Symposia
Eating Disorders
Kira Venables, B.A. (she/her/hers)
Clinical Research Coordinator
Virginia Commonwealth University
Richmond, Virginia, United States
Hannah Heintz, B.A. (she/her/hers)
Graduate Student
Virginia Commonwealth University
Richmond, Virginia, United States
Ariana Bazzi, B.S. (she/her/hers)
Research Assistant
Virginia Commonwealth University
Richmond, Virginia, United States
Amanda Makara, B.S. (she/her/hers)
Graduate Student
Virginia Commonwealth University
Richmond, Virginia, United States
Adanya D. Johnson, B.S.
Clinical Research Coordinator
Virginia Commonwealth University
Richmond, Virginia, United States
Kelsey Hagan, Ph.D.
Assistant Professor
Virginia Commonwealth University
Richmond, Virginia, United States
Ann F. Haynos, Ph.D. (she/her/hers)
Assistant Professor
Virginia Commonwealth University
Richmond, Virginia, United States
Research using empirical clustering techniques has identified three distinct personality profiles in eating disorders: Overcontrolled (OC; i.e., emotionally/behaviorally inhibited), Undercontrolled (UC; i.e., emotionally/behaviorally dysregulated), and Low Psychopathology (LP). Studies in anorexia nervosa (AN) and bulimia nervosa (BN) have found the OC group to be associated with greater restrictive eating and the UC group with more binge eating and purging and elevated depression, anxiety, and eating disorder symptoms. As no prior data-driven subtyping studies have included atypical AN (AAN), it is unclear if AAN is associated with a particular profile or interacts with personality uniquely to impact symptom expression. Here, we identified personality profiles in undergraduates with a history of AAN, AN, and BN and longitudinally examined associations between personality profiles, diagnoses, and clinical symptoms.
Undergraduates (n = 2389) completed a survey assessing personality, mood symptoms (i.e., depression, anxiety), and eating disorder symptoms (i.e., fasting, driven exercise, binge eating, purging) in their freshman, sophomore, and junior years. Latent profile analysis compared the fit of 1- to 6-profile models derived from Big Five personality factors. Logistic regression models determined whether personality profile predicted AAN, AN, and BN diagnosis. Regression models examined the influence of class membership, eating disorder diagnosis, and their interaction on mood and eating disorder symptoms. Results supported a 3-class solution: UC (n = 238; 10.0%); OC (n = 243; 10.2%); and LP (n = 1903; 79.8%) profiles. The UC profile significantly predicted BN diagnosis (p = .001) and trended towards predicting AAN diagnosis (p = .067). None of the classes predicted AN diagnosis (ps > .11). Across diagnoses, Time 1 UC profile predicted greater Time 2 depressive symptoms (B = .64, p =.046) and binge eating (B = .57, p = .022) than the LP profile. Personality profiles predicted eating disorder and mood symptoms across diagnoses. Profiles were heterogeneous within diagnoses, including AAN, suggesting they may better capture transdiagnostic clinical features versus discrete diagnoses. In general, results suggest that orienting treatment of AAN and other EDs around personality clusters and specific symptom constellations may be more fruitful than making intervention decisions according to diagnosis. However, further attention may be warranted to address affective components of AAN.