Age-Related Differences in Local and Global Processes for Social and Non-Social Information in Autism: How Do Children and Adolescents Differ?

Background:  Individuals with autism present both an atypical and distinctive visuo-perceptual profile (Mottron et al., 2006; Bertone et al., 2010).  Despite marked social and behavioural impairments, individuals with autism often excel at tasks requiring a local analysis of detailed information and preferentially attend to the constituent parts of a stimulus rather than its whole form.  It remains unknown, however, whether a bias for such local analysis is at the origin of other aspects of the social cognitive phenotype in autism, such as facial information processing. Age-related changes in local and global processes are also poorly understood.

Objectives:  The goal of this study was to assess age-related differences in local and global visual processing strategies used in social (face identity discrimination task) and non-social (Block Design task) visuo-perceptual tasks in children and adolescents with and without ASD.

Methods:  Twenty-eight participants with autism and twenty typically-developing participants were split into child and adolescent age groups (5-11; 12-17 years).  All participants performed social and non-social visual perceptual tasks under conditions favouring either a local or global analysis. Both tasks incorporated a simultaneous presentation of a central target and four surrounding choice probes on a touchscreen monitor.  In the social task, participants completed a face identity discrimination task wherein they matched identities across varied viewpoints (i.e. front, inverted and view-change).  The presentation of faces in the same-view (front or inverted) biased a local analysis, whereas in different views (view-change) biased a global analysis. In the non-social task, participants completed a block design task incorporating puzzle-like designs wherein they identified the correct match to a target design. The puzzles were manipulated in terms of perceptual coherence (PC), with high PC patterns favouring a global analysis and low PC patterns a local analysis. All responses were made with a simple press on the display. Measures of accuracy and reaction time were recorded.  

Results:   In the social task, children and adolescents with autism were significantly less accurate than typically-developing participants only when matching faces across different views. Interestingly, adolescents with autism responded significantly slower than children with autism in the view-change condition, when access to local information was minimized and a more global analysis was required. In the non-social task, all participants achieved higher accuracy on the high than low PC patterns, and adolescents made significantly more correct responses than children. All participants responded significantly faster to the high than to the low PC patterns.

Conclusions:  Our findings illustrate subtle, yet significant differences in social information processing in children and adolescents with autism, specifically when a global analysis is required. However, we found no differences between groups in the local and global processing of non-social information. Strengths in local and/or weaknesses in global processing in ASD may be more apparent in adolescence than in childhood, consistent with other age-related findings (Guy et. al, in press).