Perceptual and Acoustic Measures of Prosody in Individuals with Autism Spectrum Disorder and First-Degree Relatives

Background: Impairments in prosody (e.g., intonation, stress) can be among the most notable characteristics of individuals with autism spectrum disorder (ASD), and can significantly impact communicative interactions (Mesibov, 1992; VanBourgonien & Woods, 1992). Subtle differences in prosody have also been reported in parents of individuals with ASD, suggesting that familial characterization of prosody may help to highlight genetic influence on components of language impacted in ASD. While prior studies of prosody in ASD have used the Profiling Elements of Prosody in Speech-Communication (PEPS-C) to gather perceptual ratings of prosody and some have incorporated acoustic measures alongside these ratings, comprehensive perceptual and acoustic investigation of how specific prosodic abilities (e.g., contrastive stress, lexical stress) differ in individuals with ASD and their parents are needed to clarify key domains contributing to perception of atypical prosody in ASD.

Objectives: To assess perceptual and acoustic measurements of prosody in individuals with ASD and their parents.

Methods: Individuals with ASD (n=26), their parents (n= 43), and respective control groups (n=16 ASD controls; n=30 parent controls) completed the PEPS-C, which was rated by blinded coders. Responses to the expressive subtests of the PEPS-C from a subset of participants (ASD n=5; proband controls n=6; ASD parents n=14; parent controls n=10) were further characterized with acoustic analysis using Praat (Boersma, 2001) to obtain measures of mean fundamental frequency (f₀), standard deviation (SD) of f₀ and f₀ range, both measures of f₀ variability (acoustic analysis of data from the full sample is ongoing). Additionally, correlational analyses were conducted to examine relationships between these measures.

Results: Individuals with ASD and their parents exhibited poorer prosodic imitation, lexical stress expression, and contrastive stress expression (ps<.05). Individuals with ASD also demonstrated significantly greater atypicalities in phrase stress expression (e.g., “I saw a blue bird” vs. “I saw a bluebird”) compared to controls, and this difference was marginal between the parent groups, with the ASD parent group demonstrating greater atypicalities (p=.05). Of the seven receptive subtests, individuals with ASD performed more poorly than controls only in their understanding of contrastive stress (p<.01). Acoustic analyses did not reveal significant differences in the ASD (ps>.67) or parent groups (ps>.61). Decreased f₀ range was associated with ratings of poorer contrastive stress ability in ASD parents (r= .60, p=.04).

Conclusions: Individuals with ASD and their parents demonstrated greater prosodic atypicalities across several domains of the PEPS-C based on perceptual ratings. However, consistent with prior work, these differences were not readily apparent based on acoustic measures of prosody (Diehl et al., 2013), although associations in the ASD parent group suggest that perceptual differences may be related to decreased f₀ variability. The large overlap between subtests in which individuals with ASD and their parents differed from their respective control groups highlights the domains of imitation, lexical stress, and contrastive stress as potentially key areas contributing to overall perception of prosodic atypicalities during conversational interactions. Ongoing analyses including a larger sample size will further investigate acoustic differences and relationships with perceptual ratings of prosody in all groups.