The Characteristics of F0 Contours in Speech Directed at Infants at-Risk for Autism

Background: The mother-infant interaction mediates early development and sensory regulation, and these interactions largely comprise early sensory experiences for the infant. One way mothers interact with their baby is through infant-directed speech (IDS), which is a unique speech register used by adults when speaking to infants. In particular, IDS can facilitate socio-communicative and language development in infants. One salient aspect of IDS is the trajectory of pitch, known as F0 contours. Research shows F0 contours are associated with infant affect and attention, but less is known about how F0 contours vary across contexts. It is important to understand the role of IDS with infants who are at-risk for socio-communicative, sensory regulation or language delay. Infants later diagnosed with Autism Spectrum Disorder (ASD) display significant impairments in these areas, however there is currently no research investigating how maternal F0 contours are used with these infants. It was hypothesised that infants at-risk for ASD would be spoken to differently when compared to infants not at-risk.

Objectives: The aim of this study was to compare F0 contours used by mothers when interacting with infants at-risk for ASD, to those used by mothers interacting with infants not at-risk.

Methods: Six 12-month-old infants and their mothers participated in a 15-minute recorded interaction. Of the six infants in the study, three were clinically at-risk for ASD (high-risk), and three infants were not at-risk (low-risk). Infant risk for ASD was assessed using the parent-report First Year Inventory, which indicates two areas of risk; socio-communication and sensory regulation (Reznick, Baranek, Reavis, & Crais, 2007). 1039 maternal F0 contours were extracted from the audio recorded during the interaction, and classified into contour types (rising, bell, sinusoidal, u-shape, slowly-falling, rapidly-falling, and complex). Independent samples t-tests were performed to investigate differences in mean proportions of F0 contours used with high-risk and low-risk infants. Pearson product-moment correlations were conducted to determine any associations between ASD risk and the proportion of contours used.

Results: Mothers of high-risk infants spoke a total of 419 utterances during the interaction, compared to 620 spoken by mothers of low-risk infants. Preliminary analyses of contours used between mothers of high-risk infants compared to low-risk infants revealed a difference in frequency of slowly-falling contours, which was approaching significance (t=-2.714, df=4, p=.053). More complex contours (r=.824, N=6, p=.04) and less rising contours (r=-.811, N=6, p=.05) were associated with a high-risk for ASD. Higher sensory regulation risk was associated with less rising (r=-.853, N=6, p=.031), more complex (r=.858, N=6, p=.029), and more rapidly-falling contours (r=.829, N=6, p=.041).

Conclusions: Although the sample size is small, these preliminary results suggest that infant risk for ASD is related to the way mothers use F0 contours during interactions. This could provide evidence that infants at-risk for ASD, even as early as 12 months-old, are spoken to differently than infants not at-risk. Further research into how mothers use IDS with high-risk infants is recommended and will be presented at the conference, in order to inform early interventions that could improve outcomes for children diagnosed with ASD.