[Study of the categorization process among patients with eating disorders: a new cognitive approach to psychopathology].

Anorexic and bulimic patients have a highly distorted relationship with food and eating, even though they tend to be knowledgeable about diet and nutrition. The progress of this disease, as well as its complications and associated difficulties, are increasingly understood, while the etiopathogeny of eating disorders remains obscure. The approach that we are proposing involves the study of one of the most fundamental cognitive functions of human reasoning--the cognitive process of categorization. The purpose of this study is to understand the procedures used by these patients to construct representations of food. Categorization, one of the basic features of human cognition, allows individuals to organize their subjective experience of the surrounding environment by structuring its contents. This ability to group different objects into the same category based on their common characteristics is important for explaining the major cognitive activities of planning, memorization, communication and perception. Indeed, our categories reflect our conceptions of the world. They depend on our experiences and representations, as well as the expertise acquired in a specific field. The differences that appear in the categories created by subjects when they are asked to classify objects reveal the properties that are most salient to them and, as a result, the interests, values and ideas associated with these properties. There are three types of properties: perceptive properties, which describe the object's shape, color, odor and texture; structural properties, which relate to the object's components; and functional properties, which specify the way in which the object is used and provide an answer to the question, "What is it used for?". Subjects attribute these functional properties by means of knowledge or inference according to their representation of the object's role; such properties are especially likely to emerge during top-down (theory-driven) processing. The type of processing used (bottom-up or top-down) is dependent on a certain number of factors. We hypothesize, within the context of food product categorization, that patients suffering from eating disorders largely resort to processing based on acquired information or beliefs about the objects, i.e. top-down processing. We present two studies: a naturalistic and exploratory pilot study whose goal is to identify whether the various categorization processes used by eating disorder patients differ from those employed by subjects not suffering from an eating disorder. A second study aims to identify the different categorization procedures. During the first experiment, 68 women (17 control subjects, 17 anorexics, 17 anorexic bulimics and 17 bulimics) aged 18-39 (average age: 26.6) verbalize all representations that come to mind during a limited time period as the name of a food item is read. Eighty-nine food items are presented in alphabetical order. The list is read out loud and all comments are recorded. The data is processed in three ways : an analysis based on the positive or negative valence of each representation, an analysis based on each categories of food and an analysis of representations based on themes expressed. The three analyses (valence, categories of food and theme assigned to the representations) show differences between the representations of the four experimental groups. In fact, the anorexics and anorexic bulimics mainly express strongly negative representations about food, whereas bulimics and control produce representations whose positive and negative valences balances. These negative cognitions concern mainly meat for the control subjects and cakes for the subjects reached of TCA. Concerning theme assigned to the representations, the control subjects produce mainly cognitions relating to the hedonism, the flavor of food and their purpose on health. The anorexics and anorexics-bulimics evoke mainly the fat and sugar content of the foods. The bulimics evoke mainly cognitions relating to the effect on health and the intestinal transit time of food. These results lead one to believe that it is not the bulimic binging and purging of these patients, but rather their restrictive behavior that is the determining factor in the differences in food representations observed between the two experimental groups. During the second experiment, 60 women (15 controls, 15 anorexics, 15 anorexic bulimics and 15 bulimics) aged 18-32 (average age: 25.6) classified 27 food names according to their similarities and differences, and then explained the reasons for their categorizations. The data were analyzed in terms of similarity/difference, and the verbalizations were analyzed by content. The results indicate that 10 of the 27 foods were categorized differently by the controls and the subjects with eating disorders. Subjects classified the following foods: camembert cheese, cold cuts, cheese spread, fruit in syrup, whole milk, mayonnaise, bread, fresh fish, potatoes and plain yogurt. Bulimics and controls use similar classifications for food names, while anorexics and atypical bulimics classify foods in a similar way. Examining the categorization criteria used during verbalizations allows us to better understand these differences. The control group's major criterion seems to be the succession of dishes. These subjects group into separate categories entry foods (beef, eggs, fish, etc.), vegetables, cheese or dairy foods, and finally desserts. Additional foods, like bread and mayonnaise, belong to the same category. Other categories are nutritional criteria (for example, dairy products contain calcium) and biological criteria (for example, bananas and apples are fruits). These categorization criteria include structural properties (which describe what the object is made of) and functional, "academic" properties, those which describe how foods are used, "as in cookbooks or diet books." On the other hand, the categorization criteria expressed by anorexic patients are very different from those used by control subjects: foods that are hard to eliminate, rich, high-fat and therefore indigestible are considered to be similar. Some examples are cold cuts, potatoes, mayonnaise and prepared desserts. A second categorization criterion involves the concept of natural foods : certain foods "are unhealthy because they're processed, so they're bad for you"--one such example is cheese spread. A third criterion concerns the notion of familiar foods: poultry and eggs, for example, are "familiar to us." We are clearly seeing here the importance of functional properties in the categorization of food names: certain foods are indigestible, hard to eliminate, cause heartburn or reflux, are not natural, and thus are avoided. The categorization criteria mentioned by bulimic patients also clearly take into account the functional properties of foods. The criteria are of the following type: "it's filling, it relieves a bulimic attack, it helps prevent heartburn and constipation, etc." It appears that bulimics' categorization criteria are solely associated with these foods' imagined or real effect on the body. The categorization criteria used by anorexic bulimics seem to be especially associated with weight gain or the consumption of such foods during bulimia attacks because "they make you feel full." On the other hand, light foods, which patients allow themselves to eat, are placed in the same category. This study, which seeks to understand the cognitive functioning of eating disorder patients with anorexia and bulimia, has brought new elements to light. All patients exhibit food categorization processes that differ greatly from those displayed by control subjects. Patients also attribute greater significance to the functional properties of foods as compared to controls, who give priority to structural properties. Anorexic and bulimic patients base their food categorizations on the consequences of ingestion, in terms of health, digestion and weight gain. Their processing of food stimuli is therefore radically different and gives a dominating place to top-down processes. Additional studies should supplement these findings in order to gain a better understanding of patients' disturbed processing of information.