The interplay between voice acoustics, expressive language, and motor development in autistic children
OA Version
Citation
Abstract
School-aged autistic children exhibit heterogeneous expressive language abilities. The reasons why around 30% of them remain nonverbal or minimally verbal (MV) are still understudied. Investigating the mechanisms underlying the expressive language heterogeneity is important to inform personalized interventions and to promote optimal language outcomes. Vocal and fine motor development occur in parallel and are highly intertwined over time. However, this mouth-hand link has rarely been studied in preschool- and school-aged autistic children who develop heterogeneous expressive language abilities. Recent studies have demonstrated the contributions of speech production and fine motor skills to explaining concurrent variability in expressive language abilities in school-aged autistic children. Acoustic features such as fundamental frequency (f0) variability and cepstral peak prominence (CPP) are associated with children’s phonatory stability, offering valid measures to quantify differences in vocal motor development between autistic and neurotypical (NT) children. My dissertation consists of three studies aimed at providing a more in-depth evaluation of the relationships among voice acoustics, expressive language, and motor development in school-aged autistic children.In Study 1, acoustic differences in f0 variability and CPP were assessed between 49 autistic (14 female) and 25 NT children (10 female) aged from 3 to 7 years. The relationships between acoustic features and fine motor skills were examined in autistic children. Autistic children completed a syllable imitation task remotely to assess their acoustic features. NT children completed the same task in a lab setting. Fine motor skills were assessed using the v-scale score of the fine motor subdomain of the Vineland Adaptive Behavior Scales, Third Version (VABS-3, Sparrow et al., 2016). Autistic children were found to show lower f0 variability and greater CPP than NT children. However, these group differences were not statistically significant after controlling for age. Greater fine motor skills were negatively associated with lower CPP after controlling for age, suggesting autistic children with greater fine motor skills tended to exhibit lower phonatory stability. There was no relationship between f0 variability and fine motor skills. Our findings suggested that autistic children developed phonatory stability comparable to that of their NT peers. The preliminary evidence of the negative relationship between CPP and fine motor skills highlighted the significance of replicating these findings using high-quality acoustic measures and more rigorous statistical models to establish the potential link between vocal and fine motor development.
Building on the findings of Study 1, Study 2 further examined differences in f0 variability and CPP across MV autistic, verbal autistic, and NT children aged from 3 to 8 years using the same syllable imitation task. The unique influence of fine motor skills on acoustic features was evaluated in autistic children by controlling for age, autism severity, and nonverbal cognition. Autistic children were classified into MV (ASD-MV; N = 20, 4 female) and verbal (ASD-V; N = 34, 9 female) groups based their Autism Diagnostic Observation Schedule, Second Edition (ADOS-2, Lord et al., 2012) module assignments (ASD-MV: module 1; ASD-V: module 2 or 3). Autism severity was measured using the total calibrated severity score (CSS) from the ADOS-2, and nonverbal cognition was assessed using ratio IQ scores derived from either the Mullen Scales of Early Learning (MSEL; Mullen, 1995) or the Differential Ability Scales, Second Edition (DAS-II; Elliott, 2007). Fine motor skills were assessed using the v-scale score of the fine motor subdomain of the VABS-3 (Sparrow et al., 2016). Using the same NT children from Study 1 as the control group, we observed that the ASD-MV group exhibited significantly lower CPP than the NT group. No group difference in CPP was found between the ASD-MV and ASD-V groups, or between the ASD-V and NT groups. There was no significant difference in f0 variability across the three groups. Additionally, a suppressor effect was observed when evaluating the relationship between CPP and fine motor skills. Greater fine motor skills were negatively associated with lower CPP only after controlling for age, autism severity, and nonverbal cognition. Fine motor skills were not associated with CPP when these covariates were not included in the model. Taking nonverbal cognition as a proxy of children’s general developmental level, we further explored whether fine motor skills mediated the relationship between nonverbal cognition and CPP in autistic children. We found a significant indirect effect of nonverbal cognition on CPP through fine motor skills. f0 variability was not associated with fine motor skills. Together, MV autistic children demonstrated distinctive vocal motor developmental profiles in their phonation. They exhibited reduced voice quality compared to their NT peers. We observed a unique contribution of fine motor skills to explaining variability in phonatory stability in autistic children. The results supported our hypothesis that increased phonatory stability in autistic children reflected inflexibility in adjusting their vocal behaviors during voice production, as evidenced by fine motor skills altering the positive relationship between nonverbal cognition and CPP.
In Study 3, we investigated the relationship between expressive vocabulary and concurrent consonant inventory, and the relationship between expressive vocabulary and three motor skills (gross motor, fine motor, and oral motor) in autistic children (N = 52, 13 female) aged from 5 to 7 years. Expressive vocabulary and consonant inventory were assessed using a 15-minute natural language sample. The three motor skills were measured using parental interviews and questionnaires. Greater consonant inventory and fine motor skills were significantly associated with greater expressive vocabulary size, whereas gross motor and oral motor skills did not. These findings extended prior research on early language and motor development to a school-aged autistic population, highlighting the potential contributions of consonant inventory and fine motor skills to explaining concurrent variability in expressive vocabulary.
Findings from this dissertation underscore the importance of 1) taking a hypothesis-driven approach to characterize multiple acoustic features both between autistic and NT children and within autistic children who develop heterogeneous expressive language abilities to enhance our understanding of differences in vocal motor development across MV autistic, verbal autistic, and NT children; 2) examining the relationship between vocal and fine motor development in autistic children beyond preschool age to further our knowledge of the motor bases underlying acoustic atypicalities; 3) evaluating individual motor and language skill areas to identify strengths and weaknesses of school-aged autistic children and to inform individualized interventions to promote optimal expressive language outcomes.
Description
2025