Selective visual attention for ugly and beautiful body parts in eating disorders

Abstract

Body image disturbance is characteristic of eating disorders, and current treatments use body exposure to reduce bad body feelings. There is however little known about the cognitive effects of body exposure. In the present study, eye movement registration (electroculography) as a direct index of selective visual attention was used while eating symptomatic and normal control participants were exposed to digitalized pictures of their own body and control bodies. The data showed a decreased focus on their own ‘beautiful’ body parts in the high symptomatic participants, whereas inspection of their own ‘ugly’ body parts was given priority. In the normal control group a self-serving cognitive bias was found: they focused more on their own ‘beautiful’ body parts and less on their own ‘ugly’ body parts. When viewing other bodies the pattern was reversed: high symptom participants allocated their attention to the beautiful parts of other bodies, whereas normal controls concentrated on the ugly parts of the other bodies. From the present findings the hypothesis follows that a change in the processing of information might be needed for body exposure to be successful.

Introduction

Body-image disturbance is one of the main diagnostic features of eating disorders. Subjects with eating disorders feel fat and unattractive, and most of them show intense loathing of their bodies. They are preoccupied with their bodily appearance and show compulsive behaviors such as mirror checking and body measuring. Several authors argue that body disparagement is the key factor that underlies vulnerability to, and maintenance of, eating disorders (for example Cooley & Toray, 2001; Rosen, 1990; Tuschen-Caffier, Vögele, Bracht, & Hilbert, 2003; Stice, 2002; Wilson, Fairburn, & Agras, 1997).

It is well documented now that the body-image disturbance is not a pure perceptual aberration and is better characterized as a cognitive-emotional distortion (e.g., Cash & Deagle, 1997; Rushford & Ostermeyer, 1997; Viken, Treat, Nosofsky, McFall, & Palmeri, 2002; Williamson, 1996). Cognitive models of eating disorders propose that patients with eating disorders selectively attend to appearance cues (e.g., Hargreaves and Tiggemann, 2002; Viken et al., 2002; Williamson, Muller, Reas, & Thaw, 1999), meaning that appearance stimuli are given priority and are amplified, while other stimuli are inhibited. The selective attention to appearance cues is thought to flow from underlying knowledge structures (schemas) that filter information and direct what a subject attends to: they guide attention to, memory for, and interpretation of stimuli in ways that serve to maintain the disorder (Hargreaves and Tiggemann, 2002; Viken et al., 2002; Vitousek & Hollon, 1990; Williamson, 1996). The selective attention for appearance cues may be a maintenance factor in eating disorders.

Empirical data suggest that eating disordered patients demonstrate an attentional bias in the processing of appearance-related information. In modified Stroop color-naming tasks (see e.g., Williams, Mathews, & MacLeod, 1996), eating disordered subjects need significantly more time to name the colors of body shape-related words than they need to name the colors of neutral words (e.g., Ben-Tovim & Walker, 1991; Flynn & McNally, 1999; Labarge, Cash, & Brown, 1998). The color-naming interference for body shape-related words might mean greater attentional allocation to the body shape-related stimuli, in a very early stage of information processing. The processing resources are supposed to be automatically drawn toward the appearance-related information, even before that information has entered conscious awareness (Williams, Watts, MacLeod, & Mathews, 1988). Although the Stroop interference in eating disorders is usually supposed to reflect early preferential processing of body shape words, it has been argued lately that an increased interference in modified Stroop tasks might but need not necessarily reflect increased attention to the content of the stimuli. It is difficult to distinguish attentional effects from other non-attentional processes such as response biases or distraction as a consequence of induced emotional arousal (MacLeod, Rutherford, Campbell, Ebsworthy, & Holker, 2002). In fact, the modified Stroop task is not able to differentiate between attention, avoidance and other cognitive processes. Furthermore, because appearance cues are rather broad concepts, the modified Stroop merely offers a rough indication of (presumably attentional) processes during confrontation with appearance cues. Unsurprisingly, it shows selective interference to appearance cues in subjects who are preoccupied with their body shape. But it does not teach us what exactly happens when subjects with eating disorders are exposed to appearance cues.

In a series of elegant experiments, Tuschen-Caffier and co-workers (Hilbert, Tuschen-Caffier, & Vögele, 2002; Tuschen-Caffier et al., 2003) found that prolonged body-image exposure induced negative emotions and cognitions in eating disordered patients. The participants in Tuschen-Caffier’s studies described themselves during in vivo body-image exposure, being led by questions that the investigator read from an exposure manual. The activation of appearance schemas led to negative evaluations and lower mood in the eating disorder group. Although the eating disordered group spent significantly less time describing their waist, hips and bottom than a control group (Tuschen-Caffier et al., 2003), it is not clear whether the shorter descriptions reflect typical avoidance of these body parts, or that the participants were just feeling shame in the experimental situation, and were not able or did not want to describe the emotionally charged body parts so extensively. The use of self-reports is not the most reliable method to measure selective attention; schemas are underlying structures whereas self-reports are not only subjective to social desirability, but they are also based on the respondent’s conscious awareness.

Cognitive processes like selective attention can thus be best assessed when using non-self-report measures and more direct indicators of information processing. To examine the attentional processes related to the body-image disturbance, and overcoming interpretative difficulties (whether the bias is related to the input (attentional) or output (responses) level, inherent in the modified Stroop task and self-reports), the present study used a more direct measure of visual attention during body exposure. To further investigate the body shape-related processing bias in eating disorders, eye movement registration (electroculography) as a direct index of selective visual attention was used while eating symptomatic and normal control participants were exposed to digitalized pictures of their own body and control bodies. During the exposure, the allocation of attention was measured exactly, and after the eye movement registration, the participants identified the most “ugly” and the most “beautiful” body part of the bodies they had seen (including their own). Given that the emotional content of stimuli is supposed to capture attention, it was hypothesized that during the exposure to their own bodies, the eating symptomatic participants would, in comparison to the normal controls, allocate their attention relatively more towards their self-identified ugly body part than to their self-identified beautiful body part, thereby reinforcing their cognitive body disparagement schema. Normal controls were expected to do exactly the opposite: to focus more on their own beautiful body parts than on their own ugly body parts. The amount of attention allocated to their self-identified ugly body parts is thus expected to be larger in the eating symptomatic group than in the normal control group. When looking at other bodies, eating symptomatic subjects were expected to focus on the beautiful parts of the other bodies, whereas normal controls were not expected to show any differences in visual attention for the ugly and beautiful parts of the other bodies. Finally, it was examined whether the exposure to the bodies induced a differential mood effect across the two groups. It was hypothesized that the eating symptomatic participants would be more depressed following short body-image exposure than the normal controls.