Investigation into the role of value learning in core features of autism in toddlers

Yale University, New Haven CT

Investigators

Abstract

ABSTRACT. Limited selective social attention represents one of the best-replicated biomarkers in ASD. The mechanisms underlying differences in social attention in ASD remain to be identified. One of the mechanisms that contributes to attentional selection in neurotypical individuals is value learning. Value learning represents an implicit (i.e., occurring outside conscious awareness) process through which initially neutral stimuli (e.g., a face) when paired with a reinforcer (e.g., the face smiling) gains enhanced capacity to attract attention, or value-based salience. Our group has demonstrated that compared to typically developing and developmentally delayed children, preschoolers with ASD exhibit attenuated value learning when both stimuli and reinforcers are social (face, smiling face) and enhanced value learning when both stimuli and reinforcers are nonsocial (fractal, evolving fractal). If present early in life, this type of imbalance in propensity to learn about social and nonsocial stimuli may alter the experience-dependent development of neural circuitry supporting salience detection and learning and contribute to atypical development in ASD. Despite their potential importance for development of selective attention and cognition, the factors contributing to differences in value learning in ASD are not fully understood. Specifically, it is not clear if altered value learning results from children having trouble making associations between stimuli and reinforcers (value learning) whenever the stimuli are social (or nonsocial) in nature (stimulus effect) or whenever the reinforcers are social (or nonsocial) in nature (reinforcer effect) or when the stimulus and reinforcer type conspire together to give rise to specific patterns of value-based attentional biases. In the present proposal we plan to carefully manipulate the type of stimuli and reinforcers and consider key cognitive, affective, and biological covariates, to elucidate the processes through which initially neutral social and nonsocial stimuli gain privileged status in the attentional system in toddlers with and without autism. We also plan to investigate how individual differences in this value learning associate with core attentional and behavioral features of autism. The project will examine value learning in 18- to 30-month-old toddlers with ASD (n=100), typical development (TD) (n=50), and developmental delays (DD) (n=50) matched for chronological age and sex using a gaze-contingent eye- tracking implicit value learning task developed in our lab. To test our predictions, in Aim 1, effects of stimulus (social versus nonsocial) and reinforcer (social versus nonsocial) on indices of value learning will be examined in ASD, TD, and DD groups. In Aim 2, we will examine the associations between value learning indices and selective social attention measured using both a validated free-viewing screen-based (2a) and a novel live eye- tracking (2b-exploratory) task. In Aim 3, we will examine concurrent (18-30 months) and prospective (36 months) associations between indices of value learning (Aim 1) and selective social attention (Aim 2) and measures of severity of autism symptoms in the social and repetitive behavior domains.