Neural Mechanisms Underlying Visuospatial Functioning in Different Cognitive Subgroups in Autism Spectrum Disorder

Background: Strengths in visuospatial functioning have been found in a large proportion of individuals with an autism spectrum disorder diagnosis. Neuroimaging studies have revealed higher activation in occipito-parietal regions and diminished activation in some frontal regions in autism (Kana et al., 2013; McGrath et al., 2012; Silk et al., 2006). However, no such studies have investigated the neural mechanisms underlying visuospatial functioning in distinct subgrouping of autistic individuals based on their visuospatial abilities.

Objectives: The principal goal of this study was to identify the neural network involved in visuospatial abilities and expertise in autism in different subgroupings based on their Wechsler Block Design subtest’s performance using functional magnetic resonance imaging (fMRI) technique while performing two visuospatial tasks.

Methods: 28 male autistic participants were matched to 22 male non-autistic participants on age (19-36) and WAIS-IV non-verbal IQ (70-134). Among autistic participants, 13 had a relative strength on Wechsler Block Design (BD) subtest – a performance on BD subtest more than 1 standard deviation above their mean performance at other Wechsler subtests – and 15 had no particular visuospatial expertise. All participants performed two visuospatial tasks presented in an event-related design in a 3T MRI scanner. The first task was a classic mental rotation (MR) task with three-dimensional geometric shapes. Spatial orientation (0, 70, 140 and 180 degrees) was parametrically varied across 104 trials. The second task was an adaptation of the original BD subtest suitable for presentation in the MRI scanner. Perceptual cohesiveness of the target design (neutral, low, high) was varied across the 90 trials.

Results: Despite no between-group differences in accuracy and response times, non-autistic participants recruited a larger network of cortical regions (occipital, parietal, temporal and frontal) during both tasks, compared to autistic participants. No region was more activated in autistic participants with superior visuospatial abilities across both tasks, and in autistic participants without superior visuospatial abilities in the adapted BD task, relative to non-autistic participants. In the MR task, greater activity in some fronto-parietal regions, the insula and the cerebellum was found in autistic participants without superior visuospatial abilities compared to non-autistic participants. Further analyses compared cerebral activity in autistic participants with and without superior visuospatial abilities. Higher activation in occipital, parietal and frontal regions was observed in the autistic group without superior visuospatial abilities while performing both tasks, whereas the autistic group with superior visuospatial abilities showed higher activation in parietal regions in the adapted BD task and in the insula during the MR task.

Conclusions: Despite equivalent behavioral performance, the underlying pattern of cerebral activity is less extended in autistic individuals than in non-autistic individuals during visuospatial tasks. Interestingly, functional resource allocation is similar in autistic without superior visuospatial abilities and in non-autistic individuals when both compared to autistic with superior visuospatial abilities. Together, these results suggest more efficient visuospatial processing requiring less cognitive resources in autism, which is particularly true for those with superior visuospatial abilities.