fMRI Correlates of Relational Memory Difficulties in Autism Spectrum Disorder

Background: It is now well established that individuals from across the Autism Spectrum exhibit a rather unique pattern of memory strengths and weaknesses. Their free recall performance is usually attenuated, they experience difficulties in remembering the personally experienced past (i.e. episodic memory) and they tend not to use organisational strategies efficiently to facilitate memory. By contrast, they tend to experience few problems on tests of recognition or cued recall and their memory for factual knowledge (i.e. semantic memory) is generally preserved and often even enhanced. Although the neural basis of this memory profile remains largely unknown, it has for long been speculated that atypicalities in the functioning of the Hippocampus and areas of the Frontal Lobes may be responsible.  

Objectives: To assess, for the first time, the functional integrity of Medial Temporal and Frontal Lobe memory systems in ASD through the use of an established fMRI paradigm.

Methods: Following a paradigm developed by Addis & McAndrews (2006), we asked participants to learn a series of word triads in a 3T scanner. Word-triads were presented for 6 seconds (with 4-8s jittered intervals) and consisted of a category name and two nouns. Either none, one or both of these nouns were examples of the category name and participants were required to indicate the relevant number of category links whilst trying to remember the triads. Immediately after the encoding scan, participants completed a two-alternative forced choice recognition test in a quiet room. In this test, studied triads were presented alongside new triads that included a novel noun, and participants indicated whether they clearly ‘remembered’, simply ‘knew’ or ‘guessed’ which of the two triads they had seen during the earlier scan.

Results: So far, 10 ASD and 10 age and IQ matched typically developed individuals have been tested (12-16 per group are planned). Behavioural results, so far, indicate equivalent forced choice recognition accuracy in the two groups although reports of ‘remembering’ are significantly attenuated and ‘guessing’ responses augmented in the ASD group. Preliminary analyses of the fMRI data confirm Addis & McAndrew’s (2006) observation of successful encoding effects in the Hippocampus and Inferior Frontal Gyrus (IFG) and although data collection is ongoing the results also suggest that these effects may be attenuated in the ASD group (particularly in the Hippocampus). Our final analysis will assess the extent to which successful encoding effects in the Hippocampus and IFG (and potential group differences thereof) are modulated by the number of category-links in encoded word-triads and by the subjective reports of ‘remembering’, ‘knowing’ and ‘guessing’ of participants during the recognition test. Based on the available literature, we predict atypicalities particularly in hippocampal function in ASD but also in the connectivity between the Hippocampus and Inferior Frontal Lobe.

Conclusions: We will discuss our observations in the context of current cognitive and neurobiological theories of memory in ASD, with a particular focus on the notion that Medial Temporal Lobe dysfunction may constitute a widespread facet of the neuropathology underlying this disorder.