Do Children with Low-Functioning Autism Disorder Have A More Extensive Declarative Memory Impairment Than Children with High Functioning Autism Spectrum Disorder? Implications for the Medial Temporal Lobe

Background: We have hypothesized that the language and learning impairments that distinguish autistic disorder (AD) from Asperger syndrome (AS) and other forms of high-functioning autism (HFA) most commonly result from an impairment of semantic memory additional to the impairment of episodic memory known to occur in all forms of autistic spectrum disorder. More specifically, we hypothesise that whereas relational memory (critically dependent on hippocampally mediated recollection) is impaired across the spectrum, single item memory (critically dependent on a sense of familiarity mediated by perirhinal and/or entorhinal medial temporal lobe cortex) is impaired only in individuals with AD.

Objectives: The objective of the research to be reported is to test this hypothesis.

Methods: Study 1 tested single item recognition in adolescents with AD using shape and pattern recognition tasks, and assessed the relation between performance on these tasks and lexical-semantic knowledge as assessed by the Pyramids and Palm Trees test and by the Similarities and Vocabulary subtests from the Wechsler scales.

Study 2 tested single item and relational memory in children with AS or HFA using a shape recognition task and an action recall source memory task.

Study 3 (ongoing) tests single item and relational memory in adolescents with AD and young children with AS or HFA using the shape recognition and action recall tasks used in Study 2, plus forced choice recognition tests developed by Migo, Mayes et al. specifically to discriminate between the process of familiarity that is critical for single item memory and the process of recollection that is critical for relational memory. Two comparison groups are included: a young typically developing (TD)  group and a group of intellectually disabled adolescents without autism, equated with the AD group for age and ability.

Results: Study 1 showed that single item recognition was impaired in the AD group relative to an ability-matched TD comparison group but not to an age and ability-matched intellectually disabled comparison group. However correlation between recognition and lexical-semantic ability was specific to the AD group.

Study 2 confirmed the prediction that single item recognition would be unimpaired in an AS/HFA group relative to an age and ability matched TD group (reflecting unimpaired single item memory), but that source memory would be impaired relative to the TD group (reflecting an impairment of relational memory).

Results of Study 3 are not yet available.

Conclusions: Findings from Studies 1 and 2 provide preliminary support for the hypothesis.  However, the results of Study 1 leave open the possibility that impaired single item memory may not be specific to AD, and Study 2 (designed to assess novel methods of measuring single item and relational memory separately) did not include an AD group. The results of Study 3 should resolve both these limitations.