1 - (SYM 124) the Buffet Challenge: A Behavioral Assessment of Eating Behavior in Adolescents with an Eating Disorder

Introduction: Eating disorders (ED) are characterized by rigid eating behaviors and abnormal mealtime behaviors. Laboratory meal paradigms are a rigorous approach to accurately record food intake, choice, and ED behaviors. Here, we describe the development and preliminary psychometric results of the Buffet Challenge, a behavioral assessment of eating flexibility in youth with ED. We hypothesized that buffet intake variables (e.g., total intake, caloric density, and macronutrients) would be associated with higher ED psychopathology, higher in remitted vs. not remitted patients and would predict early weight gain and good prognosis.

Method: Participants were 56 adolescents (48 girls, 8 boys) with anorexia nervosa. At baseline, youth were 15.37 yrs with a mean BMI z-score = -0.68. They completed the Buffet Challenge at baseline, mid treatment and end of treatment. The task offers a wide range of food items with varied caloric and macronutrient compositions. We weigh food served and consumed to calculate caloric and macronutrient (i.e., fat, protein, and carbohydrate) intake and video record youth to code ED behaviors (e.g. tearing food, nibbling, label checking).

Results: All adolescents were willing to complete the Buffet Challenge at all time points and no adverse events were recorded. The most common ED behaviors across all timepoints were staring at food, fidgeting, and inappropriate napkin use. At baseline, higher mother-reported ED symptoms were associated with lower caloric density consumed (r = -.33, p = .02). At end of treatment, we found large, non-significant effect across remission groups for caloric density (H(2) = 4.96, p = .084, h² = .16, 90%CI[-.03, .56]), where caloric density was lowest in those not remitted compared to those partially/fully remitted. We observed medium, non-significant effect sizes between groups for % intake from fats (t(16) = -1.58, p = .13, Hedge’s g = -0.79, 95%CI[-1.80, 0.24]), where youth referred to a higher level of care consumed less of their meal from fats compared to those who were not.
Discussion: Our pilot of this task demonstrates its acceptability, although investigation of our hypotheses was hindered by missing data during COVID. Effect sizes suggest that caloric density may be a marker of recovery for future studies. Limitations of this study include foods that may not reflect the diverse cuisines of youth from other cultures and locations. Strengths include modifications to reflect normative adolescent eating and alignment with treatment principles and the variety of variables collected.