Differences in Associative Learning Between Infants At Risk for Autism and Typically Developing Infants within the First 15 Months of Life

Background:  

Associative learning capabilities are intact in children with Autism Spectrum Disorders (ASDs). A novel social-object associative learning task developed in our recent work also showed intact associative learning and reduced spontaneous attention to caregivers in infants at risk for ASDs as compared to TD infants at 6 months of age (Bhat et al., 2010). We suggested that associative learning tasks may be a valuable tool to evaluate multisystem impairments of infants at risk for ASDs including learning, social, and fine motor delays.

Objectives:   In the present study, we compared spontaneous and social associative learning between infant siblings of children with autism (AU sibs) and typically developing (TD) infants over the first 15 months of life (6m, 9m, and 15m) along with developmental follow-up and autism screening at 18 and 24 months of age.

Methods:   12 AU sibs and 12 TD infants were observed during a novel social-object learning task during which attention to caregiver/social cues and object/non-social cues was contrasted. Video recorded data were collected and coded for learning based on toy activation rates during the spontaneous (i.e.; infants learn on their own) and social (i.e; infants receive caregiver feedback on which switch to activate among the two switches) contexts of the three learning phases (Baseline, Acquisition, and Extinction (Bhat et al., 2010). In addition, percent duration of gaze to objects or caregiver and patterns of bilateral and unilateral hand use were also quantified.   

Results:   Our preliminary data suggest that the majority of the infants from both groups learned the association between moving a switch and activating a cause-effect toy at 6, 9, 12, and 15 months of age. Older TD infants showed equal or better learning during the social contexts as compared to the spontaneous contexts of learning. However, AU sibs spent less time looking at caregivers and more time looking at the toy or joystick when their caregivers made no attempts to engage their attention as compared to TD infants. Moreover, AU sibs did not show increased task-appropriate toy activation patterns during the social contexts of learning as compared to the TD infants. We are waiting on future outcomes for the majority of the AU sibs. However, we expect AU sibs with poor social communication outcomes to show reduced spontaneous looking to caregivers and reduced use of caregiver cues to facilitate learning.  

Conclusions:   Our preliminary data suggest that AU sibs show reduced use of caregiver cues during associative learning as compared to TD infants. Moreover, they also have reduced spontaneous engagement with caregivers as compared to TD infants. Taken together, social-object associative learning tasks may be a valuable context to determine early markers of autism in young infants at risk for ASDs, specifically, infant siblings of children with ASDs.