Links Between Looking to Audiovisual Speech and Language in Infants with Autistic and Non-autistic Siblings

• In Event: Language and Communication Across Development

Abstract

Theory and prior research suggest that visual speech cues (i.e., the eyes and mouth of the speaker) facilitate speech processing and language development. Infants attend to these cues early in life, especially during pivotal periods of language learning (Bastianello et al., 2022), and looking to the eyes and mouth has been associated with language ability, particularly in autistic individuals (Tenenbaum et al., 2014). Infant siblings of autistic children (Sibs-autism) are an important population for studying differences in language emergence, as these infants are at increased likelihood for autism and language delays (Zwaigenbaum et al., 2007). Looking to visual speech cues in Sibs-autism has been shown to have unconditional and indirect positive associations with language (Santapuram et al., 2022; Young et al., 2009). Findings linking looking to audiovisual speech and language development in this population, however, are mixed (Chawarska et al., 2022). This project extends prior work by evaluating whether links between looking behavior and later language in Sibs-autism and in infant siblings of non-autistic children (Sibs-NA) vary according to early language ability and/or later diagnostic status.

Twenty-eight Sibs-autism and 22 Sibs-NA from primarily English-speaking households participated in the study. Infants in the Sibs-autism group had at least one older sibling who was autistic. Infants in the Sibs-NA group had only non-autistic older siblings. At the first timepoint (Time 1; 12-18 months), infants completed a remote eye tracking task, wherein they watched a video of infant-directed speech. Proportion of looking time to the (a) mouth and (b) eyes were derived for analyses. Nine months later (Time 2; 21-27 months), infants’ language was evaluated via (a) the Communication and Symbolic Behavior Scales (CSBS), a standardized communication sample that was orthographically transcribed via video recordings, (b) the Mullen Scales of Early Learning, (c) the Vineland Adaptive Behavior Scales, and (d) the MacArthur-Bates Communicative Development Inventory: Words and Sentences checklist (MCDI). Participants also completed the Autism Diagnostic Observation Schedule, 2nd edition to evaluate autism diagnostic status. An aggregate index of expressive language was generated from the expressive language age equivalency scores from the Mullen and Vineland, the number of words spoken on the MCDI, and the number of different words spoken during the CSBS following z-score transformation. A series of regression analyses was run to evaluate the degree to which looking to the eyes and mouth at 12-18 months was associated with (a) later expressive language and (b) to statistically test whether associations of interest were moderated by later autism diagnostic status and/or sibling group.

Results indicate that proportion of looking time to the mouth was negatively associated with later expressive language (zero-order correlation = -.34, p = .02) and that proportion of looking time to the eyes was positively associated with later expressive language (zero-order correlation = .43, p = .002). When controlling for Time 1 language, relations did not hold. Sibling group at Time 2 moderated relations between looking behavior and expressive language (See Figure 1). Implications for theory, research, and clinical practice will be discussed.

Authors

• Jennifer E Markfeld, Vanderbilt University
  Presenting Author

• Kacie A Dunham

• Pooja Santapuram

• Jacob Feldman

• Samantha Bordman

• Sarah M Bowman

• Sweeya Raj

• Evan Suzman

• Bahar Keceli Kaysili, Vanderbilt University

• Tiffany G. Woynaroski