Cognitions are of crucial importance in the aetiology and the maintenance of eating disorders. Dysfunctional cognitions in eating disorders are related to body image, self-esteem and feeding. The aim of this paper is to review the actual knowledge in this area. First, we will display cognitive models in eating disorders. Cognitive factors in eating disorders are logical errors, cognitive slippage and conceptual complexity. Eating disorder patients seem to have a deficient cognitive development. Some cognitive models stipulate that eating disorder patients may develop organised cognitive structures schemas concerning the issues of weight and its implications for the self. These schemas can account for the persistence and for the understanding the << choice >> of the eating disorder symptomatology. Cognitive phenomena of interest are self-schema, weight-related schema and weight-related self-schema. The maintenance model of anorexia nervosa argued that, initially there is an extreme need to control eating which is supported by low self-esteem. The maintenance of the disorder is reinforced by three mechanisms : dietary restriction enhances the sense of being in control; aspects of starvation encourage further dietary restriction; concerns about shape and weight encourage restriction. The development and maintenance of bulimic symptomatology are explained by placing a high value on attaining an idealised weight and body shape accompanied by inaccurate beliefs. The cognitive model of specific family of origin experiences puts forward the development of maladaptative expectancies for eating and thinness. Second, we discuss distortions in information processing. a) In feeding laboratories, bulimics show a wide range of caloric intake and a disruption of circadian feeding patterns. In overeating bulimics, large meals occurred mainly during afternoon and evening with high fat and carbohydrate intake, but the majority of meals were of normal size and frequency. Responsivity to food cues indicates that bulimics were more responsive to sight, smell and taste of their favourite binge food, and a greater responsivity was associated with increasing cue salience. Eating disorderpatients appear to have internalised a mediated social rule concerning << good food >> and make drastic selections thus removing the possibility of choice of foodstuffs. b) Experimental processes : temporal factors in the processing of threat seem to be of importance in patients with high levels of eating psychopathology. There is no evidence for preattentive processing biases among anorectics. Changes in information processing speed after treatment were not linked to treatment condition or treatment response. c) Judgement and emotions: in eating disorder patients, distortions of depressogenic nature are found that influence the cognitive style; thoughts about eating, weight and shape are characterised by negative affective tone; negative emotions could account forbulimic behaviour; anxiety and distress are correlated to thought control strategies. Information treating seems to be impaired in a non-homogeneous way. d) Cognitive schemas are seriously maladaptive and not well investigated. In eating disorder patients, core beliefs are absolute, unconditional and dichotomous cognitions about oneself and the world. There are only few studies in this field moreover showing controversial results. Core beliefs can explain links between personality disorders and eating psychopathology. Pathological core beliefs have to be taken in to account because the influence the outcome and the efficacy of cognitive behavioural therapy. Third, the last part of this paper summarises actually available rating scales evaluating distorted cognitions in eating disorders. There are different methods for evaluation : specific and non-specific self-report questionnaires, thought-sampling procedures, methods derived from cognitive psychology. The Mizes Anorectic Cognition questionnaire (MAC) is a well-known self-rating scale with good psychometric properties. The revised form of the MAC appears to be an improvement in the area of internal consistency, sensitivity, and reliability. It is obvious that there is no particular rating scale referring to specific cognitions on food. In conclusion : the main result of this literature review reflects that the cognitive treatment in eating disorders is altered in a specific way on an emotional basis and on self-representation.
