Semantic Representations In Asperger Syndrome and Nonverbal LD

Background:  In 1995, Klin and colleagues described a convergence of features in individuals with Asperger Syndrome (AS), an Autism Spectrum Disorder, and Nonverbal Learning Disability (NLD), a learning disorder defined by deficits in visual-spatial organization. Only recently has research begun to systematically explore the nature of these similarities, and little work has investigated language. Individuals with AS and NLD have well-documented pragmatic impairments. It is possible, however, that pragmatic impairments are due in part to underlying semantic difficulties, which appear as subtle in assessment situations but are more conspicuous socially.

Objectives:  The purpose of the present study was to investigate whether individuals with AS and NLD have semantic representations that are a) less precise and b) less rich than semantic representations as expressed by peers without disabilities.

Methods: Participants were typical adults and those with a community diagnosis of AS or NLD who had successfully completed at least one course at a Canadian college or university. Conversational samples obtained during a semi-structured interview and individual’s responses to test stimuli on the Vocabulary and Similarities subtests of the Wescheler Adult Intelligence Scale (WAIS) IV were coded for Mean Length of Utterance (MLU).  Samples were also coded for MLU-Coherence, a measure of the number of utterances required to express single ideas that were relevant to the questions asked of the participants. MLU for irrelevant responses and their frequency were also compared by group. 

Results:  Participants with AS and NLD had more variable MLU scores than did their typical peers. For all participants with a diagnosis, MLU scores were highest for interview questions about strengths, weaknesses, and interests. These scores were longer for AS than NLD participants, but this trend did not absolutely distinguish the two clinical groups from each other.  Both groups produced more words to express single ideas, whether or not they were in response to the questions or test stimuli presented. Finally, the clinical participants produced more irrelevant utterances in response to the Vocabulary subtest stimuli, although scaled scores on the test were not significantly different by group.

Conclusions:  Results suggested that adults with AS and NLD may be distinguished from typical adults by subtle weaknesses in semantics. Longer, less precise utterances were produced by the adults with diagnoses in semi-structured interviews and on psychometric testing. Richness of semantic representations also appeared to be affected; responses for participants with AS and NLD had more instances of word repetitions than did controls.  Differences between clinical groups were seen for a question about interests. Overall, findings implicated similarities over differences between adults with AS and NLD; however, they also suggested a potential source of difference to be researched further.