What Are Personality Disorders

The Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV) describes several overarching aspects of personality disorders (Axis II) that theoretically distinguish them from other psychiatric disorders (Axis I). The personality disorders were originally placed on Axis II along with mental retardation because, in theory, they begin early in development and last a lifetime. However, recent research suggests that some personality disorder symptoms may respond to both pharmacotherapeutic and psychotherapeutic interventions (Sanislow and McGlashan, 1998). Personality disorders are additionally described by the DSM-IV as an enduring or chronic pattern of experience or behavior that deviates markedly from the expectations of an individual's culture involving cognition, affect, social function, or impulse control. Such a pattern:

1. is inflexible and pervasive across different domains of functioning;

2. leads to clinically significant distress or impairment;

3. is stable and begins early in life;

4. is not due to another mental disorder; and

5. is not due to the direct physiological effects of a substance or medical condition.

Detailed descriptions of each of the 10 personality disorder diagnoses can be found in the DSM-IV. However, because we believe that the clearest linkages between personality disorders and physiology may lie at the level of symptoms and their interrelationships rather than at the level of existing diagnoses, we will focus more on the symptoms that comprise the diagnoses rather than on the diagnoses themselves. The 10 personality disorders listed in the DSM-IV can be grouped into 3 clusters on the basis of similar symptoms. Specifically, people diagnosed with a personality disorder in cluster A (paranoid, schizoid, or schizotypal) tend to show odd and eccentric behavior. Interpersonally, they are often reclusive and suspicious. People diagnosed with a personality disorder in cluster B (histrionic, narcissistic, borderline, or antisocial) tend to show dramatic, emotional, and impulsive behavior. People diagnosed with a personality disorder in cluster C (avoidant, dependent, and obsessive-compulsive) tend to show anxious or fearful behavior (see Table 5.1).

Since personality disorders are difficult to diagnose reliably, incidence data are rare and variable. According to a recent review, current U.S. population estimates for any personality disorder (PD) based on DSM-III-R criteria ranged from 6.7 to 33.1 percent, with the authors concluding that lifetime prevalence of at least one personality

TABLE 5.1. Primary Symptoms of DSM-IV Personality Disorders


Personality Disorder

Primary Symptoms

Odd/eccentric (A)

Impulsive/dramatic (B)

Anxious/fearful (C)

Paranoid Schizoid


Antisocial Borderline






Distrust and suspiciousness

Detachment from social relationships; restricted emotional expression

Discomfort in close relationships; cognitive/perceptual distortions; eccentric behavior Disregard for and violation of the rights of others

Instability in interpersonal relationships, self-image, and affect; impulsive behavior Grandiosity; need for admiration;

lack of empathy Excessive emotionality; attention-seeking

Social inhibition; feeling of inadequacy; hypersensitivity to negative evaluation Submissive and clinging behavior;

excessive need to be taken care of Preoccupation with orderliness, perfectionism, and control disorder diagnosis appears to be approximately 10 to 15 percent. Estimates for the odd/eccentric cluster fell under 6 percent (with schizotypal PD being most prevalent), estimates for the impulsive/dramatic cluster fell under 8 percent (with histrionic PD being most prevalent), and estimates for the anxious/fearful cluster fell under 18 percent (with obsessive-compulsive PD being most prevalent) (Mattia and Zimmerman, 2001). However, these numbers surely overestimate the number of diagnoses actually made since most of the estimates were calculated from randomly sampled individuals that were recruited to receive a diagnostic psychiatric interview. Commonly, people with personality disorders lack awareness that they have a problem and so would be less likely to voluntarily submit themselves to such an interview.

Some research has hinted at demographic differences in the incidence of personality disorders. One of the better documented demographic differences involves gender specificity. People diagnosed with antisocial and obsessive-compulsive personality disorders are more likely to be male, whereas people diagnosed with dependent personality disorders are more likely to be female. Overall, people diagnosed with personality disorders also tend to be younger than the age of the general population, except in the case of schizotypal personality disorder (Zimmerman and Coryell, 1989).

Other differences may involve culture specificity. For instance, a lower incidence of antisocial personality disorder has been reported in some (e.g., China and Japan) but not all (e.g., Korea) Asian countries (Lee et al., 1987). Of course, these differences also raise the possibility that culturally biased value judgments influence the definition of personality disorder criteria.

Despite the heuristic and descriptive utility of DSM-IV personality disorder diagnoses in medical settings, researchers have noted several shortcomings of these categorical diagnostic schemes. First, although the DSM-IV places personality disorders on a separate axis from other psychiatric disorders (Axis I), personality disorders often co-occur with other psychiatric disorders, and often do so in predictable ways (Dolan-Sewell et al., 2001). For instance, people with antisocial personality disorder are more likely to also receive a diagnosis of substance dependence. Second, although the DSM describes personality disorders as belonging to a distinct category from normal personality, a preponderance of empirical evidence suggests that personality disorder symptoms are continuously distributed across both clinical and healthy samples (Livesley et al., 1994). This fact helps to explain why clinicians might have difficulty establishing stable "cutpoints" for distinguishing personality disorder diagnoses from normalcy. Third, personality disorder diagnoses are difficult to measure since they often show poor psychometric properties such as validity (i.e., diagnostic criteria index the targeted traits/symptoms but not other traits) and reliability (i.e., diagnostic criteria show stability across different measurement attempts) (Blais and Norman, 1997). Validity comes in many forms and can include either internal validity (i.e., criteria that index the same thing are more correlated with each other than with criteria that index something else) or external validity (i.e., criteria predict relevant external features such as etiology and prognosis). Studies of the internal validity of personality disorder criteria suggest that they show only modest convergent (O'Boyle and Self, 1990) and discriminant validity (Widiger et al., 1991). In other words, a criterion for a given diagnosis is as likely to correlate with criteria from different diagnoses as with criteria from the same diagnosis. Fourth, personality disorder diagnoses have limited clinical utility in that they do not typically help practitioners to choose between distinct pharmacological or psychotherapeutic interventions (Sanderson and Clarkin, 1994). Fifth, because they have been defined by the DSM-IV, the criteria for personality disorders have not been wholly empirically derived. Rather, they have emerged through a combination of historical precedence, clinical observations, legal necessity, and repeated deliberations by expert committees (Frances et al., 1994). As a result, the disorders and their criteria have changed somewhat with each new edition of the DSM. Together, these five shortcomings of the categorical diagnostic framework threaten to hinder investigators' abilities to define personality disorders in a quantitative and replicable way, and so might slow cumulative research on the occurrence and treatment of personality disorders.

In an attempt to circumvent these shortcomings, a number of theorists have proposed that personality disorders be defined dimensionally rather than categorically. An illustration of this distinction appears in the field of cognitive testing, where intelligence can be described either with a continuous measure such as the intelligence quotient (i.e.,

IQ) or according to a cutoff with a categorical label such as "normal" versus "retarded." In a similar manner, Siever and Davis (1991) proposed that four continuous behavioral dimensions may underlie both personality disorders at less severe levels (Axis II) and clinical psychiatric disorders at more severe levels (Axis I). These dimensions include cognitive/perceptual organization, impulse control, affect regulation, and anxiety modulation. According to their proposal, cognitive/perceptual aberrations should map onto the odd/eccentric cluster of personality disorders (i.e., cluster A of Axis II) as well as onto schizophrenia (Axis I). Poor impulse control and affect regulation should map onto some of the impulsive/dramatic cluster of personality disorders (i.e., cluster B, specifically, borderline and antisocial disorders of Axis II) as well as onto mood disorders (Axis I). Poor anxiety modulation should map onto the anxious/fearful cluster of personality disorders (i.e., cluster C of Axis II) as well as onto anxiety disorders (Axis I). These proposed dimensions suggest that several continua bridge Axis II and Axis I and can potentially account for the frequently observed co-occurrence of personality disorder and psychiatric diagnoses.

In support of this proposed continuity between Axis II and Axis I, research suggests that at least one disorder in the odd/eccentric cluster (schizotypal PD) lies on a continuum with schizophrenia (Oldham et al., 1995). This continuity particularly seems to hold in the case of negative symptoms such as affective blunting and a lack of social engagement (Chapman et al., 1994). However, research has generally not supported selective dimensional relationships between the impulsive/dramatic cluster of personality disorders and mood disorders, or between the anxious/fearful cluster of personality disorders and anxiety disorders. Instead, people with either impulsive/dramatic or anxious/fearful personality disorders have a higher risk for co-occurrence of all types of Axis I disorders (Dolan-Sewell et al., 2001). Some exceptions to this apparent lack of specificity include the findings that impulsive/dramatic cluster personality disorders uniquely co-occur with increased rates of substance abuse (Oldham et al., 1995) and that anxious/fearful cluster personality disorders preferentially co-occur with increased rates of somatoform disorders (Tyrer et al., 1997).

A second dimensional approach to assessing personality disorders has arisen from empirical data rather than from theory. Livesley and colleagues identified and culled a prototypical set of personality disorder symptoms (as judged by psychiatrists) spanning all Axis II diagnoses with minimal overlap. They then combined these items in order to construct a psychometric instrument called the Dimensional Assessment of Personality and Psychopathology (DAPP) (Livesley et al., 1992). Next, they administered the DAPP to patients with personality disorders as well as healthy volunteers. Finally, they conducted factor analyses (a mathematical method of examining the correlational structure between many items) on both patients' and healthy volunteers' responses. The investigators found that a similar factor structure described relations among symptoms in both patients and healthy volunteers, suggesting continuity across the groups, but also that the patients had more extreme scores than the healthy volunteers. While factor analysis with an oblique rotation (which allows dimensions to correlate with each other) yielded an underlying structure similar to the discrete personality disorders listed in the DSM-III-R (see column 2 of Table 5.1), factor analysis with an orthogonal rotation

TABLE 5.2. Conceptual Translation Scheme for Personality Disorder Clusters and Personality Traits Measured by a Selection of Psychometric Inventories









Low dom/

Extraversion (—)

Extraversion (—)

eccentric (A)

low aff



High dom/


Agreeableness (—)

dramatic (B)

low aff





anxious (C)




aLivesley et al. (1992). bWiggins (1988). cEysenck and Eysenck (1992). dCosta and McCrae (1992).

aLivesley et al. (1992). bWiggins (1988). cEysenck and Eysenck (1992). dCosta and McCrae (1992).

(which does not allow dimensions to correlate with each other) revealed four factors similar to those found in studies of healthy personality (i.e., neuroticism, introversion, disagreeableness, and compulsivity) (Livesley et al., 1993). Coincidentally, Clark and colleagues used a similar empirical strategy to construct a measure of personality disorder symptoms [called the Schedule for Nonadaptive and Adaptive Personality (SNAP)] and found similar results (Clark, 1993). Of greatest interest, the four orthogonal factors observed in both studies appear to comprise pathologically extreme versions of four of the "big 5" factors commonly observed in studies of normal personality (Clark and Harrison, 2001; Livesley et al., 1998) (see Table 5.2).

In sum, studies support some degree of continuity between Axis II and Axis I, given that the odd/eccentric cluster of Axis II may lie on a continuum with the Axis I diagnosis of schizophrenia. However, existing evidence even more strongly suggests that the impulsive/dramatic and anxious/fearful cluster disorders of Axis II lie on a continuum with symptoms shared not just by one but by several Axis I diagnoses (Angst and Ernst, 1993), as well as with traits that comprise "normal" personality. Thus, it may be that some personality disorder symptoms represent extreme variants of normal personality traits. Before addressing this possibility in greater detail, we briefly review how researchers operationally define and measure "normal" personality.

Emergency Panic Remedies

Emergency Panic Remedies

Eliminate Your Panic And Anxiety Attacks Instantly With These Proven Techniques! There’s
a panic

 of control.

Get My Free Ebook

Post a comment