Objectives: In this study, part of a larger, on-going project to identify autism subphenotypes (The Autism Phenome Project), we sought to identify electrophysiological markers of sensory processing subphenotypes. Our approach was to examine the electrocortical response amplitude recorded to stimuli of increasing loudness. A critical feature of this approach was the development of procedures that yielded robust data from individual very young participants with ASD.
Methods: 60-channel event-related potentials (ERPs) were elicited by randomly presented 50, 60, 70, and 80 dB 50 ms complex tones via headphones from 30 typically developing (TD) toddlers and 30 children diagnosed with ASD. Diagnostic criteria were based on ADOS, ADI-R, DSM-IV and clinical observation. All children (age 2.5 - 4 yrs.) were judged to have clinically normal hearing. ~1000 stimuli with inter-stimulus intervals of 1-2 s were presented as children passively listened to the stimuli and watched a quiet video of their choice. ERPs were derived separately for each intensity. Data analyses included examination of all-waveform overlays, animations of scalp current density topography and derivation of Laplacian waveforms from identified scalp current foci.
Results: For all children, clearly defined auditory ERPs were obtained to at least one intensity level. TD children, compared with children with ASD, generally had more well-defined ERPs to the lower intensity levels and showed a pattern of graded responses with larger cortical activity evoked by louder stimuli. The pattern for children with ASD was much more variable. However, four distinct loudness dependency response profiles were identifiable: 1) a pattern that resembled the typical response of increasing ERP amplitude with increasing stimulus intensity (N=12); 2) a pattern of little variation between intensity levels (Min-diff) (N=6); 3) a pattern of increasing response amplitude that included secondary or “echo” cortical activations (N=7); and 4) a striking pattern of response amplitude reversal with the largest responses seen to 50 dB stimuli, with decreasing response amplitude to sounds of increasing loudness. (N=5). The groups differed by age, with the inverse group younger than the echo and Min-dif groups (38 mo vs. 45 mo). Initial examination of neuropsychological data associated with these subgroups, analyzed with age as a covariate, show no difference in DQ scores (Mullens), nor overall ADOS scores. However, ADOS behavioral subscores for the Inverse group were significantly lower than for Echo and TD-like groups and marginally lower than the Min-diff group. The Min-diff group was significantly more impaired on the ADI-R behavioral subscale than each other group.
Conclusions: These data suggest that there are distinct electrophysiological sensory response profiles for subgroups of children with ASD that may account for the observed phenotype of atypical reactions to sounds in some children, and which bear relation with other phenotypic measures.