
Posted by Ms. Natalie Jones on 08 Jan 2012 12:36:18 PM GMT

Categorising personality disorder...
Posted by Mr. Brian Thompson on 23 Dec 2011 06:18:02 PM GMT

0 comments posted so far

Please use this functionality to flag objectionable, inappropriate, inaccurate, and offensive content to WebmedCentral Team and the authors.


Author Comments
0 comments posted so far


What is article Popularity?

Article popularity is calculated by considering the scores: age of the article
Popularity = (P - 1) / (T + 2)^1.5
P : points is the sum of individual scores, which includes article Views, Downloads, Reviews, Comments and their weightage

Scores   Weightage
Views Points X 1
Download Points X 2
Comment Points X 5
Review Points X 10
Points= sum(Views Points + Download Points + Comment Points + Review Points)
T : time since submission in hours.
P is subtracted by 1 to negate submitter's vote.
Age factor is (time since submission in hours plus two) to the power of 1.5.factor.

How Article Quality Works?

For each article Authors/Readers, Reviewers and WMC Editors can review/rate the articles. These ratings are used to determine Feedback Scores.

In most cases, article receive ratings in the range of 0 to 10. We calculate average of all the ratings and consider it as article quality.

Quality=Average(Authors/Readers Ratings + Reviewers Ratings + WMC Editor Ratings)