Correlates of selective information processing related to food, shape, and body parts

The purpose of the present study was to examine clinical and demographic correlates of selective information processing of material representing food, body parts, and shape within a population of college women. A computerized version of the modified Stroop color-naming task was used to investigate processing of food and body shape material, while a number of questionnaire measures were used to collect clinical and demographic information.;Analysis of results revealed interference effects for both body parts and shape relative to corresponding matched control categories. Unexpectedly, a facilitation effect was found for food words relative to matched control words. This finding may indicate that only those women with severe eating disorder pathology are likely to display an information processing deficit in relation to food. Importantly, other researchers have also failed to find a disruption effect for food words in non-eating-disordered subjects (Tucker & Schlundt, 1995; Ogden & Greville, 1993).;Of great relevance, correlates of the interference and facilitation effects differed by word category. For example, the food facilitation effect was best accounted for by measures of ineffectiveness, general psychopathology, and feeling out of control regarding food. Current dieting status and drive for thinness were the strongest predictors of the body part interference effect, supporting the results of Tucker & Schlundt (1995). The shape interference effect was, curiously, more strongly related to general psychopathology than to eating disorder symptomatology. Similarly, a combined shape/body interference effect was best predicted by general psychopathology, although purging was also able to account for a significant amount of variance. This latter finding, importantly, is simply the reverse of what Cooper & Fairburn (1993) found in their study of bulimic women.;Finally, the paper-and-pencil measures were subjected to factor analysis, the results of which support the utility of extending Tobin et al.'s (1991) multifactorial model of bulimia to a population of college women. Eating disorder symptoms certainly appear to differ quantitatively, rather than qualitatively, among women regardless of diagnostic status.