Successful attentional set-shifting in 2-year-olds with and without Autism Spectrum Disorder

Abstract

The development of executive function is necessary for flexible and voluntary control of behavior. Deficits in executive function are purported to be a primary cause of behavioral inflexibility-a core clinical symptom-in Autism Spectrum Disorder (ASD). Attentional set-shifting has traditionally been measured with the Dimensional Change Card Sort, however, this task requires following verbal instructions. Here, we used a novel visual search task that does not require verbal instructions in conjunction with eye-tracking to test attentional set-shifting in 2-year-old toddlers diagnosed with ASD (N = 29) and chronological age-matched typically developing controls (N = 30). On each trial, a relevant and an irrelevant target were embedded in a set of feature-conjunction distractors, and toddlers were tasked with searching for the relevant target. Critically, after a set of trials the targets switched roles (i.e., the previously relevant target became irrelevant, and the previously relevant target became irrelevant). We measured visual search performance prior to and following a target switch. We found that both groups of toddlers could readily switch targets, and found strikingly similar performance between typically developing toddlers and toddlers with ASD. Our results challenge the centrality of deficits in attentional set-shifting to early behavioral inflexibility in ASD.

Fig 1. Example stimuli.

Panel (A) shows the experimental design. During the Baseline phase, toddlers were tasked with finding the relevant target (T_R). The identity of T_R reversed during the switch phase: the T_R became the irrelevant target (T_IR) and the T_IR became the T_R. Panel (B) shows the event sequence within a trial. Animations are depicted in red. Panel (C) shows the Areas of Interest (AOI) and a heat map of fixations during a Familiarization trial (top left), a Training trial (top right), and a Baseline trial (bottom center) for a TD participant.

Fig 2

Predicted pattern of results under (A) complete switching and (b) under no switching (complete perseveration) models.

Fig 3

Hit rates (A) and average fixation durations (B) for the T_R and the T_IR. Bars are 1 standard error of the mean. Average distractor (D) hit rate and fixation duration is plotted for reference. Mean and SE values are reported below.

Fig 4. Average fixation latency for the T_R and the T_IR.

Bars are 1 standard error of the mean. Mean and SE values are reported below.