General clinical description of personality disorders

Armand W. Loranger Chapter.. References

During the past two decades there has been an unprecedented interest in the personality disorders. This is due in no small measure to the introduction in 1980 of the American Psychiatric Association's multiaxial classification system.(1) This influential nosology assigned a separate axis to the personality disorders. An indication of the dramatic impact of that innovation is a statistic on diagnostic practice at a leading university hospital in the United States. (2) There was a more than twofold increase in the diagnosis of personality disorders (19.1 per cent to 49.2 per cent) when 5143 consecutive admissions during the last 5 years of DSM-II (1975-1979) were compared with 5771 patients admitted during the initial 5 years of DSM-III (1981-1985). The increase was not due to artefacts like a change in the theoretical orientation of the staff, differences in the pathways of referral, demographics of the patient population, or reimbursement formulas. This was demonstrated when the transition from one classification system to the other occurred at midyear in 1980. When the first and last 6 months of that year were compared, the statistics were very similar (16.6 to 52.4 per cent).

Of no small significance is the fact that 97 per cent of the patients with a personality disorder also had other psychiatric conditions, typically anxiety, mood, alcohol, drug, and eating disorders. Presumably in the pre-DSM-III era, clinicians believed that such Axis I diagnoses usually provided an adequate description of a patient's current illness, and a sufficient explanation for the subjective distress or social and occupational impairment that often accompanied it. This raises the interesting question of whether personality disorders were previously underdiagnosed, or are now overdiagnosed. Although a definitive answer is not immediately forthcoming, the question does bring to the fore some fundamental issues about the nature of personality disorders.

According to DSM-III and its revisions, personality disorders are characterized by maladaptive traits that cause subjective distress or significant impairment in social or occupational functioning. The behaviour is deeply ingrained and inflexible, displayed in a wide range of personal and social situations, enduring rather than episodic, and has its onset by adolescence or early adulthood. Although these requirements also describe certain features of some mental state or Axis I disorders, they are not absolutely essential for their diagnosis.

Not all the mentally ill have personality disorders, but like everyone else they do have personalities. Since many of the criteria used to diagnose personality disorders are exaggerations of traits observed in the general population, there may be a temptation to lower the threshold for a particular criterion and label a trait pathological, when it is within the broad, somewhat arbitrary, limits of normality. This may be more likely to occur when the classification system requires that a decision regarding the presence or absence of a personality disorder be made for all patients.

It is sometimes difficult, even for the experienced clinician, to disentangle personality disorders from other psychiatric conditions. The availability of a separate axis for personality disorders may bias the decision-making process in that direction. This may be compounded by the error introduced by trait-state artefacts, (3) a phenomenon known to astute clinicians long before it was discovered by psychometricians. It is not always easy to penetrate the symptoms of other disorders, and to reconstruct a patient's premorbid personality. A common example is the depressed patient who appears introverted, but whose family and acquaintances are quick to point out that he is actually quite extraverted when his usual self. Clinicians may succumb to such artefacts when they view a patient in the midst of an episode of illness.

This is not to suggest that most personality disorders are merely Axis I disorders masquerading as Axis II. For example, I do not share the view of some that borderlines are merely misclassified bipolar II disorders. However, the more prominent that anxiety or an abnormal mood are in the clinical picture, the greater the potential for error. All other things being equal, there is less likelihood of misdiagnosing an antisocial or schizoid disorder than a borderline or avoidant one. In dealing with egosyntonic conditions, it is usually not necessary to rule out anxiety and mood disorders as the reason for the clinical presentation. However, with egodystonic disorders, there are circumstances when the ultimate diagnostic decision may have to await the successful treatment of the Axis I disorder, in order to determine whether there is any residue of what previously appeared to be the manifestations of a personality disorder.

Little is known about the other reasons for the frequent co-occurrence of personality disorders and mental state disorders. (4) In the absence of epidemiological data, it is possible that the current statistics on the subject, which are derived from treated cases, may overstate the relationship. Those with two disorders may be more likely to seek treatment than those with only one. For patients with similar disorders, for example social phobia and avoidant personality, the explanation may reside in overlapping symptomatology, shared aetiologies, or an imperfect nosology.

One disorder may also create a greater vulnerability to another, as does, for example, AIDS with pneumonia. The situation is further complicated by conditions like substance abuse and eating disorders, where the chronology of their appearance, and therefore the direction of the putative causal connection with personality, is not always readily identifiable. Some Axis I and Axis II distinctions are also somewhat arbitrary, particularly when the disorders have a wide range of severity or are viewed as part of a spectrum. ICD-10 lists schizotypal under schizophrenia and related conditions, while DSM-IV considers it a personality disorder.

The term 'comorbidity' is frequently used to describe the association of personality disorders with other psychiatric conditions. If the term is meant to imply the presence of a distinctly different clinical entity, this is probably ill-advised. At the present time, in the absence of more definitive data on the subject, the expression 'co-occurrence' is probably preferable. The added presence of a personality disorder is often but not invariably associated with a poorer response to the treatment of mental state disorders like anxiety and depression.(5) But there are also reports suggesting that the presence of some specific personality disorders, for example the dependent type, may augur a more favourable response to treatment, perhaps due to greater compliance.(5)

It is debatable whether reserving a separate axis for personality disorders has actually led to their overdiagnosis. Still it is worth considering the pros and cons of having a different kind of Axis II. Personality disorders might be returned to the primary axis where they were in DSM-II, and still are in ICD-10. Axis II could then be used to provide a concise description of the most salient premorbid personality traits of all patients, with the possible exception of those with a personality disorder. Implicitly this would revive a distinction made in early twentieth-century German psychiatry. Traits were considered pathogenic when they were aetiologically linked to the principal manifestations of the clinical syndrome, as with personality disorders. They were pathoplastic when they merely coloured or modified the expression, course, or outcome of a clinical syndrome. Consider, for example, two patients with unipolar depression, one introverted, modest, and conscientiousness, the other extraverted, exhibitionistic, and dependent. The comprehension and management of their depressions would likely present quite different challenges to the clinician.

Such an alternative classification would not be without problems of its own. Like the current system it might be susceptible to trait-state artefacts. A more formidable hurdle, perhaps, would be determining the array of normal traits that should be used to characterize a patient's personality, as well as the method of assessing them. One would soon encounter the considerable divide that separates clinicians from psychologists devoted to the study of normal personality, most notably the conspicuous lack of integration and cross-fertilization of ideas.

Psychiatrists generally favour categorical classification systems, while students of normal personality usually prefer dimensional taxonomies. Categorical classification has a long clinical tradition, and it has proved to be a very effective shorthand form of communication. Ideally, a category should not only specify the defining features of a disorder, but it should also have points of rarity with normality and other disorders. Personality disorders like many other mental as well as physical disorders do not always conform to this ideal, but that does not negate the usefulness of categorical classification. Advocates of dimensional taxonomies also argue that because many abnormal personality traits exist on a continuum with normality, dichotomizing them diminishes the reliability of diagnosis. There is no inherent reason, however, that clinicians have to choose between dimensional and categorical classifications. Dimensions can be used to supplement categorical information, witness their long-standing use in the diagnosis of mental retardation and hypertension.

P>The methods of assessment employed by clinicians and personality psychologists often differ. The delineation of the personality disorders emerged primarily from clinical observation. Therefore, the interview approximates the clinical diagnostic process more than the self-administered questionnaires favoured by many psychologists. However, there are data from the pre-DSM-III era(6) and the field trials of DSM-III and ICD-10(7) indicating that clinicians tend to agree less about the diagnosis of personality disorders than they do about many other mental disorders. This has been the inspiration for the development of several semistructured clinical interviews to improve diagnostic reliability. These instruments were primarily intended to encourage investigators to employ more uniform methods of case identification in their research, in order to facilitate the comparison and generalization of results. The interviews, however, also have heuristic value, and some clinicians have found them useful in their practice.

Perhaps the most widely used and extensively tested semistructured interview of this kind is the International Personality Disorder Examination ( IPDE)(8,9) It was developed for international and cross-cultural use, as part of the World Health Organization ( WHO) instrumentation package, and is available in separate modules (1°> based on either ICD-10(1! or DSM-IV.(12> The 67 items for ICD-10 and 99 items for DSM-IV are arranged topically under six headings: Work, Self, Interpersonal relationships, Affects, Reality testing, and Impulse control. In addition to diagnoses, the interview provides dimensional scores for each disorder based on its diagnostic criteria, regardless of whether the subject has the disorder. There is a detailed item-by-item scoring manual that defines the scope and meaning of each criterion, and that provides anchor points for scoring. The interview takes into account the age at onset and the duration of the behaviour, and it requires substantiation of responses with anecdotes and examples. The scoring also has provision for information from informants. At the conclusion of the interview, the final algorithmic integration of the scores is done clerically or with a personal computer. The IPDE is designed for use by experienced clinicians, namely those capable of making independent psychiatric diagnoses. It is available in more than 20 languages, and training in its use is available at WHO centres around the world. Table !

displays a sample item from the interview.

n vi i «-i ij r ■ 11 i ha i i tt ■ ■ ■ rm ta bit n i r+i* i rr - r ■ ■ ■ ! Hr f r ■ nn —si i r!

n vi i «-i ij r ■ 11 i ha i i tt ■ ■ ■ rm ta bit n i r+i* i rr - r ■ ■ ■ ! Hr f r ■ nn —si i r!

Table 1 Sample item from the International Personality Disorder Examination (IPDE)

The IPDE was subjected to a worldwide field trial that involved 716 patients examined by 58 psychiatrists and clinical psychologists at 14 clinical facilities in Austria, England, Germany, India, Japan, Kenya, Luxembourg, The Netherlands, Norway, Switzerland, and the United States. The inter-rater agreement and temporal stability of the interview were comparable to that reported with similar instruments for the psychoses, anxiety, mood, and substance use disorders. This was true despite the fact that the IPDE was subjected to an unusually exacting test with culturally and linguistically diverse patients in North America, Europe, Africa, and Asia. Although the WHO study was not epidemiological in nature, it did provide evidence that the personality disorders in DSM and ICD, despite their obvious origins in European and American psychiatry, were present, identifiable, and clinically meaningful, in a variety of nations and cultures. Not surprisingly, the study also found that dimensional measures were more reliable than categorical ones, additional evidence in support of the argument for retaining both methods of assessment.

The literature suggests that self-administered personality questionnaires are especially prone to false-positive diagnoses and should not be used to make psychiatric diagnoses/:1..3) This does not undermine their potential value as screening instruments. (14) Nor does it diminish their usefulness as economical, objective measures of the traits that underlie a particular diagnostic category. Indeed, if future nosologies should include an axis for listing normal premorbid personality traits, it is inevitable that the optimal method of assessment would be some form of personality questionnaire. A number of such inventories already exist, but because of the terminology and constructs they employ, many of them do not appeal to clinicians, or they are wedded to a particular theory of personality not favoured by some.

The author recently developed such a personality questionnaire with the clinician in mind, but one which also maintains a link with the extensive literature on normal personality traits. It consists of 375 items, each with seven response options, and it assesses 25 normal traits as well as the dimensions underlying each of the 10 DSM-IV personality disorders. It also measures seven broad personality constructs derived from a factor analysis of the combined normal and personality disorder scales. To make the inventory acceptable to a greater number of psychiatrists and psychologists, the normal traits were selected without reference to any particular theory of personality. A collaborative group of clinicians considered the traits meaningful and potentially useful. The scales measuring the traits were refined and developed through extensive field-testing with 2000 subjects in the general population, and several hundred non-psychotic psychiatric patients. To acquaint the reader with the kind of normal traits with potential relevance to clinical practice, Table.2 describes the lower-order or more specific traits, and Table.,...? presents the higher-order or more general constructs, that were derived from a factor analysis of the combined normal and DSM-IV personality disorder scales.

Table 2 Proposed list of clinically relevant normal personality traits








Andiv Pp [-1


WrtLwt (-]






Halnm PD

bt^M^ni H


hjr: iiuinc H}











Table 3 Broad personality constructs derived from factor analysis of 25 normal traits and 10 DSM-IV personality disorder dimensions

The history of medicine is replete with diseases that have shed light on normal physiology, and discoveries regarding normal physiology that have informed and advanced the understanding of pathology. It would be surprising if a similar potential for mutual enrichment were not latent in the increased integration of research on normal and abnormal personality. This should enhance our understanding of the origins and organization of personality, its disorders, and their biological substrate.

Develop The Guts To Talk With Anyone

Develop The Guts To Talk With Anyone

Learn How Helping Others Benefits You And How You Can Begin Accomplishing Powerful Goals In The Process. The secrets behind Living A Confident Life and Inspiring Others!

Get My Free Ebook

Post a comment