Prosodic Abilities in High Functioning Autism

Background: Language is crucial to human communication, and it is usually studied in terms of the words and sentences used for conveying ideas, beliefs, and instructions to others; however, other aspects of language are equally important. These other features impact not only on what we say but also on how we say it. These aspects of language are usually referred to as prosody. They include stress and prominence patterns, rhythm, phrasing and intonation, and have several communicative functions. Difficulties with prosodic skills can dramatically influence daily conversations, social interactions, and even typical language development.

Objectives: We aim to analyse prosodic impairments in a group of children with high functioning autism (HFA) compared to typically developing peers (TD), and study possible interactions with other deficits in autism.

Methods: 15 children (3 girls, 12 boys) with HFA (6 - 9 years; M = 7.40, SD = 1.12; IQ ≥ 70) matched to a TD group on age and gender (n= 15). We examined prosody (using the Profiling Elements of Prosody in Speech-Communication; PEPS-C), non-verbal intelligence, attention, executive functions, language, and receptive vocabulary.  

Results: An analysis of variance (ANOVA) on the tests results obtained, with Group as a factor, showed significant differences in the following PEPS-C subtests: Turn-End Reception (F (1, 28) = 5.701, p = 0.024; η² = .169) and Focus Expression (F (1, 28) = 4.490; p = .043; η² = .138). After controlling for the effect of non-verbal intelligence, significant differences between groups were also found in the tasks Short-Item Discrimination (F (1, 28) = 34.542, p = 0.001; η² = .571), Long-Item Discrimination (F (1, 28) = 23.028, p < .001; η² = .470), Long-item Imitation (F (1, 28) = 10.906, p = 0.003; η² = .295), Turn-End Reception (F (1, 28) = 61.074, p < .001; η² = .701), Affect Reception (F (1, 28) = 56.589, p < .001; η² = .685), and Focus Reception (F (1, 28) = 6.850, p= 0.015; η² = .209). No effect of the variables executive function, attention, language, and receptive vocabulary was found. To determine the contribution of non-verbal intelligence to the impaired performance in prosodic tasks, this variable was entered in a regression model as a predictor of the overall PEPS-C results. This regression model explained about 82% of the variance. 

Conclusions: The effect size of non-verbal intelligence, as well as the regression model results, showed that this skill had an important explanatory role for the differences observed in prosodic abilities of individuals with autism. Consequently, it could be concluded that prosodic abilities are affected by cognitive strengths or difficulties. This result has led us to consider that different approaches of autistic cognition (driven by neural differences) can be applied to linguistic traits observed in autism. Cognitive theories can guide future research, and can help to clarify the different causes of inconsistent results in studies of impaired prosody in autism.