3 - (SYM 126) Identifying Distinct Latent Profiles of Executive Functioning Among Adolescents and Adults with Anorexia Nervosa and Adolescent Healthy Controls

Introduction: Inefficiencies in executive functioning (EF) are implicated in the etiology of anorexia nervosa (AN). EF typically develops during adolescence and is consistent with the onset of AN. Identifying distinct EF profiles in adolescents and adults with AN may aid in identifying those at risk for a prolonged illness course and guide intervention decision-making.
 


Methods:  We conducted secondary analyses on data from adolescents (n = 559; mean age = 15.4; 92% female) and adults with AN (n = 74; mean age = 26.3; 99% female) and adolescent healthy controls (HC; n = 118; mean age = 15.1; 98% female). Specifically, we conducted latent profiles on five Delis Kaplan Executive Function System scores: Color-Word Interference conditions 3 and 4, Verbal Fluency category switching speed and total accuracy, and Trail Making Test condition 4. We then examined differences across profiles in illness severity measures: eating disorder cognitions (EDE score), BMI z-score, length of illness, and weight suppression.
 


Results: Analysis of the adolescent sample revealed the best fit for a three-profile solution (entropy = 0.83; AIC = 16419.78; BIC = 16532.26). Profile 1 (n=244) exhibited high scores in cognitive flexibility, verbal fluency, and inhibition, while Profile 2 (n=252) and Profile 3 (n=63) displayed moderate and low scores across these domains, respectively. We found no significant differences in weight suppression, EDE, or BMI z-score across profiles; however, youth in Profile 2 had a longer course of illness (p =.02). For adults, a two-profile solution demonstrated best fit (entropy = 0.81; AIC = 2198.83; BIC = 2325.55). Profile 1 (n=51) exhibited higher scores on measures of cognitive flexibility, inhibition, and moderate verbal fluency, whereas Profile 2 (n=23) displayed lower flexibility and inhibition scores and moderate verbal fluency. We saw no significant differences between profiles in markers of illness severity. Most HCs belonged to one profile (n=109) characterized by moderate levels of flexibility, verbal fluency, and inhibition, although the two-profile solution (entropy = 0.99; AIC= 3165.19; BIC= 3317.58) demonstrated the best fit. The remaining adolescent HC belonged to Profile 2 (n=9) and displayed moderate flexibility and verbal fluency but varying performance on inhibition.
 


Conclusions:  These data demonstrate different EF profiles across HC and among adolescents and adults with AN. Identifying distinct profiles may facilitate treatment personalization for those with AN who may benefit from adjunctive treatments.