Differentiating Old from New-but-Similar: Tracking Episodic Memory Profiles in Children with Autism

Background:  

Autism spectrum disorders (ASD) are marked by an uneven profile of memory performance; with reports of selective deficits in episodic memory, yet intact semantic memory (Shalom, 2003). A key feature of successful episodic memory is distinguishing old experiences from similar but distinct new ones, a task thought to require hippocampal pattern separation (Kirwan et al., 2012). Previous research finds that adults with ASD struggle to correctly differentiate previously seen stimuli from similar yet distinct distractors (i.e. lures), instead tending to label lures as new (South et al., 2015). Little is known about this crucial memory capacity in children with autism. 

Objectives:  

In this study, we sought to characterize the ability of children with ASD to distinguish old from new-but-similar items, in multiple contexts.

Methods:  

16 children with autism ages 8-11 and 13 age-, IQ-, and gender- matched TD peers completed the Behavioral Pattern Separation Object (BPS-O) and NEPSY-II Memory for Designs Delayed (MDD) tasks. The BPS-O is a modified visual object recognition task where items are identified as “old” (i.e. previously seen stimuli), “new” (i.e. stimuli not previously seen in the context of the experiment) or “similar” (i.e. lures that are similar but not identical to previously presented stimuli). MDD is a standardized spatial memory task, where participants distinguish between target and distractor designs to recreate a layout of 10 abstract designs after a delay. 

Results:  

On the BPS-O, children with ASD were less likely to correctly identify lures than TDs (p=0.019), and more likely to call lures “new” (p=.007). While children with ASD demonstrated unimpaired spatial location performance (p=0.251), they were more likely to incorrectly select distractors on the MDD task than TDs (p=0.039). Among children with ASD, the number of distractors selected in MDD correlated with performance on BPS-O lure identification, with increased selection of distractors related to lower performance identifying lures (r=-.57, p=.021). This same correlation was not observed in TD children (r=-.23, p=.449). 

Conclusions:  

Consistent with studies of adults on the spectrum, our results suggest that difficulty differentiating between old and new-but-similar items may be a hallmark of memory impairments in children with ASD. Crucially, spatial memory was intact in children with ASD, indicating that not all forms of episodic memory are impaired in this group. Further, the unique relationship between two distinct measures of lure detection suggests a common origin for this capacity among children with autism, potentially relating to hippocampal functioning.