Delayed but Not Deviant Developmental Trajectories Related to Language Impairment in Children with Autism Spectrum Disorder: Neural and Behavioral Evidence

Background:  It has been estimated that half of all children with ASD have co-occurring language impairment (ASD+LI, Blumberg, Bramlett, Kogan, Schieve & Jones, 2007). Eigsti and Bennetto (2009) found that the language abilities of these children lag behind the typical developmental trajectory (i.e., showing delay in development rather than deviance). Neural studies have proposed that atypical brain responses to sound underlie the language impairments in children with ASD (e.g., Oram Cardy, Flagg, Roberts & Roberts, 2008). However, brain function studies to date have not determined whether atypical neural response patterns are reflective of a developmental delay or deviance trajectory.

Objectives:  To explore whether patterns of developmental delay or deviance in brain responses to sounds are related to language delays in children with ASD+LI.

Methods:  Twenty-two children with ASD (N=13 without LI: ASD-LI; N=9 with LI: ASD+LI) participated. To explore the neural responses to sound, a 128-channel EGI system recorded auditory evoked potentials (AEPs) elicited by 225 trials of a 50ms, 490Hz tone. The Clinical Evaluation of Language Fundamentals – 4 (CELF-4)was administered to measure language ability. The following statistical analyses were performed to identify patterns of delay or deviance:

• For language abilities: Scatter-value (S-value, VanMeter, Fein, Morris, Waterhouse & Allen, 1997) was calculated based on the individual test items that each participant answered incorrectly using the equation S = sqrt(P x ∑W), where peak (P) = the last correctly answered item and weight (W) = percentage of children with typical development (TD) who answered correctly (calculated from our previously established normative database). Children with a developmentally deviant language ability will show more intra-test scatter than their peers and earn higher S-values.
• For neural responses: For each participant, an AEP segment from 0-500ms post-stimulus presentation was compared to normative AEPs we previously established from 74 children with TD. Using intraclass correlation coefficient (ICC) as an indicator of resemblance, the AEPs of children with ASD were assigned an age-equivalent based on the comparison that yielded the highest resemblance score. Low resemblance of a participant’s waveform compared to all normative waveforms result in a low ICC value, indicating a pattern of deviance in neural responses.

Results:  As expected, children with ASD+LI scored lower on the CELF-4 compared to children with ASD-LI. However, the ASD+LI group did not have higher S-values on the CELF-4 than the ASD-LI or language-matched TD groups (see Table).In other words, there was no evidence to suggest a pattern of developmental deviance in the language performance of children with ASD+LI. Despite being equivalent in chronological age, children with ASD+LI had significantly younger AEPs than children with ASD-LI. The ASD+LI group did not have lower ICC-values when compared to ASD+LI or language-matched TD groups. This finding suggests that brain responses to sounds in the ASD+LI group are delayed relative to, rather than deviant from, a typical developmental trajectory.

Conclusions:  Our data suggest that developmental delay in the neural responses to sound underlies the language delay in children with ASD+LI. Both the neural and behavioral data point to trajectories characteristic of younger children with typical development.