Cette étude se propose d’évaluer les relations entre les troubles de la personnalité et la psychopathie. Un regard particulier sera porté sur le trouble borderline vu son implication pour la prise en charge. L’hypothèse générale est qu’il existe des liens entre la psychopathie et la présence d’autres troubles de la personnalité, principalement de l’axe II cluster B (narcissique, antisocial, histrionique, borderline) et parmi ceux-ci un lien particulier avec le trouble de la personnalité borderline. La population est composée de 80 détenus adultes de sexe masculin (âge : M=31,48 ans ; SD=11,06). Les outils d’évaluation sont : la PCL-R [Hare, RD. The Hare psychopathy checklist-revised manual. Toronto: Multi-Health System Inc; 2003], le SCID-II [J Personal Disord 9 (1995) 83–91], le MINI [Lecrubier Y, Weiller E, Hergueta T, et al. MINI, Mini International Neuropsychiatric Interview. French Version; 1998] et la WAIS-III [Eur Rev Appl Psychol 4 (2001) 437–41]. Nos résultats suggèrent que les troubles narcissique et antisocial sont positivement associés à la PCL-R (score global, facteurs 1 et 2, ces deux derniers représentant les dimensions interpersonnelle et comportementale de la psychopathie) ; un lien avec le trouble paranoïaque est aussi apparu (cluster A). Le trouble borderline est seulement associé au score global et au facteur 2. Les régressions linéaires multiples (procédure Stepwise) suggèrent que les diagnostics de personnalités antisociale et paranoïaque ont un effet sur le score total à la PCL-R et le facteur 2. Les troubles de personnalités antisociale et narcissique ont un impact sur le facteur 1. Contrairement à ce qui était attendu, et bien que corrélée au score total à la PCL-R et au facteur 2, la personnalité borderline ne prédit significativement ni le score total ni celui de chaque facteur de la PCL-R. Ces résultats suggèrent qu’impulsions et caractéristiques antisociales (facteur 2), narcissiques et paranoïaques sont essentielles à repérer.

Recent clinical and empirical works are based on Cleckley’s clinical observations in which psychopathy is viewed as a personality disorder, characterised by a lack of emotions, callousness, unreliability and superficiality. Hare operationalised Cleckley’s concept of psychopathy by developing the Psychopathy Checklist-Revised composed of 20 items that load on two factors in majority: factor 1 (personality aspects of psychopathy) and factor 2 (behavioural manifestations), close to the antisocial personality disorder (DSM-IV criteria). Comorbidity is strong with antisocial personality disorder but also with histrionic, narcissistic and borderline disorders.

As results of categorical studies relative to comorbidity suggest a strong comorbidity between psychopathy and other personality disorders, and particularly cluster B disorders (axis II, DSM-IV), this study assesses the relationships between psychopathy (dimensional approach) and personality disorders (categorical approach) and particularly with the borderline personality disorder. The aim of this study is also to underline the complementarity of categorical (SCID-II) and dimensional approaches (PCL-R), and the utility of the standardised clinical examination.

We hypothesised positive associations between psychopathy and other personality disorders, mainly with the cluster B axis II (narcissistic, antisocial, histrionic, and borderline). Among those disorders, a particular link exists with the borderline personality disorder, considering that their association may attenuate the pathological level of the psychopathy. The sample included 80 male inmates from French prisons (age: M=31.48; SD=11.06). Each participant was evaluated with the PCL-R to assess the level of psychopathy and the SCID-II to assess the possible presence of personality disorders. The MINI and the WAIS-III were used to exclude respectively those who presented an axis I comorbidity (mood disorders and psychotic disorders established at the moment of the testing), or a backwardness (IQ<70). Correlations and multiple linear regressions analysis (with the Stepwise procedure) were used to analyse the data.

As expected, the results suggested positive correlations between narcissistic, antisocial personalities and scores of psychopathy (from 0.36 to 0.63); paranoid personality was less expected (from 0.32 to 0.47). Borderline personality was associated with both the total score of psychopathy (0.24) and the score of factor 2 (0.30). Linear regression analysis revealed that the antisocial and paranoid personalities predicted the total score (R²=38%) and the factor 2 (R²=45%) of the PCL-R. Antisocial and narcissistic personalities predicted factor 1 (R²=22%). However, in the different models, contrary as predicted, the borderline personality was not a significant predictor.

First, these results underline the importance of impulsivity above all for the cluster B personality disorders and secondly, the importance of considering impulsivity with antisocial (factor 2), narcissistic and paranoid characteristics. Moreover, because of the transversality of impulsivity, the literature outlined the cross-over between cluster B disorders and psychopathy. These different studies could have important clinical consequences (risk of violence, therapeutic indications and forecast). These results also emphasize the necessity of standardised examinations. Implications for treatment are outlined: the treatment may be adapted according to the comorbidities having an effect on psychopathy that is antisocial with paranoid personalities, and antisocial with narcissistic personalities.