Peer Interaction and Loneliness in High-Functioning Children with Autism

N. Bauminger, C. Shulman and G. Agam, Journal of Autism and Developmental Disorders, Volume 33, No 5, October 2003, 489 - 507.

Social interaction with peers and the understanding and feelings of loneliness were examined in 18 high-functioning children with autism and 17 typically developing children matched for IQ, chronological age, gender, and maternal education. Observations were conducted on children's spontaneous social initiations and responses to their peers in natural settings such as recess and snack time, and children reported on their understanding and feelings of loneliness and social interaction. Overall, children with autism revealed a good understanding of both social interaction and loneliness, and they demonstrated a high level of social initiation. However, they spent only half the time in social interactions with peers compared to with their matched counterparts, and they interacted more often with a typically developing child than with another special education child. Despite the intergroup differences in frequency of interaction, a similar distribution of interactions emerged for both groups, who presented mostly positive social behaviours, fewer low-level behaviours, and very infrequent negative behaviours. Children with autism reported higher degrees of loneliness than their typical age-mates, as well as lower association between social interaction and loneliness, suggesting their poorer understanding of the relations between loneliness and social interaction. Research and practice implications of these findings are discussed.

Several findings are of particular interest regarding differences found between the children with autism and their typically developing peers. Indeed, as expected, children with typical development revealed a higher level of participation in peer interaction (both initiations and responses) compared with high-functioning children with autism, both in the general categories of social interaction (positive, negative and low-level) and in most of the specific social behaviours in each general category. However, the distribution of social interaction behaviours was identical for the two groups, whereby the majority of behaviours were positive (eye contact, sharing, social communication); second were low-level interactions (mostly looking, functional communication, and close proximity); and last, very few negative behaviours (such as physical or verbal aggressiveness) were noted in either group. This profile corroborates Hauck et al 's (1995) findings regarding low-functioning children with autism, suggesting that, similar to the typical children, this profile may universally characterise autism regardless of functioning level. However, in contrast with the low-functioning children with Hauck's research, the high-functioning children in our study were much more socially active with peers.

In this study, high-functioning children with autism initiated and responded to peers at about half the rate of typical controls, even during unstructured outdoor recess activity, which is considered the most challenging social framework for these children. A surprising finding was the relatively high rate of social initiations revealed by the children with autism in this study. Loneliness is a strong social motivation-actually the strongest drive in typically developing children to initiate or to take part in social relationships and interactions with peers. This study probed whether children with autism would be able to understand the complexity of loneliness, including its more emotional aspect of closeness and affective ties. Surprisingly, these children were as good as their typically developing counterparts in understanding that a close friend might protect them from loneliness and that the presence of many people without a close friend would not protect them from lonely feelings.

In terms of social versus emotional nature of loneliness feelings, contrary to our expectations, children with autism in this study reported themselves to be lonelier compared with typically developing children on both the emotional and social dimension of loneliness. The dual roots of these children's loneliness hold implications for conceptualising the social/emotional deficit in autism.

We would like to conclude the study with an emphasis on several of its implications, both for research and practice. The lower frequency and quality of social interaction, combined with the higher reported loneliness in high-functioning children with autism as compared with typical controls, may indicate that these children would like to take part in more satisfying social interactions - but probably do not have the knowledge of how to do so. Furthermore, the poor associations between a good understanding both of peer interaction and of loneliness with the actual manifestations of social interactions with peers found for the children with autism may suggest that these children lack an intersubjective understanding of social relationships and interactions with peers.