Is Language Regression Related to a Theory of Mind?

Background:

 Regressive autism is characterized by the loss of acquired language and social skills at approximately 18 to 24 months in addition to a diagnosis of an autism spectrum disorder (ASD). Many studies have attempted to explain the different developmental trajectories between children with regressive and non-regressive autism; however, the phenomenon of regressive autism continues to be poorly understood. Because deficits in theory of mind are characteristics of autism, it is important to determine if and how children with regressive autism demonstrate differences in theory of mind compared to children with ASD but without regression. To our knowledge, this is the first study to examine theory of mind in relation to regression in children diagnosed with ASD.

Objectives:

The primary goal of the present research was to examine whether children with ASD who had and had not experienced early language regression differed in their performance on traditional theory of mind tasks.

Methods:

Thirty-three children with ASD from a national study also participated in this substudy. Families were mostly Caucasian, middle-class, and well-educated. ASD diagnoses were made independently. Based on the ADI-R, 18 children had experienced early word loss; 15 had not. At a mean age of approximately nine years, children completed a battery of theory of mind tasks: appearance-reality tasks (including nonverbal), first-order false-belief tasks, and positive and negative belief-desire tasks. Each task was administered twice. For each set of theory of mind tasks, children received a score of zero for failing both assessments, a score of one for passing one of the two assessments, and a score of two for passing both assessments. The number of times a child succeeded on both assessments across the various tasks was used to create a composite score with a maximum value of six.

Results:

After examining receptive language as a possible covariate, ANOVA was used to determine whether children with language regression performed differently from those without language regression on each set of theory of mind tasks and the composite score. Results indicated that children who reportedly had language regression performed better on the nonverbal appearance-reality questions (M = 1.57, SE = .15) relative to children who had not experienced language regression (M = 1.12, SE = .13; F(1, 27) = 4.73, p < .03); other  task scores and the composite score did not differ between groups.

Conclusions:

These study findings suggest that the presence of early language abilities may facilitate performance on some aspects of theory of mind, particularly those requiring nonverbal responses. One possible interpretation is that the children with early access to language may have been able to learn about certain aspects of communication before the regression occurred, whereas children with persistent deficits in language from early in life may not have had this opportunity. Although replication with a larger sample is needed, this work contributes to the literature on cognitive differences in children with ASD by indicating that variability in performances across tasks exists as a function of the presence or absence of regression early in life.