Developing an ecologically valid EEG paradigm to parse language subgroups in autism

Abstract

The present work describes a series of EEG / ERP studies conducted to develop a short passive paradigm that is sensitive to the detection of cortical markers of auditory processing and social motivation. A novel oddball paradigm was developed which utilized naturalistic speech and nonspeech stimuli and contrasting synthetic speech and nonspeech stimuli. Application of this paradigm to samples of typically developing children revealed evidence for faster auditory processing of natural speech stimuli relative to synthetic speech stimuli (Study 1), and evidence for robust orienting responses to stimuli that followed a stream of speech sounds but not a stream of nonspeech sounds (Study 2). Study 3 applied the paradigm to three language subgroups in autism (minimally verbal, phrase speech, verbally fluent) and a typically developing comparison group, and demonstrated that cortical markers of auditory processing (P1, N2) and social motivation (P3a) revealed a distinct profile for each of the three language subgroups of children with autism. These findings demonstrate the importance of including naturalistic stimuli in EEG studies of language processing and also suggest that distinct language-related mechanisms may give rise to the different forms of language impairment that occur in the context of autism.