EXPLORING LONELINESS THROUGH COGNITIVE PROFILES: UNDERSTANDING EXECUTIVE FUNCTION AND MENTALIZING IN AUTISTIC AND NON-AUTISTIC YOUTH

Abstract

Adolescence is a transitional period of development where rates of loneliness increase. Autistic youth are more likely to experience loneliness than their non-autistic peers, making adolescent youth on the spectrum particularly vulnerable to poor social outcomes. Difficulties with executive function (EF) and mentalizing are well-documented in autistic youth; these challenges may contribute to social difficulties. Given the heterogeneous presentation of autism symptoms, particularly executive functioning and mentalizing abilities, we conducted a latent profile analysis (LPA) to determine if distinct cognitive profiles emerge around 2 executive functioning and 2 mentalizing variables. We then examined whether these profiles are associated with loneliness outcomes. Results demonstrate that none of the 4 cognitive variables of interest: Flanker, DCCS, Faces, and STOMP are significantly related to loneliness when variance in age, overall IQ, and diagnostic group are accounted for, thus suggesting the need for a heterogeneous approach to understand cognitive skills. LPA analysis revealed that a 3-profile solution best fits the data. The profiles were significantly related to loneliness outcomes, with the profile characterized by poor flexibility reporting the highest rates of loneliness. However, this result did not hold when we controlled for diagnostic group membership. When diagnostic group membership was accounted for, Profile 1 (the higher cognitive performance) and Profile 2 (the average cognitive performance) were marginally significantly different on loneliness outcomes. These results, although preliminary, demonstrate that a profiles approach to cognitive skills provides more explanatory power to loneliness outcomes compared to a linear approach.