Mechanisms of Voice Control in Autism Spectrum Disorder: A Family Study

Background: Impairments in prosody (e.g., intonation, rate), have been noted since the earliest descriptions of social communication impairments in autism spectrum disorder (ASD; Kanner, 1948). Additionally, subtle differences in prosody have been identified in a subset of parents of individuals with ASD with the broad autism phenotype (BAP; Losh et al., 2012). Existing investigations of prosody in ASD have focused primarily on perceptual judgments and global acoustic measures (e.g., fundamental frequency), which are beneficial in characterizing prosodic profiles within this population; however, these studies have failed to investigate the feedforward and feedback control mechanisms, which support planning, production, self-monitoring, and self-correction of speech (Guenther, Tourville, & Bohland, 2015), that may underlie prosodic atypicalities in ASD.

Objectives: This study investigated potential breakdowns in sensorimotor monitoring and correction of prosody in individuals with ASD and their parents.

Methods: A pitch-altered voice auditory feedback paradigm in which participants vocalized and received real-time auditory feedback of their voice was implemented (Burnett et al., 1998). Participants included individuals with ASD (n=20), their parents (n=20), and respective controls (n=20 proband controls; n=23 parent controls). In the first task (“vowel”), participants vocalized “ah” for five seconds, while five randomly-timed pitch alterations occurred. Simultaneously, event-related potentials (ERPs) to pitch alterations were recorded using a 32-electrode electroencephalography (EEG) system. In the second task (“speech”), participants asked the question “You know Nina?” and one pitch alteration occurred on the third syllable. Vocal responses from both tasks were examined in relation to prosodic abilities assessed by the Profiling Elements of Speech Prosody-Communication assessment (Peppé & McCann, 2003).

Results: The ASD group had larger responses to pitch-altered voice auditory feedback in the vowel task compared to controls (p=.03). Increased vocal response magnitudes during vowel production in both ASD and their controls, were associated with poorer expressive prosodic abilities (ps < .05). The ASD parent group did not differ from parent controls in vocal response magnitude during the vowel task, but preliminary ERP results demonstrated smaller N1 amplitudes in fronto-central electrodes in ASD parents (p =.09). Both ASD and ASD parent groups demonstrated reduced vocal responses in the speech task compared to respective control groups (ps <.02). Differences in the parent groups were driven by parents with the BAP (p < .02).

Conclusions: In the ASD group, larger vocal responses to pitch-altered voice auditory feedback during vowel production but attenuated responses during speech, suggest difficulty incorporating auditory feedback cues into vocal control mechanisms. ASD parents, specifically those with the BAP, also demonstrated attenuated responses during speech, suggesting that response magnitudes to pitch-altered voice auditory feedback during speech may index genetic liability to ASD. ERP results from ASD parents during vowel production may indicate attenuated abilities in detecting mismatches between intended and actual sensory consequences. Correlational analyses provide additional support for the relationship between vocal responses to pitch-altered auditory feedback and prosodic production. Ongoing analyses investigate differences in response magnitude between vowel and speech tasks to determine if individuals with ASD and their parents demonstrate task-dependent responses similar to controls (Chen et al., 2006).