Background: Autistic individuals exhibit atypical patterns of sensory processing that are known to be related to quality of life, but which are also highly heterogeneous. Previous investigations of this heterogeneity have ordinarily used questionnaires and have rarely investigated sensory processing in Typical Development (TD) alongside Autism Spectrum Development (ASD).
Methods: The present study used hierarchical clustering in a large sample to identify subgroups of young autistic and typically-developing children based the normalized global field power (GFP) of their event-related potentials (ERPs) to auditory stimuli of four different loudness intensities (50, 60, 70, 80 dB SPL): that is, based on an index of the relative strengths of their neural responses across these loudness conditions.
Results: Four clusters of participants were defined. Normalized GFP responses to sounds of different intensities differed strongly across clusters. There was considerable overlap in cluster assignments of autistic and typically-developing participants, but autistic participants were more likely to display a pattern of relatively linear increases in response strength accompanied by a disproportionately strong response to 70 dB stimuli. Autistic participants displaying this pattern trended towards obtaining higher scores on assessments of cognitive abilities. There was also a trend for typically-developing participants to disproportionately fall into a cluster characterized by disproportionately/nonlinearly strong 60 dB responses. Greater auditory distractibility was reported among autistic participants in a cluster characterized by disproportionately strong responses to the loudest (80 dB) sounds, and furthermore, relatively strong responses to loud sounds were correlated with both auditory distractibility and noise distress. This appears to provide evidence of coinciding behavioural and neural sensory atypicalities.
Limitations : Replication may be needed to verify exploratory results. This analysis may ignore some variability related to classical ERP latencies and topographies. The sensory questionnaire employed was not specifically designed for use in autism. Variability in sensory responses unrelated to loudness is ignored, leaving much room for additional research.
Conclusions: Taken together, these data demonstrate the broader benefits of using electrophysiology to explore individual differences. They illuminate different neural response patterns and suggest relationships between sensory neural responses and sensory behaviours, cognitive abilities, and autism diagnostic status.
Autism, heterogeneity, hierarchical clustering, event-related potentials (ERPs), subgroups, sensory processing
This is a list of supplementary files associated with the primary manuscript. Click to download.