The issue of the relative merits of categorical versus dimensional approaches cuts across most domains of psychopathology. However, in the domain of personality disorders it is especially significant, in part, because of the long tradition of research in personality psychology based on dimensional models. Currently, clinicians using DSM-IV must decide whether a patient meets criteria for one or more personality disorders, each considered a separate and distinct category.
If the true nature of personality disorders is dimensional, it would suggest that the most appropriate epidemiological approach might be to determine the mean and standard deviation of the population on the appropriate dimension. Consistent with such an approach, prevalence could be regarded as the proportion of the population that exceeds a threshold associated with impairment and/or distress.
The DSM-IV Options Book (American Psychiatric Association, 1991) described the strengths and weaknesses of the current categorical system. The Options Book pointed out that a dimensional approach would improve flexibility, possibly improve reliability and validity for certain disorders, and save information lost in categorical classification. The inclusion of a dimensional approach in an appendix of DSM-IV was proposed (American Psychiatric Association, 1991). However, lack of consensus about the dimensions to include was a disadvantage of adopting a dimensional approach (American Psychiatric Association, 1991) and DSM-IV continues to use a categorical system.
Costa and McCrae (1990) suggested that personality disorders, at least as they are assessed by the scales created by a number of different investigators, represent an extreme standing or a configuration of extreme standings on the dimensions of normal personality. This might be termed the ''defining model,'' in which the disorder results from exceeding the threshold on some dimension (or dimensions); the dimension is causally related to the disorder and exceeding the threshold is a sufficient cause. A medical example of this model would be cases of essential hypertension without a demonstrable underlying pathophysiological cause for elevated blood pressure. That is, the factors that go into determining the blood pressure of the afflicted individuals are the same as those for the general population. Blood pressure, in general, is normally distributed in the population and some individuals, for the same multifarious genetic and environmental factors that determine the blood pressure of human beings in general, fall at the high end of the distribution. Because blood pressure at the high end of the distribution is associated with excess morbidity and mortality, it is justifiably considered a disorder: hypertension. The defining model of personality disorder is similar to this example. Some individuals, due to the same genetic and environmental factors that influence everyone, are at the high or low end of a dimension or dimensions of personality and this standing defines (it is) their personality disorder.
The defining model, however, does not describe the only possible manner in which dimensions of normal personality may relate to personality disorders. There are at least two other models that are equally plausible. In the ''descriptive (trivial) model'' the disorder can be described in terms of a dimension (or dimensions), but the relationship is not causal and is uninformative for understanding the disorder. A medical example of the descriptive model would be explaining Downs syndrome in terms of the ''universally occurring'' dimensions of height and IQ. That is, individuals with Downs syndrome could be described as being at the low end of the continuum of height and at the low end of the continuum of IQ. However, this is an uninformative approach. A telling question would be, ''Are people with Downs syndrome short for the same reason that short people without Downs syndrome are short?'' If the answer is no, then trying to understand Downs syndrome through the mechanism that determines height for most people would not be informative. Individuals with Downs syndrome differ qualitatively from the general population. If a personality disorder is caused by some factor or factors that are independent of "normal" personality, it still might be described in these terms but such a procrustean approach will be counterproductive for understanding the nature of the phenomenon.
In the "predisposing model,'' exceeding a threshold on the relevant dimension is a risk factor for the disorder and could be a necessary cause, but it is not a sufficient cause. Phenylketonuria (PKU) can serve as a medical example of this model. An individual who is homozygous for a defective gene that leads to the production of phenylalanine hydroxylase is vulnerable to the development of PKU. Such a genotype is a necessary but not sufficient cause. In order for the individual to manifest PKU, he or she must be exposed to dietary phenylalanine. If such an individual ingests phenylalanine, damage to the nervous system results. Without such exposure, there is no damage to the nervous system. For example, schizotypal personality disorder could be due to an extreme position on a universal continuum of introversion: extroversion plus possessing a schizophrenia genotype. In this example, introversion is necessary but not sufficient to produce schizotypal personality disorder.
Block and Ozer (1982) discussed a similar issue with regard to the use of typologies in the study of personality. They made a distinction between type-as-label and type-as-distinctive-form. Type-as-label refers to the practice of establishing categories by classifying individuals above some threshold on an underlying continuum as being in the category. The category so established is then given a name or label. The defining model discussed above corresponds to type-as-label. The alternative definition of "category" or "type" described by Block and Ozer is stronger and more implicative. Type-as-distinctive-form refers to "a subset of individuals characterized by a reliably unique or discontinuously different pattern of covariation... with respect to a specifiable (and nontrivial) set of variables'' (Block, 1971). The use of a dimensional approach to personality disorders is not appropriate if these disorders fit the type-as-distinctive-form model. Meehl (1979) made a similar distinction between a communicative taxon and a "true" taxon.
It seems likely that some aspects of personality disorders are related to deviation within normal dimensions of personality while other features are best considered categorical. It is unlikely that any single model will adequately describe all personality disorders. For example, Gunderson et al. (1991) have suggested that the more severe personality disorders (e.g., schizotypal, paranoid, and borderline) may be discrete clinical syndromes with discrete etiological pathways, while less severe personality disorders (e.g., compulsive, avoidant, and dependent) represent extremes of normally occurring traits. It is conceivable that certain types of personality pathology represent a "common final pathway'' for a number of different etiological factors, while others might be uniquely related to a specific standing on universal dimensions of personality.
It is unlikely that the diagnostic system will be changed because personality disorders can be described in dimensional terms; change is only likely to occur if it is demonstrated that personality disorders should be described dimensionally. The result of solely descriptive or clinical research is unlikely to be decisive in resolving the issue of categorical versus dimensional approaches. Research on etiology and pathophysiology may be more promising for adducing the type of evidence required. For example, data from genetic studies might shed light on whether factors contributing to conscientiousness and neuroticism in the general population are responsible for the occurrence of compulsive personality disorder. Another example of such an approach would be to determine whether the relationship between MAO activity and sensation seeking seen in normal populations could predict risk for antisocial personality disorder. A delineation of relationships between dimensions of normal personality and personality disorders is an important step towards acquiring the knowledge required to create a diagnostic system that accurately reflects the nature of personality disorders.
Was this article helpful?
Do You Suffer From High Blood Pressure? Do You Feel Like This Silent Killer Might Be Stalking You? Have you been diagnosed or pre-hypertension and hypertension? Then JOIN THE CROWD Nearly 1 in 3 adults in the United States suffer from High Blood Pressure and only 1 in 3 adults are actually aware that they have it.