Un biais de mémoire explicite en faveur de stimuli positifs et négatifs liés à l’apparence corporelle et au poids a été évalué chez 42 jeunes femmes étudiantes à l’université. Cette recherche visait à étudier dans quelles mesures les personnes à risque ou souffrant de troubles du comportement alimentaire montrent un biais mnésique congruent avec leurs préoccupations liées à l’apparence et au poids.

Les participantes ont été invitées à s’exposer à une liste de mots cibles et contrôles tout en réalisant une tâche d’encodage incident auto-référentielle, puis ont réalisé une tâche de rappel libre. Des mesures auto-rapportées ont été proposées aux participantes afin d’évaluer l’insatisfaction corporelle ainsi que les conduites alimentaires.

Les analyses montrent que l’ensemble des participantes rappellent plus de mots liés à l’apparence corporelle et au poids comparativement aux mots non liés. Aucune différence n’a été trouvée entre les différents groupes. De plus, les participantes ne rappellent pas plus de mots négatifs que de mots positifs ou neutres en lien avec l’apparence corporelle et le poids.

Ces résultats sont en contradiction avec les hypothèses découlant du modèle cognitif des troubles du comportement alimentaire. Les implications cliniques et pour de futures recherches de ces résultats sont discutés.

Memory bias for weight and shape related words was investigated in 42 young women university students. According to the eating disorders cognitive model proposed by Vitousek and Hollon, 1990 [3], women suffering from eating disorder develop self-schemas about weight and shape considered to be at the core of eating disorder pathology. Self-schemas about weight and shape appear to bias information processing with a confirmatory objective. Cognitive biases for congruent information, such as memory bias for negative information about weight and shape, are believed to be implicated in the development and maintenance of eating disorders (Williamson et al., 1999) [4]. The literature supports this cognitive model in clinical samples, with women suffering of anorexia nervosa, bulimia nervosa or binge eating disorder (McNally, 2001; King et al., 1991) [5,6]. Results, however, are inconsistent in non-clinical samples or women with subclinical eating disorders (Hunt and Cooper, 2001; Israeli and Stewart, 2001) [10,11]. These inconsistencies could be linked to the use of different methodologies to evaluate memory bias or the lack of pre-test methods for experimental material. Taking into account suggestions from previous researcher in this field, the objective of this study was to investigate whether a non-clinical sample of women with eating disorder or high body dissatisfaction demonstrate a memory bias congruent with their concerns.

Participants (n=42) were invited to view target and control words, i.e., positive, negative and neutral words related to weight and shape and positive, negative and neutral words not related to weight and shape. They performed a self-referent encoding task during the exposition to words and recall memory was subsequently assessed. Self-report measures were taken in order to evaluate body dissatisfaction and eating behaviour.

Non-parametric statistical analysis indicated that all women demonstrated a memory bias to weight and shape related words compared to control words. Indeed, Wilcoxon test is significant (z=2.82, P=0.005), indicating that all women selectively recall more weight and shape related words than not related ones. On the other hand, no difference was found between groups. Indeed, Mann-Whitney T test is not significant (z=−0.57; P=0.57), indicating that women with body dissatisfaction do not recall more weight and shape related words than women satisfied. Likewise, women suffering of eating disorders do not recall more weight and shape related words than women without eating disorder (z=−0.38: P=0.70). Moreover, women did not recall more negative weight and shape related words (those congruent with their preoccupations) than positive or neutral ones.

These findings are in contradiction with hypothesis based on the eating disorders cognitive model (Vitousek and Hollon, 1999; Williamson et al., 1999) [3,4]. Indeed, all women demonstrated a memory bias for information related to self-schemas about weight and shape. These results suggest that the self-schema about weight and shape was activated for all women and thus available for information processing. Therefore, availability of this specific self-schema about weight and shape do not appear to be the core of eating disorders in so far as it seems to be adopted by women without body dissatisfaction or eating disorder. The preferential use of this self-schema about weight and shape by women suffering of eating disorder in information processing, instead of another self-schema, could be a factor in the development and maintenance of eating disorders. Results show also that women did not selectively recall congruent information, i.e., negative words related to weight and shape. These results could have implications for therapeutic interventions in eating disorders. Indeed, it could be relevant to target not only worries about be or becoming fat, but also the desire to be thin. Indeed, results suggest that this theme, drive for thinness, could be an important factor of vulnerability to certain women and could lead them to adopt unhealthy behaviours in order to achieve an ideally thin body. Therefore, prevention programs should target the development and assimilation of self-schemas not related to weight and shape as well as the drive for thinness.