Because few studies have identified biological factors that predict the persistence of eating pathology, we tested the hypotheses that elevated responsivity of brain regions implicated in reward valuation to thin models and high-calorie binge foods would predict the persistence of eating pathology. We analyzed data from 146 women (Mage = 21.87 ± 3.81) with threshold or subthreshold anorexia nervosa, bulimia nervosa, binge eating disorder, or purging disorder who completed functional magnetic resonance imaging scans assessing neural response to thin models and binge foods at baseline, were randomized to two eating disorder treatments or a waitlist control condition and completed diagnostic interviews that assessed change in symptoms over 2.5-month follow-up. Elevated activation in regions associated with memory (parahippocampal gyrus r = .38; hippocampus r = .25) and with reward valuation and emotional salience (amygdala r = .35) in response to thin versus average-weight models predicted future persistence of an eating disorder symptom composite (all analyses controlled for treatment condition). Neural response to high-calorie binge foods did not predict the persistence of eating disorder symptoms. There was no evidence that either treatment moderated the relation of baseline neural responsivity to thin models to future persistence of eating disorder symptoms, though power for these post hoc analyses was limited. Results suggest that overvaluation of the thin ideal increases the risk for the persistence of eating disorder symptoms but provided no support for the hypothesis that overvaluation of high-calorie foods would predict symptom persistence. Results imply that treatments should seek to reduce the overvaluation of the thin ideal. (PsycInfo Database Record (c) 2023 APA, all rights reserved).
Trial registration: ClinicalTrials.gov NCT03261050.