Part II
5
Joel Paris
How Personality Disorders Ended Up with Three Different Diagnostic Systems
Researchers studying personality disorders (PDs) are currently facing a situation in which there are three separate methods of classifying these disorders: two in DSM-5 (American Psychiatric Association, 2013) and one in ICD-11 (Tyrer, Crawford, Mulder, & Blashfield, 2011). If one were to consider the Research Domain Criteria (RDoC; Insel et al., 2010) as another alternative, there would be four.
This dilemma reflects the problem of defining complex psychopathological constructs that describe symptoms, traits, and psychosocial dysfunction. It also reflects longstanding disagreements as to whether PDs can be considered categories in much the same way as medical illnesses, or whether they reflect extreme and pathological variants of normal traits that can be scored dimensionally (Widiger, 2007). Behind this problem lies still another question: is there any real difference, other than levels of functioning, between personality traits and PDs?
Most experts have agreed that the system used in DSM-IV, describing ten categories of disorders (as well as a “not otherwise specified” diagnosis), was unsatisfactory. However, they could not agree on how to replace it. Discussions of how best to proceed have focused on three main questions. The first is whether there is enough data to support a different set of categories. The second is whether we should define PDs as endpoints on a continuous spectrum describing personality trait variations. A third is whether a hybrid model that combines both approaches would be preferable.
These questions have long divided the research community. Traditionally, psychiatrists, reflecting their medical background, have tended to prefer a categorical classification, and clinical psychologists have generally followed their lead (Jablensky, 2016). However, over the past few decades, an influential group of researchers trained in trait psychology have begun to argue that categories are artificial, emphasizing the absence of any sharp boundaries between trait variations and diagnosable disorders (Widiger, 2007). This view is supported by a large body of research, both in community and clinical samples (Krueger and Bezdjian, 2009). Support for the continuity of traits and disorders also comes from a large body of behavioral genetic research (Livesley, Jang, & Vernon, 1998). Nonetheless, it remains unclear as to whether trait dimensions can account for the prominent symptoms seen in some PDs, especially the borderline category (Herpertz et al., 2017).
Research on PD diagnoses has been active for several decades. As of 2017, Medline lists over 11,000 articles and PsychInfo lists 19,000 articles. Three diagnostic categories tend to dominate the literature. For borderline PD, 5800 articles are listed on Medline and 10,000 are listed on PsychInfo; for antisocial PD, both Medline and PsychInfo list over 5000 articles. (When psychopathy is added to antisocial, Medline lists over 9000, whereas PsychInfo lists over 22,000, indicating that research on these traits is more frequent than research on the related PD diagnosis.) Schizotypal PD is third in frequency, with 2400 papers on Medline, and 3900 on PsychInfo. However, many of these publications are clinical reports, and only about half are retained if one limits the search using the term “research.”
Thus, most PD research has focused on the categories of borderline PD and antisocial PD. Such a large body of research is usually based on the assumption, supported by the structure of the DSM, that the categories of borderline and antisocial PD are entities as coherent as major depression or schizophrenia. However, even the most well-established diagnoses in psychiatry lack precise boundaries. For example, it is difficult to separate sadness from depression (Horwitz & Wakefield, 2007) or alcohol use from abuse (McArdle, 2008). This issue is particularly relevant for antisocial and borderline PD categories. Yet, despite this challenge, it is equally true that these two diagnostic categories are as readily recognizable to clinicians as most other diagnoses in current use. Thus, in these ways, the concerns specific to differentiation of categories seem to be largely a focus in the research literature and far less of a concern among clinicians.
The DSM Categorical System
PDs, categorically defined, are listed in all diagnostic systems of mental disorders. In the DSM system, the most major changes came with the 3rd edition (American Psychiatric Association, 1980), which included, for the first time, borderline PD and narcissistic PD.
One problem was that the original DSM system had no formal description of what defines a PD. This definition only came with DSM-IV (American Psychiatric Association, 1994), which introduced, for the first time, overall criteria for diagnosing a PD. The definition focuses on an enduring pattern of inner experience that affects cognition, mood, interpersonal functioning, and impulsivity.
The ten-category system of diagnosis listed in DSM-IV has been retained in Section II of DSM-5. However, a serious problem lies with the fact that, as Zimmerman and colleagues showed (Zimmerman, Rothschild, & Chelminski, 2005), about half of patients who meet overall criteria for a PD do not fit into any of these categories. Thus, about half of all patients with PDs end up being diagnosed as “not otherwise specified” (NOS), or, in the terminology of DSM-5, “unspecified.” This lack of precision is a serious problem for the categorical system, and suggests that a different method of classification is required.
A second problem with the ten DSM categories is that there has been almost no research on most of them. Although obsessive-compulsive and avoidant PDs were included in the Collaborative Longitudinal Personality Disorders Study (CLPS; Gunderson et al., 2011), they have not been studied in detail, and there is even less research on schizoid, paranoid, and dependent PDs. The large database on borderline and antisocial PDs is a striking exception, and helps to explain the resistance of researchers interested in these conditions to replacing them with dimensional scores. Given that abandoning these categories would make thousands of past research studies irrelevant, it might make more sense (or, at least, would be least disruptive) to reduce the number of categories by eliminating those that lack an evidence base.
PDs are most likely to attract research when they create serious clinical problems. In borderline PD, patients frequently present to emergency rooms and display clinically important symptoms. In antisocial PD, research has been stimulated by the forensic system and by social concerns about containing criminal behavior. The other eight PD categories in DSM-5 could probably be removed without dramatic impact on research or clinical practice.
Another approach would be to make use of the three “clusters” described in DSM-5: Cluster A for “odd” PDs: Cluster B for dysregulated PDs, and Cluster C for anxious PDs (Zimmerman et al., 2005). But yet again, there is a dearth of research specifically focusing on these constructs.
Turning to the specific categories in Cluster A, schizotypal PD has a relatively small research literature, although one recent review quoted 79 publications (Rossell, Futterman, McMaster, & Siever, 2014). Many follow the ICD system in viewing this condition as better placed in the schizophrenic spectrum. Research on the related categories of schizoid and paranoid PDs is rather scant, and some research reports have combined all three diagnoses together (Esterberg, Goulding, & Walker, 2010). Given that one of the earlier iterations of the DSM-5 system proposed doing just that, combining these categories might be reconsidered.
Turning to Cluster B, histrionic PD is mainly of historical interest in that it derives from the concept of “hysteria” (Novalis, Araujo, & Godinho, 2015). On the other hand, the relatively new diagnosis of narcissistic PD has recently been enriched by a body of research on narcissistic traits (Campbell & Miller, 2011), and although the full disorder has not been well researched, the construct of pathological narcissism seems to be clinically useful.
In Cluster C, research on avoidant PD is thin, although the CLPS study (Gunderson et al., 2011) found that like other PDs, one tends to observe improvements over time. Also, avoidant PD overlaps with generalized social anxiety in both phenomenology and genetic risk (Reichborn-Kjennerud et al., 2007). It is therefore possible that these conditions might benefit from the same treatment methods. Much of the work on dependent PD has come from one research group (Beitz & Bornstein, 2010). Very little has been published on obsessive-compulsive PD, other than the CLPS study, which found some degree of improvement over time.
Findings that PD categories tend to overlap, particularly within clusters, raise the question as to whether we actually need ten of them. This argument, made several decades ago by Tyrer and Alexander (1979), proposed that four categories (i.e., sociopathic, passive-dependent, anankastic, and schizoid) could be sufficient. However, as problems with most categories proved difficult to resolve, Tyrer et al. (2011) recommended dealing with overlap by replacing them with ratings on five trait domains (i.e., asocial, emotionally unstable, obsessional [anankastic], anxious/dependent, and dissocial).
Up to now, the International Classification of Diseases, 10th edition (World Health Organization, 1992) has also used a categorical system for the diagnosis of PD that is in many respects similar to the DSM. But, as the date for publication of ICD-11 approaches, this system may be abandoned (or at least greatly modified) to allow for dimensional assessment of traits and disorders.
Dimensional Systems
Trait psychologists, schooled in psychometrics, study personality by developing self-report questionnaires that are reliable, have external validity, and can be scored dimensionally. This approach to measuring psychopathology does not apply only to PDs. In many diagnoses, there is no sharp demarcation between traits and disorders, as is clearly the case for depression (Angst and Merikangas, 1997; Horwitz & Wakefield, 2007) and anxiety (Lang and McTeague, 2009; Horwitz & Wakefield, 2012). Even in psychosis, one sees patients who fall in a space between diagnosable illness and trait variation (Potuzak, Ravichandran, Lewandowski, Ongur & Cohen, 2014). This approach to the measurement of psychopathology goes back several decades. Moreover, neurobiological studies of mental disorders show that biomarkers tend to correlate with dimensional measurements, but only rarely with categorical diagnoses (Caspi, Houts, Belsky, & Moffitt, 2014).
These findings influenced the editors of DSM-5 to hope, at least initially, to make all diagnosis in psychiatry dimensional (Kupfer & Regier, 2011). Because they saw PDs as the poster child for this change, Kupfer and Regier instructed the committee in charge of PD classification to develop a dimensional system.
This approach could draw on a rich literature in trait psychology that has been studying dimensional approaches to personality and psychopathology for decades. Hans Eysenck (1967), a pioneer in this field, attempted to describe all trait variations and mental disorders as admixtures of two broad factors: Neuroticism and Extraversion – later adding a third he called “Psychoticism” (which would be better labeled as impulsivity).
Building on the first two of these dimensions, the most intensively researched trait system, the Five-Factor Model (FFM), was developed (Costa & Widiger, 2013). Its broad domains (Neuroticism, Extraversion, Agreeableness, Conscientiousness, and Openness to Experience) have been the basis of thousands of research studies. The FFM has been supported by a majority of trait psychologists, has been shown to have cross-cultural validity, and helps to account for variance in personality traits (Costa & Widiger, 2013). That is why the FFM became by far the most frequently used system of its kind.
One might have thought that the American Psychiatric Association would have adopted the well-researched FFM to describe PDs and their underlying traits. However, they did not. One reason may be that in contrast to the usual method of psychiatric assessment, which relies on ratings by expert clinicians, the FFM is based on scores from self-report questionnaires. Another reason could be that the FFM, validated in normal populations, was not considered to do as good a job with psychopathology as with normal variation. Thus, the DSM-5 committee on PDs ended up striking out in a new direction and developing a new dimensional system.
Personality Disorders in DSM-5
An oncologist I know was very surprised when I told him that the final form of DSM-5 was determined, not by generally agreed upon facts, but by the vote of a committee. But voting may be the only way to proceed when the facts are unclear or unavailable. Many scientific domains have struggled with the problem of fuzzy categories. Biologists have argued for decades about how best to classify species. Even the astronomical community was required to vote on whether Pluto should remain a planet or be demoted to a dwarf planet.
The problem is that a decision of a scientific committee can be determined by the opinions and biases of its members. For this reason, the DSM-5 committee on PD diagnosis consisted of a mixture of experts who favored categories and those who favored dimensions. Its chair, Andrew Skodol (then at Columbia) had been one of the researchers in the CLPS study, but was receptive to a directive to produce a dimensional system.
The initial proposals of the committee were published on the web for commentary. The result was a compromise between the two camps. The group developed a “hybrid system” in which some categories were retained, but were built entirely from trait profiles (American Psychiatric Association, 2013). Thus, each category is not, as in previous editions, diagnosed using a list of criteria that can only be rated as present or absent, but is derived from a characteristic pattern of traits that can be scored. The original draft proposed to reduce the number of categories from the ten in DSM-IV to five: antisocial/psychopathic, avoidant, borderline, obsessive-compulsive, and schizotypal. Not coincidentally, these were also the five categories studied in CLPS.
As drafts of the DSM-5 proposals were published on the internet, the system went through several iterations. Originally, there were five categorical diagnoses, but at one point a sixth, narcissistic PD, was added. This change was due to the views of a strong lobby of investigators who did not want to see the diagnosis disappear. Thus, narcissistic PD remained in the magic circle of categories that could be constructed from trait profiles.
These results were the outcome of a number of political compromises that led to a split in the committee. Two members (John Livesley of Canada and Roel Verheul of the Netherlands), who had hoped for a fully dimensional system, considered the hybrid model to be incoherent and resigned from the committee in protest.
As the deadline for publication of DSM-5 approached in 2013, a decision was needed. In the end, the American Psychiatric Association determined that the proposed changes were too radical to make with a relatively small evidence base (Silk, 2016). Once again by vote, the Board of Trustees moved the hybrid system to Section III of the manual, containing diagnostic constructs that require more research. It has since been described as an “alternative” system for diagnosis.
One might have thought that the decision of the APA not to adopt the hybrid system would have made its eventual success less likely, which is what I thought at the time (Paris, 2013), but that is not what happened. Instead, its advocates considered the decision to be only a temporary setback that could be overcome by publishing more research. Under the leadership of the University of Minnesota psychologist Robert Krueger, co-editor of the Journal of Personality Disorders and a member of the DSM-5 committee, over 300 research studies using the alternative system were published between 2013 and 2017. Most of these studies were carried out by trait psychologists who have a strong investment in making PD diagnoses dimensional (Krueger & Markon, 2014). The goal of their program is to remove the old PD categories and replace them entirely with the alternative system (often referred to in these research publications by its supporters as the “DSM-5 system”).
DSM-5 was given an Arabic rather than a Roman numeral so that it could be revised, as DSM-5.1 and 5.2 without having to wait another 19 years for the next manual. This expectation could be viewed as either visionary or overly optimistic. In the past, some researchers have been unhappy when changes in the DSM came too soon, requiring them to use new constructs and new instruments.
The five trait domains in the model (negative affectivity, detachment, antagonism, disinhibition vs. compulsivity, and psychoticism) have a resemblance to previous schemata in trait psychology, particularly the FFM. However, this system, when applied to clinical practice, is not scored by self-report but by clinical ratings. This raises the question of whether clinicians, even when trained to rate personality traits, can produce scores that are reliable and that have external validity.
In the alternative model, all PD patients are rated on five personality trait domains. The first four (negative affectivity, detachment, antagonism, disinhibition vs. compulsivity), resemble the FFM. The fifth domain, psychoticism, does not appear in most trait models because it describes problems that are less common in community populations. (Unlike the similar term used by Eysenck, this domain describes a tendency to have psychotic symptoms.) However, it was considered necessary to use this construct to describe PD patients, some of whom have quasi-psychotic experiences.
The alternative system offers a more precise definition of PDs than was found in DSM-IV, focusing on pathology affecting self (identity and self-direction) and interpersonal relations (empathy or intimacy). Ratings involve a series of stages and procedures: (1) Is impairment in personality functioning present? (2) If present, rate the level of impairment in self and interpersonal functioning. (3) Is one of the six defined types present? (4) If so, record the type and the severity of impairment. (5) If not, is the PD trait specified or unspecified? (6) One can record the PD trait specified (PDTS), identify and list the trait domain(s) that are applicable, and record the severity of impairment. (7) If a PD is present and a detailed personality profile is desired, one can evaluate the trait facets. (8) If neither a specific PD type nor PDTS is present, the PD is unspecified, but one can evaluate trait domains and/or trait facets. At each of these steps, clinicians are asked to use a Likert scale (from 1 to 5) for scoring.
Each of these ratings requires judgment calls to determine what is normal, what is extreme, and what is truly dysfunctional. One might ask whether busy clinicians can carry out such a demanding procedure. Research has already shown that most clinicians ignore the precise instructions of the DSM system for diagnosis (First, Bhat, Adler, Dixon, & Goldman, 2014), and that these ratings reflect a global impressionistic opinion.
The work of the committee also produced a self-report measure that can be used to aid diagnosis. This is the Personality Inventory for PID-5 (Krueger, Derringer, Markon, Watson, & Skodol, 2012), which can assess the five trait domains and 25 maladaptive personality trait facets of the DSM-5 alternative system. This 220-item inventory has also been abbreviated to 100 items (PID-5-BF; Al-Dajani, Tralnick, & Bagby, 2016; Maples et al., 2015). It is offered for clinical use in monitoring change in patients, but will probably be mainly used as a research instrument.
Let us summarize the current impact of the alternative DSM-5 system. At this point, the model is shaping the current direction of research, but is not yet being used in clinical practice. The clinical community remains more comfortable with categories, which are a strong tradition in medicine and clinical psychology. Also, even if the alternative model is fully adopted in future versions of DSM-5, it will take time to be accepted. Krueger and Markon (2014) blame conservative political forces for this situation, seeing the categorical model as unscientific and the dimensional model as progressive and empirically based. This view has also been supported by Zachar and First (2015), who see the transition to what might be included in “DSM-5.1” as just a matter of time.
Few will disagree that the categorical model used in previous editions of DSM is outdated. What we do not know is how a drastic change in the conceptualization and assessment of PDs would affect clinical practice. We know from research that PDs are often misdiagnosed or entirely ignored (Zimmerman et al., 2005). Putting them in Axis II, as was done in DSM-III and DSM-IV, only made the situation worse, as clinicians were likely to write “Axis II, deferred” in their assessments of patients, allowing them to focus on the disorders they thought they knew how to treat. What would be the effect if PDs were the only diagnoses in the manual to be dimensionalized? Might this make ignoring personality pathology easier? And, if so, would patients with PDs suffer from not being assessed adequately? We just do not know. Finally, since patients these days expect a diagnosis and often look these up on the web, how would this system affect patient education and compliance with treatment?
The International Classification of Diseases (ICD) System
The DSM system is not the official classification of mental disorders. By treaty, almost all countries accept the International Classification of Diseases (ICD), published by the World Health Organization (1992), as the standard. The coding of disorders in DSM always allows for direct translation into an ICD equivalent. However, whereas DSM predominates in North America, in other parts of the world, the ICD system is used.
Previous editions of this manual, such as ICD-10 (World Health Organization, 1992), described categories generally similar to those in the DSM, with a few exceptions. Specifically, borderline PD was a subcategory of “emotionally unstable” PD, and narcissistic PD was never included as a diagnosis.
While controversies about the DSM system were getting international press coverage, WHO quietly and systematically prepared its own revised classification, the ICD-11. The preliminary plans for this system have been published on the web. Once again, PDs are a test case for adopting a dimensional system of diagnosis. This project was led by the British psychiatrist Peter Tyrer, who has also been a co-author of most of the preparatory research.
The proposed system for ICD-11 (Tyrer et al., 2011) asks clinicians to rate personality dysfunction on a five-point scale (none, difficulty only, mild, moderate, severe). There are no PD categories, but the manual instructs clinicians to rate patients on five trait domains (asocial, emotionally unstable, obsessional [anankastic], anxious/dependent, and dissocial). This procedure is much simpler than the alternative DSM-5 model, as there are fewer decision points. It is not known whether these five domains closely track the five domains of the DSM-5 alternative model. Further, although the procedures are much simpler than DSM-5’s alternative system, it is not known whether busy clinicians can be trained to make reliable ratings.
The most radical aspect of the ICD-11 proposal, and a point of departure from both the DSM-5 alternative system and the ICD system as a whole, is to eliminate all categorical diagnoses of PDs. One advantage of making diagnoses on the basis of a trait profile is that so many cases do not fit any of the existing categories. On the other hand, one can question the logic of dimensionalizing PDs while leaving other potentially dimensional constructs, such as anxiety and depression, as a set of categories. And, again, one can question whether the forensic community would accept the complete disappearance of antisocial personality and psychopathy in favor of a high score on a dissocial domain, and whether clinicians (or researchers) would be satisfied giving patients previously diagnosed with BPD high scores on scales of negative emotions and disinhibition. Since any revolution in diagnostic systems can be problematic, one needs to ask what strength of evidence is required to justify drastic change. At this writing, a proposal is under negotiation to produce a compromise that would not eliminate the best-researched categories (Herpertz et al., 2017).
Diagnosing Borderline Personality Disorder
BPD is common in the community, affecting somewhat less than 1 percent of the population (Lenzenweger, Lane, Loranger, & Kessler, 2007). It is also the most familiar category of PD in clinical practice; Zimmerman et al. (2005) found that about 5 percent of all outpatients meet criteria for this disorder. Like all PDs, BPD is associated with an abnormal sense of self and problems in interpersonal relations. However, unlike other PD categories, it has a highly symptomatic clinical presentation. Thus, BPD is not a classically egosyntonic PD, but is associated with a wide range of egodystonic symptoms (Zanarini et al., 1998), including chronically low and/or unstable mood, a range of impulsive behaviors, and micropsychotic symptoms.
These features help to explain why the diagnosis of BPD is often missed in practice. Zimmerman and Mattia (1999) found that only half of patients meeting DSM criteria for this disorder are recognized in outpatient clinics. Often, clinicians focus on abnormalities of mood, leading them to prescribe antidepressants. It is easy to view these patients as suffering from major depression if one ignores the mood swings that characterize BPD (Gunderson & Phillips, 1991). This is an example of why it is important to recognize BPD. Both Cochrane (Binks et al., 2006) and the National Institute for Health and Care Excellence (NICE, 2009) concluded from meta-analyses that antidepressants are of little value in this clinical population.
It is also very common to see BPD patients who have been diagnosed by previous clinicians as having bipolar disorder. This is due to the current popularity of the concept of a bipolar spectrum that could be used to account for the mood symptoms seen in BPD (Paris, 2012). However, few patients with BPD ever have hypomanic episodes (Paris, Gunderson, & Weinberg, 2007). Instead, they have mood instability with prominent anger and shifts in mood that usually last a few hours and are highly sensitive to environmental adversities. The pattern of daily (or hourly) mood instability does not support viewing BPD as a sub-clinical form of bipolarity, nor does data drawn from family history, biological markers, or outcome (Paris et al., 2007). Unfortunately, when patients with BPD are misdiagnosed as having bipolar disorder, they can be prescribed lithium or anticonvulsant mood stabilizers – neither of which has been shown to be effective for this population (Binks et al., 2006; National Institute for Health and Care Excellence, 2009).
Another diagnostic issue in BPD is the separation from psychotic disorders. It is not widely known that about half of all patients with BPD experience auditory hallucinations at some time in the course of their illness (Schroeder, Fisher, & Schafer, 2013). These symptoms are usually related to stress and emotion dysregulation, and when they have these experiences, patients almost always recognize that the voices they have heard are imaginary. One also sees brief psychotic episodes in BPD patients that can require psychopharmacological intervention (Zanarini et al., 1998).
Another source of misdiagnosis in BPD occurs when clinicians focus on problems with attention and consider them to suffer from attention deficit hyperactivity disorder (ADHD) (Paris, Bhat, & Thombs, 2015). However, to diagnose ADHD in adults, one needs a childhood history of these symptoms. Moreover, difficulties with attention can be associated with a wide range of disorders. Olfson and colleagues (Olfson, Blanco, & Greenhill, 2013) have shown that the prescription of stimulants in office practice to patients of all kinds has become much more frequent in recent years. Further, even if patients with BPD report benefit from taking these agents, it should be kept in mind that they also often support a focus on tasks in people without psychopathology, and there is no evidence that they have specific value in the treatment of BPD (Binks et al., 2006).
Given that BPD is associated with symptoms that overlap with many other mental disorders, one often sees misdiagnosis. This would not be a problem if alternative models still generated a similar category. But if we were to remove the category of BPD entirely, and replace it with a trait profile, patients might not benefit from the body of research demonstrating that there are specific therapies that work in this population (Zanarini, 2009). Thus, eliminating the diagnosis of BPD could have real consequences for effective management.
Diagnosing Antisocial PD and Psychopathy
Mental health clinicians do not often see patients with antisocial PD, who comprise only 3–4 percent of the PDs seen in practice (Zimmerman et al., 2005). However, almost 50 percent of prisoners meet criteria for this disorder (Fazel & Danesh, 2002), making it one of the most common in forensic settings. Antisocial PD is one of the few categories in the DSM that requires a childhood onset (in this case, of conduct disorder). And, because more than a quarter of children with conduct disorder will develop antisocial PD in adulthood (Zoccolillo, Pickles, Quinton, & Rutter, 1992), many of these patients will have been evaluated by child psychiatrists or psychologists.
Antisocial PD has been described in all editions of the DSM, and was also in previous editions of the ICD, using the label dissocial PD. There is an older literature describing a related construct of psychopathy, which describes patients with a lack of anxiety or fear and a bold interpersonal style (Hare & Neumann, 2008). Some experts consider psychopathy to be a more severe form of ASPD (Coid & Ullrich, 2010). Others see it as a different diagnosis entirely (Hart & Hare, 1996). Although psychopathy was at one point considered to be a variant of antisocial PD in preparing the DSM-5, it is not included as a specifier in the alternative model of PDs.
Although there is little evidence for effective treatment of antisocial PD, it is still important to make the diagnosis. Since this condition is marked by impulsive and dysregulated behaviors, some of these patients could be considered to have bipolar disorder, and receive medications designed for bipolarity.
PDs, Personality, and Normality
The diagnosis of PDs in DSM-5 and ICD-11 requires clinical judgment. At what point are we seeing variations in personality traits that lie within the limits of normality versus a disorder that can clearly be viewed as pathological? If trait psychologists are correct in seeing disorders as exaggerated traits, then the cutoff point must be arbitrary.
In fact, all categories of mental disorder have unclear boundaries (Frances, 2013). Thus, diagnostic constructs are subject to “concept creep” (Haslam, 2016), in which they tend to expand over time. One good example concerns narcissistic PD, for which some researchers have described another type of pathology called “vulnerable narcissism” (Dickinson & Pincus, 2003) to describe patients who are not grandiose but highly sensitive to criticism.
In a critique of overly expansive diagnostic constructs, Frances (2013) pointed out that mental disorders cannot be diagnosed without demonstrating functional impairment and/or distress. Thus, no matter how striking a trait profile is in any patient, they should not receive a PD diagnosis unless these characteristics are producing distress or a notable loss of functioning.
What Is Needed to Create a Valid Diagnostic System for PDs?
Much ink has been spilled in the categorical–dimensional debate, giving the impression to outsiders that PD researchers are a fractious lot. However, both ways of diagnosing these disorders have advantages and disadvantages. At this point, we just do not know enough to resolve the issue. Without a more detailed understanding of the etiology of PDs, classifying them can only be provisional.
The current controversies about diagnosing PDs are reminiscent of the classical tale of the blind men and the elephant. DSM-5 and other diagnostic systems face similar problems in determining the boundaries of other disorders whose causes remain unclear, including schizophrenia, bipolar disorder, depression, and substance dependence (Paris, 2013). And, whereas the RDoC system has the ultimate goal of developing an etiologically based classification, it is many decades from that goal. Moreover, reducing complex mental phenomena to brain circuitry may turn out to be an impossible mission (Paris & Kirmayer, 2016).
When we know more about why people develop PDs, we will be in a better position to develop a scientific classification. Until then, it is probably best to be pragmatic. As Zimmerman et al. (2005) have shown, PDs, in spite of their high clinical prevalence, are being missed or ignored by many clinicians. At this point, it could be a priority to settle on a single system that, even if flawed, is simple enough to be user-friendly in practice.
Conclusions
In summary, the following conclusions seem warranted:
1.The categorical system in DSM-5 is poorly validated, but describes at least two categories of great clinical significance (borderline and antisocial PDs).
2.Dimensional systems proposed for the classification of PDs can be validated by their relationship to trait profiles, but have uncertain clinical utility.
3.The alternative (hybrid) system proposed in Section III of DSM-5 is an attempt to combine the benefits of the categorical and dimensions approaches, but is currently too complex for routine clinical use.
4.The question of which system for the diagnosis of PDs is best remains unsettled, as there are advantages and disadvantages associated with all options.
5.The problem could be eventually illuminated by a better understanding of the etiology of PDs.
References
Al-Dajani, N., Tralnick, T. M., & Bagby, R. M. (2016). A psychometric review of the Personality Inventory for DSM-5 (PID-5): Current status and future directions. Journal of Personality Assessment, 98, 62–81.
American Psychiatric Association. (1980). Diagnostic and Statistical Manual of Mental Disorders (3rd ed.). Washington, DC: American Psychiatric Association.
American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders (4th ed.). Washington, DC: American Psychiatric Association.
American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.
Angst, J., & Merikangas, K. (1997). The depressive spectrum: Diagnostic classification and course. Journal of Affective Disorders, 45, 31–39.
Beitz, K., & Bornstein, R. F. (2010). Dependent personality disorder. In J. F. Fisher & W. T. Dononhue (Eds.), Practitioner’s Guide to Evidence Based Psychotherapy (pp. 230–237). New York: Springer.
Binks, C. A., Fenton, M., McCarthy, L., Lee, T., Adams, C. E., & Duggan, C. (2006). Psychological therapies for people with borderline personality disorder. Cochrane Database of Systematic Reviews [Article CD005652].
Campbell, W. K., & Miller, J. D. (Eds.) (2011). Handbook of Narcissism and Narcissistic Personality Disorder. New York: Wiley.
Caspi, A., Houts, R., Belsky, D. W., & Moffitt, T. E. (2014). The p factor: One general psychopathology factor in the structure of psychiatric disorders? Clinical Psychological Science, 2, 119–137.
Coid, J., & Ullrich, S. (2010). Antisocial personality disorder is on a continuum with psychopathy. Comprehensive Psychiatry, 51, 426–433.
Costa, P. T., & Widiger, T. A. (Eds.) (2013). Personality Disorders and the Five Factor Model of Personality (3rd ed.). Washington, DC: American Psychological Association.
Dickinson, K. A., & Pincus, A. L. (2003). Interpersonal analysis of grandiose and vulnerable narcissism. Journal of Personality Disorders, 17, 188–207.
Esterberg, M. A., Goulding, S. M., & Walker, E. F. (2010). Cluster A personality disorders: Schizotypal, schizoid and paranoid personality disorders in childhood and adolescence. Journal of Psychopathology and Behavioral Assessment, 32, 515–528.
Eysenck, H. J. (1967). The Biological Basis of Personality. Springfield, IL: Charles C. Thomas.
Fazel, S., & Danesh, J. (2002). Serious mental disorder in 23000 prisoners: A systematic review of 62 surveys. Lancet, 359(9306), 545–550.
First, M. B., Bhat, V., Adler, D., Dixon, L., & Goldman, B. (2014). How do clinicians actually use the Diagnostic and Statistical Manual of Mental Disorders in clinical practice and why we need to know more? Journal of Nervous and Mental Disease, 202, 841–844.
Frances, A. (2013). Saving Normal. New York: HarperCollins.
Gunderson, J. G., & Phillips, K. A. (1991). A current view of the interface between borderline personality disorder and depression. American Journal of Psychiatry, 148, 967–975.
Gunderson, J. G., Stout, R. L., McGlashan, T. H., Shea, T., Morey, L. C., Grilo, C. M., … Skodol, A. E. (2011). Ten-year course of borderline personality disorder: Psychopathology and function from the Collaborative Longitudinal Personality Disorders Study. Archives of General Psychiatry, 68, 827–837.
Hare, R. D., & Neumann, C. S. (2008). Psychopathy as a clinical and empirical construct. Annual Review of Clinical Psychology, 4, 217–246.
Hart, S. D., & Hare, R. D. (1996). Psychopathy and antisocial personality disorder. Current Opinion in Psychiatry, 9, 129–132.
Haslam, N. (2016). Concept creep: Psychology’s expanding concepts of harm and pathology. Psychological Inquiry, 27, 1–17.
Herpertz, S. C., Huuprich, S. K., Bohus, M., Chanen, A., Goodman, M., Mehlum, L., … Sharp, C. (2017). The challenge of transforming the diagnostic system of personality disorders. Journal of Personality Disorders, 31, 577–589.
Horwitz, A. V., & Wakefield, J. C. (2007). The Loss of Sadness: How Psychiatry Transformed Normal Sorrow into Depressive Disorder. New York: Oxford University Press.
Horwitz, A. V., & Wakefield, J. C. (2012). All We Have to Fear: Psychiatry’s Transformation of Natural Anxieties into Mental Disorders. New York: Oxford University Press.
Insel, T. R., Cuthbert, B., Garvey, M., Heinssen, R., Pine, D. S., Quinn, K., … Wang, P. (2010). Research domain criteria (RDoC): Toward a new classification framework for research on mental disorders. American Journal of Psychiatry, 167, 748–751.
Jablensky, A. (2016). Psychiatric classifications: Validity and utility. World Psychiatry, 15, 26–31.
Krueger, R. F., & Bezdjian, S. (2009). Enhancing research and treatment of mental disorders with dimensional concepts: Toward DSM-V and ICD-11. World Psychiatry, 8, 306–310.
Krueger, R. F., Derringer, J., Markon, K. E., Watson, D., & Skodol, A. E. (2012). Initial construction of a maladaptive personality trait model and inventory for DSM-5. Psychological Medicine, 42, 1879–1890.
Krueger, R. F., & Markon. C. (2014). The role of the DSM-5 personality trait model in moving toward a quantitative and empirically based approach to classifying personality and psychopathology. Annual Review of Clinical Psychology, 10, 477–501.
Kupfer, D. J., & Regier, D. A. (2011). Neuroscience, clinical evidence, and the future of psychiatric classification in DSM-5. American Journal of Psychiatry, 168, 172–174.
Lang, P. J., & McTeague, L. M. (2009). The anxiety disorder spectrum: Fear imagery, physiological reactivity, and differential diagnosis. Anxiety, Stress, & Coping, 22, 5–25.
Lenzenweger, M. F., Lane, M. C., Loranger, A. W., & Kessler, R. C. (2007). DSM-IV personality disorders in the National Comorbidity Survey Replication. Biological Psychiatry, 62, 553–556.
Livesley, W. J., Jang, K. L., & Vernon, P. A. (1998). Phenotypic and genetic structure of traits delineating personality disorder. Archives of General Psychiatry, 55, 941–948.
Maples, J. L., Carter, N. T., Few, L. R., Crego, C., Gore, W. L., Samuel, D. B., & Markon, K. E. (2015). Testing whether the DSM-5 personality disorder trait model can be measured with a reduced set of items: An item response theory investigation of the personality inventory for DSM-5. Psychological Assessment, 27, 1195–1210.
McArdle, P. (2008). Use and misuse of drugs and alcohol in adolescence. British Medical Journal, 337, 46–50.
National Institute for Health and Care Excellence. (2009). Borderline Personality Disorder: Recognition and Management (NICE Clinical Guideline No. 78). Retrieved from www.nice.org.uk/guidance/CG78/
Novalis, F., Araujo, A., & Godinho, P. (2015). Historical roots of histrionic personality disorder. Frontiers in Psychology, 6, 1463–1467.
Olfson, M., Blanco, W., & Greenhill, L. L. (2013). Trends in office-based treatment of adults with stimulants in the United States. Journal of Clinical Psychiatry, 74, 43–50.
Paris, J. (2012). The Bipolar Spectrum: Diagnosis or Fad? New York: Routledge.
Paris, J. (2013). Anatomy of a debacle: Commentary on “Seeking clarity for future revisions of the personality disorders in DSM- 5.” Personality Disorders: Theory, Research, & Treatment, 4, 377–378.
Paris, J., Bhat, V., & Thombs, B. (2015). Is adult ADHD being over-diagnosed? Canadian Journal of Psychiatry, 60, 324–328.
Paris, J., Gunderson J. G., & Weinberg, I. (2007). The interface between borderline personality disorder and bipolar spectrum disorder. Comprehensive Psychiatry, 48, 145–154.
Paris, J., & Kirmayer, L. (2016). The NIMH research domain criteria: A bridge too far. Journal of Nervous and Mental Diseases, 204, 26–32.
Potuzak, M., Ravichandran, C., Lewandowski, K. E., Ongur D., & Cohen, B. (2014). Categorical vs dimensional classifications of psychotic disorders. Comprehensive Psychiatry, 53, 1118–1129.
Reichborn-Kjennerud, T., Czajkowski, N., Torgersen, S., Neale, M. C., Orstavki, R. E., & Kendler, K. S. (2007). The relationship between avoidant personality disorder and social phobia: A population-based twin study. American Journal of Psychiatry, 164, 1722–1728.
Rossell, D. R., Futterman, S. E., McMaster, A., & Siever, L. J. (2014). Schizotypal personality disorder: A current review. Current Psychiatry Reports, 16, 452–460.
Schroeder, K., Fisher, H. L., & Schafer, I. (2013). Psychotic symptoms in patients with borderline personality disorder: Prevalence and clinical management. Current Opinion in Psychiatry, 26, 113–119.
Silk, K. R. (2016). Personality disorders in DSM-5: A commentary on the perceived process and outcome of the proposal of the Personality and Personality Disorders Work Group. Harvard Review of Psychiatry, 24, 309–310.
Tyrer, P., & Alexander, J. (1979). Classificiation of personality disorder. British Journal of Psychiatry, 135, 163–167.
Tyrer, P., Crawford, M., Mulder, R., & Blashfield, R. (2011). The rationale for the reclassification of personality disorder in the 11th revision of the International Classification of Diseases (ICD-11). Personality and Mental Health, 5, 246–259.
Widiger, T. A. (2007). Dimensional models of personality disorder. World Psychiatry, 6, 79–83.
World Health Organization. (1992). International Statistical Classification of Diseases and Related Health Problems (10th revision, ICD-10). Geneva: World Health Organization.
Zachar, P., & First, M. B. (2015). Transitioning to a dimensional model of personality disorder in DSM 5.1 and beyond. Current Opinion in Psychiatry, 28, 66–72.
Zanarini, M. C. (2009). Psychotherapy of borderline personality disorder. Acta Psychiatrica Scandinavica, 120, 37–41.
Zanarini, M. C., Frankenburg, F. R., Dubo, E. D.. Sickel, A. E., Trikha, A., & Levin, A. (1998). Axis I comorbidity of borderline personality disorder. American Journal of Psychiatry, 155, 1733–1739.
Zimmerman, M., & Mattia, J. (1999). Differences between clinical and research practices in diagnosing borderline personality disorder. American Journal of Psychiatry, 156, 1570–1574.
Zimmerman, M., Rothschild, L., & Chelminski, I. (2005). The prevalence of DSM-IV personality disorders in psychiatric outpatients. American Journal of Psychiatry, 162, 1911–1918.
Zoccolillo, M., Pickles, A., Quinton, D., & Rutter, M. (1992). The outcome of childhood conduct disorder: Implications for defining adult personality disorder and conduct disorder. Psychological Medicine, 22, 971–986.