EEG Markers of Learning from Joint Engagement in Toddlers at High Risk for Autism Spectrum Disorder

Background: Joint engagement (JE), defined as periods of joint attention shared between an adult and infant (Carpenter et al., 1998), emerges between 9-15 months of age (Adamson et al., 2009). JE promotes language, cognitive development and social skill acquisition (Mundy et al., 2007). Children with autism spectrum disorder (ASD) exhibit early social communication deficits including in joint attention (Mundy et al., 2007). Typically-developing toddlers have shown evidence of object discrimination based on JE using electroencephalographic (EEG) methods (Hutman et al., 2015); however, this has not been investigated in toddlers showing early signs of ASD. Examination of object discrimination based on JE in toddlers showing signs of ASD may improve understanding of social communication deficits, which can inform JE-based interventions.

Objectives: (1) To investigate if JE modulates object discrimination in toddlers (12-24 months) at high risk for ASD (HR, n=16) compared to typically-developing toddlers (TD, n=11) using a novel EEG paradigm that couples a live interaction exposure with an EEG test phase. (2) To identify if quality of JE between child and experimenter during the exposure phase relates to EEG markers of learning.

Methods: Toddlers were deemed HR if they scored within the mild-to-severe concern range on the Autism Diagnostic Observation Schedule-Toddler Module. During a 4-minute exposure, the experimenter presented two sets of toys in a social and non-social manner. Child behavior was coded to evaluate quality of JE with the experimenter. Participants then passively viewed images of toys presented during exposure while high-density EEG was recorded (128-electrode, EGI Inc.). The variable of interest was frontal mean Nc (400-800ms), thought to be a marker of attention.

Results: (1) A 3-way mixed-design ANOVA examined main and interaction effects of condition, region, and group. There was no effect of condition by group, F(1,25)=.004, p=.95. There was an interaction effect of condition by region, F(2,24)=7.664, p=.003. Post-hoc tests revealed significant differentiation of conditions in the middle region, t(26)=-2.07, p=.049. (2) Correlations were then conducted to evaluate associations between EEG markers and quality of JE. Time spent looking at the examiner during the social condition correlated with Nc absolute difference value in the middle region, r(27)=.49, p=.01.

Conclusions: Both HR and TD toddlers show increased familiarity to objects presented with JE. This suggests that toddlers showing early signs of ASD were able to learn via JE with an examiner in a comparable manner to TD children. Given that early intervention for ASD often relies on learning in a social context, these result are promising. The level of object discrimination based on JE, however, varied with regard to social behavior during the exposure phase. Specifically, toddlers who spent more time looking at the examiner during JE showed higher levels of discrimination, indicating salience of object learning may increase with eye contact. Further characterizing child behaviors that relate to learning based on JE may inform intervention treatment targets and predict response to treatment. Future research will aim to relate EEG markers of learning to gains in an intervention focused on improving JE, and to diagnostic outcome of HR toddlers.