Personality Subtypes and Clinical Symptoms in Undergraduates with Histories of Anorexia Nervosa, Atypical Anorexia Nervosa, and Bulimia Nervosa

Objective: Empirical clustering has identified three distinct personality profiles (over-controlled, under-controlled, and low psychopathology subtypes) among eating disorders, which have predicted clinical outcomes. No prior data-driven subtyping studies have included atypical anorexia nervosa (AAN); thus, it is unclear whether the same personality subtypes emerge within this diagnostic group and similarly predict outcomes. Methods: In this study, undergraduates (N=2384) completed surveys assessing personality, eating disorder, and other psychological symptoms at two timepoints. Participants were classified as having histories of AAN, anorexia nervosa (AN), bulimia nervosa (BN), and no eating disorder. Latent profile analysis derived subtypes from personality measures. Regression models determined whether personality subtypes were differentially associated with eating disorder diagnosis and if personality subtype, diagnosis, and their interactions impacted symptoms. Results: Results supported a 3-class solution: Over-controlled (n=243); Under-controlled (n=238); and Low Psychopathology (n=1903) profiles. The Under-controlled subtype predicted AAN and BN diagnoses (ps<.010). Across diagnoses, Time 1 Under-controlled subtype was associated with greater Time 1 depression (p=.017) and alcohol use disorder (p<.001) symptoms and Time 2 binge eating (p=.022). No personality subtype by diagnosis interactions emerged. Conclusions: When including AAN participants, derived personality profiles mirrored prior findings. AAN diagnosis was associated with the Under-controlled subtype; thus, affective regulation may warrant attention in this group. Although personality subtype predicted clinical symptoms, relations did not differ across diagnoses. Integrating personality type into clinical decisions for eating disorders, including AAN, rather than solely relying on diagnostic categories, may inform more precise interventions.