Assessing Lateral Connectivity in Autism Using a Crowding Paradigm

Background: Crowding, often defined as the deleterious influence of nearby contours on visual discrimination, is a form of inhibitory interaction and is ubiquitous in spatial vision (i.e., Levi et al., 2008).  Generally, the ability to discriminate a target stimulus is more difficult when distractors (or flanking stimuli) are simultaneously present within a proximal spatial zone of interaction. For example, the ability to discriminate the orientation of a Landolt C optotype is maximally degraded when the bars are simultaneously presented within a certain distance of the C, referred to as the critical distance. The most congruent physiological explanation for this effect implicates lateral connections (both excitatory and inhibitory) between orientation-specific neurons in low-level visual areas (i.e., V1).

Objectives: To assess lateral connectivity mediating low-level spatial information processing in autism by measuring the far visual acuity of a Landolt-C optotype alone, and with flanking stimuli presented at different distances from the target.

Methods: Ten autistic and nonautistic participants matched for chronological age (mean of 22 years), full-scale IQ, gender and handedness were asked to identify the gap-opening orientation of high-contrast Landolt-C stimuli. The gaps in the C- optotypes were presented at either left, right, top or bottom positions. Five different conditions were assessed; C-optotypes were presented without flanking bars (far visual acuity condition), or with four flanking bars (crowding conditions) placed either abutting the optotype, or at a distance of one, two, and three gap-opening widths from the C-optotype. Flanking bar sizes and distances were always proportional to the size of the C-optotype (one gap width across and 5 gaps width in height), positioned to the left, right, above and below the target (Flom et al.,1963). Far visual acuity, defined as the minimal C optotype size needed to correctly identify gap-opening orientation, was tested for each condition using an adaptive staircase procedure.

Results: Contrary to recent findings (Ashwin et al., 2008), no significant between-group differences in visual acuity were found for the no flanker condition. As expected, crowding decreased visual acuity for both autistic and nonautistic groups. However, whereas the critical distance (flanker distance most affecting acuity) for the nonautistic participants peaked at the one gap-opening width, autistic far visual acuity was affected at all gap widths, albeit to a lesser extent. The largest between-group difference was found at 3 gap widths where no flanker effect on acuity was evident for the nonautistic group, whereas crowding decreased visual acuity for the autism group.

Conclusions: The present study demonstrates that although autistic participants present visual acuity comparable to nonautistic individuals, autistic acuity is differentially affected by crowding when assessed near the acuity limit. We suggest that these results are consistent with atypical lateral connectivity within early visual areas in autism, particularly within the context of spatial information processing. Results will be discussed within the context of current theories suggesting a low-level origin for atypical visual information processing in autism.