Neurocognitive Factors Associated with Neural Specialization for Letters in ASD

Background:  Problems with neural specialization are hypothesized to negatively impact social function in ASD. However, recent evidence suggests that children with ASD display normative patterns of specialization for non-social information, such as letters of the alphabet, despite social impairments. Event related potentials (ERPs) demonstrate similar time course and morphology for specialized face and letter perception at occipitotemporal sites (N170). Neural and behavioral facets of social perception are well studied in ASD, but few studies have focused on factors related to preserved processing of letter and word-related information in ASD.  Because some children with ASD display advanced skills and a high level of interest and proficiency in reading, understanding the brain bases of letter perception and their behavioral correlates holds potential to clarify areas of strength in ASD and potential strategies to target compensatory function.

Objectives:  This study examined neural and behavioral correlates of letter processing in ASD. Our goals were to examine variability in neural responses to letters as a function of (a) performance on standardized assessment of phonemic decoding, (b) verbal cognitive abilities, and (c) use of verbal and written language in daily life. 

Methods:  Participants were thirty-six children with high-functioning ASD (Mean age = 11.2 years) and eighteen TD children (Mean age = 12.6 years) matched for sex, age, and IQ. ERPs were recorded with a 256 electrode Geodesic sensor net while participants completed a letter recognition task. ERP analysis focused on N170 amplitudes and latencies evoked by letter and pseudoletter stimuli.  Reading ability was assessed with the Letter-Word Identification and Word Attack subtests of the Woodcock-Johnson Tests of Achievement-III; verbal IQ was measured by the Wechsler Abbreviated Scale of Intelligence.  The Written Communication Subscale of the Vineland Adaptive Behavior Scales assessed writing skills of each participant.  Circumscribed interest in reading was measured with the Yale Special Interest Survey. 

Results:  A main effect of condition indicated that both groups demonstrated shorter P1 latency (F(1, 52) = 16.28, p ≤ .01) and increased amplitude of the N170 (F(1, 52) = 14.67, p ≤ .01) for letters relative to pseudoletters. Single word reading ability was correlated with longer right hemisphere N170 latency in the letter condition (r = .64, p ≤ .01). Preliminary analyses indicate positive correlations between ERP parameters and adaptive measures of written communication and verbal cognitive ability.  

Conclusions:  Individuals with ASD displayed normative patterns of specialization for letters of the alphabet, in terms of both strength and latency of neural response. Likewise, the performance of children with ASD on standardized tests of decoding was comparable to the control group and in the average range. This study demonstrates that variability in neural specialization of letters is associated with both laboratory and real-life functioning in ASD. Results suggest that investigations of areas in preserved function in ASD can inform development of interventions to (a) apply extant strengths to boost areas of vulnerability and (b) harness intact brain systems for compensatory function in ASD.