L’objectif de cette étude était de déterminer la validité convergente, prédictive et concourante du Questionnaire d’Alexithymie de Bermond et Vorst (BVAQ-B) dans une population clinique (ie, TCA, n = 59) et de témoins (n = 191). L’échelle d’alexithymie de Toronto à 20 items (TAS-20), qui est l’autoquestionnaire le plus couramment employé, a été utilisée comme critère externe afin de déterminer la validité convergente du BVAQ-B. Une échelle de mesure de la dépression (Inventaire de dépression de Beck : BDI), d’anxiété (échelle d’anxiété trait et état de Spielberger : STAI-T et E) ainsi que deux échelles de mesure d’anhédonie (échelle d’anhédonie sociale : SAS et échelle d’anhédonie physique : PAS) ont été utilisées afin de comparer la validité concourante du BVAQ-B et de la TAS-20. Les résultats mettent en évidence que la validité convergente et prédictive du BVAQ-B en population clinique et chez les témoins sont satisfaisantes. Par ailleurs, l’étude de la validité concourante a permis de préciser les liens entre l’alexithymie et d’autres dimensions de la régulation émotionnelle. Enfin, cette étude montre que les patientes et les témoins présentent une alexithymie de nature différente. Les témoins alexithymiques seraient caractérisés par une incapacité à mentaliser puis à élaborer des cognitions associées à leurs affects. Les patientes alexithymiques seraient caractérisées par une incapacité à éprouver leurs affects et, par conséquent, par l’impossibilité à mentaliser puis à élaborer des cognitions qui pourraient y être associées. Ce résultat va dans le sens de la littérature qui questionne la mesure adaptative de l’alexithymie dans certaines populations, mais pas toutes.

Alexithymia core features are the difficulties in identifying and describing feelings ; the difficulties in distinguishing feelings from the bodily sensations of emotional arousal ; an impaired symbolization, as evidenced by a paucity of fantasies and other imaginative activity ; and a tendency to focus on external events rather than inner experiences. Several measures of alexithymia have been developed, including interviewer-rated questionnaires and self-report questionnaires. Among the self-report questionnaires, the 20-item Toronto Alexithymia scale (TAS-20) is the most commonly used, but it fails to measure all the core features of alexithymia. A recently developed instrument, the Bermond-Vorst Alexithymia Questionnaire (BVAQ), allows the measurement of the alexithymia core features, as well as an additional one. It appeared to present good psychometric properties, notably the abbreviated BVAQ-form B. The results of recent studies comparing the psychometric properties of the TAS-20 and the BVAQ have recommended the BVAQ over the TAS-20. However, this questionnaire needed further validation. Objectives – Thus, the aim of the present study was to determine the convergent, discriminant and concurrent validity of the Bermond-Vorst Alexithymia Questionnaire – form B (BVAQ-B) in a clinical sample of 59 eating disorder patients, as well as in 191 controls. The TAS-20 constituted the gold standard for the assessment of the BVAQ-B’ convergent validity. To compare the concurrent validity of the BVAQ-B and the TAS-20, participants also completed several self-reports investigating different dimensions of emotion regulation capacities : the 13-item Beck Depression Inventory (BDI), the Spielberger State and Trait Anxiety Inventory (STAI-form Y), as well as the Chapman and Chapman Physical and Social Anhedonia Scales (PAS and SAS). One way analyses of variance were used for mean scores comparisons. Convergent validity was determined using Pearson coefficients of correlation. Results – Results of the analyses suggested the BVAQ-B has a satisfying convergent and discriminant validity. This was observed in both the clinical and control samples. Moreover, the comparison of the convergent validity of the BVAQ-B and the TAS-20 revealed several differences between these two alexithymia self-report questionnaires. The BVAQ-B appeared less sensitive to the subjective emotional state of the participants than the TAS-20. Whereas it was argued the TAS-20 overlaps with other emotional state scores, the BVAQ-B would allow to measure alexithymia more specifically. In addition, the present results allowed to further determine the relations between alexithymia and other dimensions of emotion regulation capacities. The analyses confirmed that alexithymia is linked to other emotion regulation dimensions such as depression and anxiety. Moreover, alexithymia was associated with physical and social anhedonia, two dimensions that received less interest in the alexithymia literature to date. This study also showed that control and clinical sample have different emotion regulation capacities. Eating disorder patients were not only more alexithymic and more depressed, but also more anxious and more anhedonic than the controls. Finally, this study revealed that alexithymia differs whether the alexithymic individuals are patients or controls. Healthy alexithymic individuals (ie, individuals categorized as alexithymic in the control group) seemed characterised by a selective deficit of emotional cognition, with sparing of emotional experience (Bermond’s type II alexithymia). Alexithymics individuals of the eating disorder group seemed particularly unabled to experience affect. This pattern could correspond to Bermond’s type I alexithymia, which is characterised by the absence of emotional experience and, consequently, by the absence of the cognition accompanying the emotion. In summary, results of the present study add to the literature debating on whether alexithymia is similar in different types of population.