Gamma-Band Deficits During Language Processing in Adults with Autism and First-Degree Relatives of Children with Autism: An MEG Study

Background: Language ability, specifically a deficit in phonological processing ability, has been proposed to be one of six core broad autism phenotypes. Deficits involving phonology have been observed in a significant subset (ranging from 24 to 77%) of children with autism and have been reported in studies of unaffected first-degree relatives.  Despite this, no neuroimaging studies have investigated phonological processing in individuals with autism or their first-degree relatives.

Objectives: Magnetoencephalography (MEG) was utilized to investigate the neurobiology of phonological processing in adults with autism as well as parents of individuals with autism.  The aim of the present study was to examine the presence of phonological processing deficits in individuals with autism as well as unaffected first-degree relatives in order to provide evidence in favor of or against the inclusion of phonological processing deficits as a core broad autism phenotype (BAP) trait.

Methods: Sixteen parents of a child with autism, thirteen adults with autism and seventeen controls performed a phonological priming task while undergoing whole-cortex MEG.  The task consisted of four prime-target word conditions including homophones (e.g., PAUSE-paws) and related stimuli.  Primes were presented below perceptual threshold (i.e., 30ms).  Subjects, who were not informed that stimuli consisted of word pairs, performed a lexical decision task (i.e., is it a word or a nonword?) on all lowercase targets.

Results: In our priming condition that placed heavier demands on phonological decoding skills, adults with autism exhibited reduced evoked gamma band activity relative to both parents and controls in the left supramarginal gyrus (SMG) centered around 200 ms post-stimulus onset.  Adults with autism exhibited an additional later reduction in evoked gamma within the left superior temporal gyrus (STG) relative to controls, and an additional earlier reduction in induced gamma band activity relative to first-degree relatives in the left STG and SMG.  Reductions in evoked gamma activity were also observed in adults with autism for unprimed relative to primed stimuli in the left SMG relative to the control group, but in the right STG relative to the parent group.  No differences in gamma activity were observed between the parent and control groups.

Conclusions: These results are consistent with phonological processing deficits in individuals with autism. While not providing clear evidence for phonological processing deficits as a core BAP trait, the present study does provide evidence for alternate neural strategies during phonological processing in unaffected first-degree relatives. Furthermore, these alternate strategies were found to be related to gamma-band activity, abnormalities in which have been suggested to be a candidate endophenotype in autism.  These results suggest the potential of future imaging investigations of phonological processing in unaffected first-degree relatives of children with autism.  In addition, the present results do provide both behavioral and functional imaging evidence for phonological impairments in individuals with autism that could have potential translational significance for language interventions in autism.