Visuospatial Cognition and Dimensions of Autistic Traits: Reduced Lateralization of Attention and Superior Search Relate Specifically to Social Difficulties

Background: Research has identified two visuospatial cognitive characteristics of individuals with high levels of autistic-like traits (ALT) – facility in embedded figures search and reduced left-visual-field (LVF) attentional bias – relative to individuals with low ALT levels. However, ALT are multi-dimensional, with social and non-social factors essentially independent for instruments such as the Autism-Spectrum Quotient (AQ). Therefore to advance understanding of atypical cognition on the autism spectrum and its likely neural correlates, we isolated contributions of two ALT dimensions.

Objectives: Our aims were to (1) check the independence of the Social Difficulties (SD) and Patterns/Interests/Details (PID) AQ factors; (2) investigate whether facility in embedded figures search and reduced LVF attentional bias can be attributed to elevated SD traits, elevated PID traits, or their combination; and (3) examine the relationship between search performance and LVF bias as potential support for a link to differences in right hemisphere function.

Methods: First, 702 volunteer undergraduate students were screened for SD and PID AQ scores. Next, volunteers were invited for further testing if their scores fell in the upper or lower 30% for each of the SD and PID distributions. This provided a 2 x 2 design (High SD-High PID [n=25]; High SD-Low PID [n=25]; Low SD-High PID [n=26]; Low SD-Low PID [n=25]), with the four groups matched on age and gender. These 101 participants (M_age = 20.61 years) then completed the Lueven Embedded Figures Test (L-EFT) and a greyscales task to assess LVF bias.

Results: Across the 702 screened students, SD and PID scores were independent, correlating r = 0.02, p = 0.58. On the greyscales task, High SD students showed a reduced LVF advantage (M=54.53%; SE=2.98% left responses) compared to Low SD students (M=61.94%; SE=2.16% left responses; F(1,97)=4.00, p<.05) whereas High and Low PID students did not differ in LVF bias (p=.53). Similarly, on L-EFT accuracy, High SD students (M=76.69%; SE=.96%) outperformed Low SD students (M=73.16%; SE=1.41%; F(1,97)=4.24, p<.05) whereas High and Low PID scorers were close in accuracy (p=.58). Finally, there was a trend for superior L-EFT performance to correlate with reduced LVF bias (r = -0.17, p= .097).

Conclusions: Facility in embedded figures search and reduced left-visual-field bias are characteristics of individuals who report pronounced social difficulties rather than of individuals who report pronounced attention to patterns, special interests or sensory details. The independence of the social and non-social dimensions of autistic traits, and their contrasting patterns of association with visual search and attentional bias, caution against investigating autistic traits as a single spectrum. The superior visual search and reduced attention to the left side of space associated with social difficulties support speculation that each effect reflects reduced involvement of the right hemisphere relative to the left in visuospatial cognition.