Eating behavior in bulimia.

Despite our strong belief in the utility of laboratory studies of eating behavior, we also note several caveats on the data thereby obtained. First, it must be assumed that subjects' behavior is influenced by the laboratory environment and is not identical to eating behavior in a "normal" setting. Second, not all bulimic subjects who were screened for these studies actually participated, so that it is possible that the sample of patients from whom we obtained data differed in some ways from a general clinical population of women with bulimia. Nonetheless, we believe that our data provide compelling evidence that the disturbed eating behavior characteristic of bulimia nervosa can be profitably studied in the laboratory. Even under structured laboratory conditions, most bulimic patients rated one of their multicourse meals as typical of a binge, and, during that meal, consumed a much larger amount of food and ate more rapidly than did controls who were asked to overeat. The significant correlations between the sizes of the multicourse and single-course binge meals and between the size of laboratory binge meals and the size of the "naturally occurring" binge meals reported to the dietician suggest that a reproducible phenomenon is being examined. The results of our studies suggest that the abnormalities of eating behavior in bulimia nervosa cannot be viewed simply as a disturbance of carbohydrate consumption or even as the episodic consumption of a certain type of food. Rather, eating behavior in this syndrome appears more generally disturbed. The most striking difference between the binge and the nonbinge meals of bulimic patients and between the binge eating of patients and the overeating of normal persons is the amount of food consumed, not the macronutrient composition of the meals. In addition, for all four meal types, the patients were hungrier after the end of the meal than were the controls, even though the patients' average caloric intakes were generally larger and their average hunger ratings before the meals did not differ from those of the controls. Certainly, self-induced vomiting may contribute to this abnormality, but it was also observed after nonbinge meals when vomiting did not occur. Together, these data are consistent with the notion that the essential appetitive abnormality in bulimia nervosa lies in the control of the amount of food consumed, not in the consumption of a particular macronutrient or type of food. Patients with bulimia nervosa appear less responsive than normal to the signals that lead to the termination of a meal.(ABSTRACT TRUNCATED AT 400 WORDS)