3 - (IOP 34) Affect, Food Craving, and Momentary Demand as Reinforcement Mechanisms of Loss-of-control Eating: an Ecological Momentary Assessment Approach

Loss-of-control eating (LOCE) is a perceived inability to stop or to refrain from food consumption. Affect, food craving, and reward dysfunction are all implicated in LOCE, although the exact nature of the relationship has yet to be fully captured. Some reward mechanisms make positive affect (PA) a state-level vulnerability, but negative affect (NA) is consistently a LOCE predictor. Some research indicates that craving precedes affect changes but other findings suggest affect bolsters craving prior to LOCE. Given the labile nature of affect, reward valuation, and craving, a 10-day ecological momentary assessment design probed the associations among these LOCE contributors within- and between-subjects. A national sample of U.S. adults completed a screening survey and clinical interview for eligibility, including at least 1 LOCE episode weekly. They responded to 6 surveys sent daily at random times, and initiated surveys before and after eating. PA and NA were rated Positive and Negative Affect Schedule items. An ambulatory Food Purchase Task (FPT) was developed to assess momentary food demand (intensity, breakpoint, and Omax). Hypothetical purchase tasks assess the reward value of a commodity. The FPT prompted selection of a snack food participants “want the most right now”. An image was shown of a single 20-gram serving (e.g., two Chips Ahoy cookies). Intensity was assessed via the number of servings one would eat for free, breakpoint via the maximum amount one would pay for one serving ($0 - $50), and Omax via the maximum total amount to eat as much as desired ($0 - $100). Craving was probed with a single Likert-type item. An adapted item from the Eating Disorder Examination Questionnaire asked if one lost control over eating (LOCE) if eating was endorsed. Higher PA and NA at Time 1 were hypothesized to predict Time 2 food demand associated with craving, leading to Time 3 LOCE (Time 1 affect → Time 2 food demand → Time 2 craving → Time 3 LOCE).
Participants (N = 78, 87% female) were 72% White, 10% Black/ African American, 9% Asian, 9% American Indian/ Alaskan Native, and 1% Other. Seven percent were multi-racial. 53% did not meet eating disorder criteria, 37% met binge eating disorder criteria, and 9% bulimia nervosa criteria. Within-subjects NA led to higher food demand, increasing craving, and predicting LOCE. Heightened PA led to LOCE but only by way of craving. Interestingly, between-subjects the relationship between craving and LOCE was negative. NA was also associated with future LOCE by way of craving but not demand, although NA was associated positively with breakpoint. PA had no associations. It may be that for a sample characterized by frequent LOCE, these processes are only interpretable at a momentary level. It is also possible there are differences attributable to pathology duration. In incentive sensitization theory, “liking” (enjoying) and “wanting” (craving) are separate neurological paths, with “liking” more relevant early in recurrent LOCE and "wanting” asserting more influence maintaining pathology over time. While FPT have been used in eating research prior, this study is the first published with an ambulatory design, allowing multilevel analyses. Results show a need to consider food demand and related metrics at the momentary level.