Mentalizing Knowledge of the Self Versus Others: Distinct Clinical Predictors of Social Maladjustment in Children with Higher Functioning Autism

Background: Prior research has documented that self-referenced processing enhances word memory in comparison to other semantic forms of word processing in neurotypical individuals; a phenomenon coined as the self-referenced memory (SRM) effect (Symons & Johnson, 1997).  However, individuals with autism fail to exhibit this enhanced recognition memory for self-referenced words.  For example, Henderson et al. (2009) reported that higher functioning children with autism (HFA) failed to show the standard SRM effect and instead showed comparable recognition of self- and other-processed words.  Additionally, better self-referenced processing was inversely related to social symptom severity.  This ability to mentalize self-knowledge is an integral component of social cognition, and its impairment in autism may be distinct from theory of mind (ToM) abilities, or the ability to mentalize about others (Hobson et al., 2006).  Although both skills are necessary for the development of an understanding of the relations between self and others, impairments in these distinct aspects of mentalizing may lead to different social maladjustments.     

Objectives: This study investigated both SRM performance and ToM, as two distinct constructs, and their relations to self-reported social maladjustment in HFA children.

Methods: Participants were 94 HFA children (IQ > 70) aged 8 to 16, with a clinically confirmed diagnosis.  A SRM task was completed in which they read a list of words and decided whether the word described them (self-condition), a fictional character, or contained a certain number of letters.  They then identified familiar words from a longer list.  Participants also completed The Children’s Eyes Task and the Behavior Assessment System for Children, 2^nd-Edition.  Independent variables were recognition performance for words in the self-condition (d′ self) and total scores on the eyes task.   

Results: Bivariate analyses indicated that SRM and ToM were correlated, r = .34, p < .05. Controlling for verbal IQ, age, and symptom severity, results of regression analyses indicated that d′ self, but not ToM, predicted self-reported School Problems, R² = 0.44, adjusted R² = 0.13, F(5, 57) =  2.78, p < .05, β = -.28, t(57) = -2.05, p < .05.  Similarly, d′ self predicted Internalizing Problems, R² = 0.40, adjusted R² = 0.16, F(5, 56) =  2.14, p = .07, β = -.33, t(56) = -2.28, p < .05, but ToM did not.  In contrast, Personal Adjustment was predicted by ToM, R² = 0.44, adjusted R² = 0.13, F(5, 57) =  2.54, p < .05, β = -.48, t(57) = -3.12, p < .01, but not  d′ self.

Conclusions: This study provides support that self-processing and ToM are distinct indices of mentalizing that may impact specific aspects of social adjustment in HFA children.  SRM may uniquely predict school problems because poor self-awareness may impede social learning within the complexity of a school setting.  Personal adjustment, particularly during late childhood and adolescence, may be particularly dependent upon the ability to understand others’ intentions and emotions, which were assessed in the Eyes task. Results will be discussed in the context of theories of self- and other- processing as they relate to social learning and development.