Une approche des troubles de la personnalité par le modèle à cinq facteurs
Résumé
Dans cet article, les auteurs examinent et discutent le statut actuel de la notion de trouble de la personnalité décrite par le Manuel diagnostique et statistique des troubles mentaux texte révisé (DSM-IV). Ils exposent une conception alternative des troubles de la personnalité et défendent la thèse de l’existence d’un continuum entre la personnalité « normale » (le modèle à cinq facteurs) et les troubles de la personnalité (conception dite : « spectrale »). Enfin, ils présentent les éléments d’une méthode, basée sur cette conception, permettant – dans une perspective dimensionnelle – d’améliorer la validité du diagnostic des troubles de la personnalité. /
The Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV) describes the more stable and trait-liked aspects of psychopathology as “personality disorders” (PDs) and places them on a separate Axis (Axis II). Ten specific PDs (e.g., borderline) are articulated; each involves cognition, affect or behaviour that deviates from cultural norms is stable over time and causes impairment or distress for the individual and/or the immediate environment. PDs are distinguished from one another by the particular pattern of cognition, affect and behaviour. Numerous critiques of the DSM PDs have been published over the last 25 years. One of the most egregious problems is caused by the use of a categorical approach to the diagnosis of PDs in which individuals receive a diagnosis if they surpass a certain “cut-point” (e.g., five of nine borderline symptoms), but do not receive a diagnosis if they fall below the cut-off (e.g., four or fewer symptoms). These cut-points were not based on empirical data and there is little evidence to suggest that they capture distinct groups of individuals. The use of a categorical model causes a number of problems including the assumptions of homogeneity within the groups (e.g., nondisordered individuals should “look” alike) and heterogeneity across the groups (e.g., those with the disorder should not look like those without the disorder). Another concern with the PD diagnoses is the extremely high level of co-occurrence found among the PDs (e.g., avoidant and dependent PDs) and across Axes (e.g., borderline PD and major depression). If the PDs are independent, nonoverlapping constructs, then the disorders should not co-occur at the rates typically found. Opponents of the current PD diagnostic system argue that PDs need to be reconsidered using dimensional models of normal and abnormal personality traits. One increasingly popular perspective on personality pathology with empirical support involves considering PDs as extreme variants of the dimensions of the five-factor model (FFM). Many trait psychologists agree that five dimensions (i.e., neuroticism, extraversion, openness to experiences, agreeableness, conscientiousness) are both necessary and sufficient to represent the underlying structure of personality traits. Within this perspective, differences between abnormal and normal functioning are thought to be more gradual and quantitative, rather than reflecting qualitative distinct categories. Numerous individual studies have suggested that measures of the FFM relate to the DSM-IV PDs in a consistent manner and that the use of the narrower facets or traits provides a richer, more discriminating description of the PD constructs. The facets of the FFM have been used successfully to translate the DSM-IV PD criteria and describe the prototypical individual with a specific PD. The considerable overlap between the FFM dimensions and PDs has led some to suggest that the latter should be conceptualized as disorders of personality. Widiger et al. suggested a four-step process to assess personality pathology, involving: assessment of an individual's standing on the domains and traits of the FFM; screening for problems relevant to high or low standing on FFM traits; judging if the problems are of clinical significance and optionally, summarizing the personality profile in terms of a DSM-IV PD pattern. While it is unclear whether or when the FFM or a similar model of personality will be adopted for use in the DSM nosology, there is substantial evidence that such a model can be used to understand and assess personality pathology in a valid and useful manner. The ability of the FFM to resolve issues of dimensionality and comorbidity, as well as provide improved clinical utility and treatment planning, suggests significant advantages can be gained through the use of such a model.