Food Images and Attentional Biases related to Eating Disorder Pathology in Nonclinical Sample

Previous research has observed that individuals diagnosed with eating disorders exhibit attentional biases toward food stimuli which are affected by food composition (high calorie vs low calorie foods). However, these attentional biases have been found to vary by subtype of eating disorder and have not been studied extensively in nonclinical populations. In order to explore if these same attentional biases are present in nonclinical populations, 36 Union College students were recruited, restricted to women, with age M = 20.09 SD = 1.11 years and BMI M = 24.79 SD = 6.76 kg/m². Participants were told they were taking part in an advertisement study and were visually presented with 30 image pairs. Using the Eyelink1000 eye tracking system, participants’ gaze was tracked when presented with different combinations of high calorie and low calorie, high calorie and neutral (e.g., non-food product advertisements), and low calorie and neutral image sets. After each image pair, participants rated how likely they would be to buy each product, how exciting the product images were, and how appetizing the food image(s) was/were. Participants were then asked to complete an advertisement product questionnaire which contained distractor consumer questions and the binge eating, cognitive restraint, purging, restricting, and excessive exercise subscales of the Eating Pathology Symptoms Inventory (EPSI). Participants’ initial fixations, dwell time, run count, and initial saccades were examined to assess initial orientation, initial avoidance, and attentional maintenance. EPSI score was found statistically significant in predicting initial fixation count of high calorie images, such that higher EPSI scores were associated with a greater amount of initial fixations. Based on our results, there were trends suggesting that, for neutral and low calorie images, higher EPSI scores may be associated with lower initial fixation count and duration and low calorie foods with higher fixation durations; however, the current study was underpowered to detect small and medium effects. In predicting all other eye tracking behaviors (run count, initial saccade, and dwell time), EPSI score was not a statistically significant predictor. Further research is needed to determine if initial fixation count of healthy images might be a useful tool in predicting eating disorder risk in nonclinical populations. Additional limitations and future directions will be discussed.