Altered Modulation by Communicative Signals On Cognitive Performance in Children with Autism

Background:  In typical development, eye contact facilitates performance on specific socio-cognitive tasks. For example, typically developing infants tend to process others’ communicative signals more accurately if preceded by direct gaze, signaling that the infant is the intended addressee. Preliminary evidence suggests that eye contact may be less beneficial for children with Autism Spectrum Disorder (ASD).

Objectives: Here we asked whether facilitation could be observed in typically developing children during non-social cognitive tasks. Moreover, we predicted that modulation by communicative signals on cognitive performance would be altered in children with ASD.

Methods: We assessed the ability of neurotypical children (n = 27) and children with ASD (n = 12) to repeat series of digits in a modified digit span task adapted from a gold standard and widely used intelligence test. The children were 4 to 10 years old, and the groups were matched on full scale IQ. The test administrator, sitting in front of the child, first established the child’s maximum digit span by gradually increasing the difficulty of the task. When established, series of difficult (the child’s maximum digit span) and easy (2 digits) trials were given in blocks of four. On every second trial the administrator either looked at the child (the gaze condition) or looked down towards the test protocol (the no-gaze condition). Performance during difficult trials was evaluated as a function of group and condition. In addition, we measured the children’s eye movements during the task to evaluate attentional differences, using live eye-tracking (Tobii TX300).

Results: As predicted, we observed a differential modulation of the gaze/no-gaze manipulation on performance (digit span) in the two groups (p < .05). Direct gaze increased performance in typically developing children but not in children with ASD. Live eye-tracking data showed that both groups looked away from the face of the test administrator during thinking and answering phases of the interaction.

Conclusions: This study suggests that in typical development, positive effects of eye contact can be observed even outside the socio-cognitive domain. Moreover, it indicates that eye-contact is less beneficial for children with ASD than for children with typical development. The eye-movement data suggest that both groups modulated their eye contact adaptively as a function of the task demands. In addition to their potential clinical significance, the results have implications for our understanding of the interplay between social context and cognition in ASD during the first years of life.