Virtual Social-Attention, Anxiety and ADHD in Older Children with ASD

Background: Atypical social attention is a cardinal feature of Autism Spectrum Disorders (ASD). However, our understanding of social attention is school aged children with ASD is less precise than in preschool children. The former would benefit from the application of measures of types of social-attention that matures between 8 to 18 year-of-age and from research that examines the possible moderators of attention in ASD, such as social anxiety or ADHD.  

Objectives:

This paper describes a new paradigm for social attention research that assesses social orienting and attention disengagement to social partners in a virtual-reality public speaking task. Two studies of 91 children will be presented that examine the sensitive of this paradigm to diagnostic group differences in 8- to 16-year-olds. The studies also examine the hypothesis that social versus object-oriented attention is more sensitive to diagnostic group differences, and that social anxiety and ADHD symptoms moderate performance group differences in social attention. 

Methods:  

37 higher functioning children with ASD (11.92 years, SD = 1.2) and 54 children with typical development (11.75 years, SD = 1.1) were participants. The group Full Scale IQs were 108 (16.3) and 116 (14.84) for the ASD and TD samples respectively.  Parent report on the Conners-3 ADHD and child self-report on the Multidimensional Anxiety Scale for Children (MASC) were obtained. Children were presented with two 3-minute trials in a 360 degree, 3D virtual classroom populated by 9 peers (avatars) at a classroom table. The children responded to concrete questions about themselves while attempting to direct attention to the faces of each of the avatars on each trial. Five Frequency of Looks, and Average Duration of Looksscores were calculated for the Central Avatar (CA) and pairs of Avatars 1, 2, 3, or 4 positions to the left or Right of CA.

Results:  

The Frequency of Looks measures, but not the attention duration measures, displayed diagnostic group sensitivity and specificity of 76% and 74%, (p < .004).  A Diagnostic Group by Avatar Position quadratic interaction for the Five Frequency of Looks revealed that significant group differences occurred for avatar positions 2 and 3 Left/Right of CA,  F (1,74) = 9.96, p < .002, eta² = .12.  More robust group differences were observed with social versus non-social avatars F (1, 41)  =  8.50, p < .006, eta2 = .17. Finally, Social Anxiety and ADHD interacted and moderated social attention in the HFA sample, (1, 23) = 5.81, p < .025, eta²= .20.

Conclusions:  

The results indicate that virtual public speaking tasks may offer a new paradigm for delving more deeply into the course of ASD related developmental impairment in social attention in school aged children. The results are consistent with previous findings on the maintenance of more robust social versus non-social attention disturbance in school aged children with ASD.  The results also highlight the need to consider complex nature of the moderator processes in fully informative research on the heterogeneous nature of development in ASD.