(PS8-A24) Ambivalence Model of Craving: Latent Profile Analysis (LPA) of Food Approach and Avoidance Inclinations and Associated Eating-related Constructs

With treatment outcomes remaining poor, the need to identify novel mechanisms that may contribute to the development and maintenance eating disorders is critical for treatment research to advance. The Ambivalence Model of Craving (AMC) proposes four motivational states that represent combinations of competing desires to consume (approach) and not consume (avoid) substances that are associated with different patterns of use. The current study tested whether a four profile model observed for substance use would be supported for food craving. Associations between food cravings and eating disorder symptoms and eating and shape expectancies also were examined. A total of 407 undergraduate participants (54.3% female; M_Age = 20.99; M_BMI = 25.05; 47.17% Non-Hispanic White) completed an online cross-sectional survey measuring food approach and avoidance craving (FAAQ), eating disorder symptoms (EDDS for DSM-5), and thinness and eating expectancies (EEI; TREI). Latent profile analysis (LPA) was used to identify motivational profiles, and multiple linear regressions were used to examine associations between food craving, and eating disorder symptoms, and expectancies.Although four profiles emerged as the best fit to the data, the mean approach and avoidance scores were comparable in each profile (i.e., high-high, high/moderate-high/moderate, low/moderate-low/moderate, low-low). Post-hoc sensitivity LPAs were conducted with participants at high risk for an eating disorder (n = 186; 45.7%) with similar patterns emerging. Consistent with previous findings, greater approach (exp(B) = 1.13, p < .001) and avoidance cravings (exp(B) = 1.20, p < .001) were associated with more eating disorder symptoms. Avoidance craving was associated with expectancies about thinness/restriction and that eating leads to being out of control (ps < .001). Approach craving was associated with expectancies that eating manages negative affect, eating as a reward, eating leads to being out of control, and eating alleviates boredom (ps < .003). Findings did not support a four profile model of food cravings for either the full or high-risk samples, suggesting that approach and avoidance food cravings may have an additive – not interactive – effect. Consistent with AMC theory, approach and avoidance food cravings were associated with eating and thinness expectancies. Future research should investigate the temporal relationship among food craving, eating and thinness expectancies, and eating disorder symptoms to understand how these cognitive-behavioral factors develop over time in order to identify effective eating disorder treatment targets.