Memory for Items, Contexts and Relations in Adults with Autism Spectrum Disorder

Background:

Two prominent accounts of memory difficulties in adults with Autism Spectrum Disorder (ASD) suggest difficulties in episodic memory (Bowler et al., 2007) and complex memory (Williams et al., 2015). Both accounts implicate problems remembering relations between items or between items and their context (Bowler et al., 2015), and difficulties in these areas may be the result of problems in binding items to one another, in binding items to their context, or in forming a three-way relation between two items and their context (ternary relations; Halford, 1992). Recent research has shown a developmental progression in typical children’s ability to recombine items and context as well as to form relations between three components such as two items and their context (Yim et al., 2013). Forming and retaining connections between items and context have recently been identified as the prerequisites for the development of episodic memory (Yim et al., 2013), and they may be a component of autistic individuals’ difficulties in this area.

Objectives:

This study aimed to use Yim et al.’s (2013) paradigm to identify the components in the memory binding process that are most difficult for ASD individuals and to point to potential compensatory mechanisms.

Methods:

Fifty-four pairs of ASD and typically developing (TD) individuals matched on gender, chronological age (CA; M_CA = 43.19), and intelligence quotient (IQ; M_FIQ = 111) were tested on one of three associative learning tasks that varied in the number of bindings between items and context to be formed for task success. One group of participants studied lists of word pairs in the form of AB CD, where each list contained different pairs of items such as tree-shoe (AB) and strawberry-sofa (CD). A second group of participants studied lists of the form AB AC, which recombined familiar items from the first list with new items in the second list, for example, bicycle-cup (AB) and bicycle-nail (AC). A third group studied AB ABr lists, where the first studied list (AB) contained pairs such as door-glass and flag-balloon, and the second set (ABr) presented a recombination of item-pairs from the first list introducing no new items, for instance, door-balloon. Each list was accompanied by a specific contextual cue.

Results:

Overall accuracy was lower for persons with ASD compared to TD individuals on all tasks (F(1,102) = 12.49, p= .001). A multinominal model (Kim et al., 2013), using error rather than accuracy data showed that ASD participants named more items that were unrelated to the studied items, they more often confused item-item relations, and they showed particular difficulties in binding item-pairs to their contextual cue. These last two findings reflect ternary binding difficulties. However, ASD participants performed better than TD individuals in distinguishing between the two studied lists, i.e., identifying whether individual items belonged to a cue or a target list.

Conclusions:

This study confirms particular difficulties in ASD in forming ternary relations in memory. However, it also suggests that persons with ASD rely on their superior capacity to remember item-list relations as a compensatory mechanism.