A Longitudinal Study of Gestures Used By Mothers of Infant Siblings of Children with Autism Spectrum Disorder

Background:  Typically developing children develop gestures before speech, and differences in types of gestures predict their future language ability (Iverson & Goldin-Meadow, 2005). Environmental factors, particularly parental behavior, influence child language development. Maternal gesture rates closely relate to infant gesture rates, which, in turn, predict language development (Rowe et al., 2008). Diversity of maternal speech was shown to predict concurrent and later child language ability (Rowe, 2012). Children with autism spectrum disorder (ASD) show language and gesture delays as early as 12 months (Özçalışkan et al., 2015). Infants with an older sibling with ASD are at higher risk for ASD and also show delays in language and communication. Talbott et al. (2015) reported that mothers of 12-month high-risk infants who were not later diagnosed with ASD (HRA-) gesture more than mothers of both infants later diagnosed with ASD (HRA+) and low-risk controls (LRC). For HRA- and LRC infants, these gesture rates correlated with 18-month language scores.

Objectives:  The current study expands on previous research by studying maternal gestures in relation to child gestures and language longitudinally. Our research questions were a) Do mothers of HRA-, HRA+, and LRC infants differ in gestural rates at 12 and 18 months? b) Do maternal gestures correlate with child gestures? c) Do maternal gestures predict child language through 3 years?

Methods: Mothers and 12- and 18-month infant dyads (N=73, n_HRA-=25, n_HRA+=18, n_LRC=30) were observed during a 10 minute free-play as part of a longitudinal study. Speech and frequencies of gesture types (deictic, representational, conventional) were transcribed and coded according to a gestural coding scheme by Özçalışkan & Goldin-Meadow (2009). The Mullen Scales of Early Learning (MSEL) were administered to assess language at 12, 18, 24 and 36 months, and the Autism Diagnostic Observation Scale (ADOS) was administered to determine ASD diagnosis at 36 months (confirmed by clinical evaluation).

Results:  Although mothers of HRA- children gestured more than mothers of HRA+ or LRC children, differences in total gestures were not statistically significant at either 12 (X²=4.17, p=0.17) or 18 (X²=0.41, p=0.52) months (Figure 1). Maternal gestures were significantly correlated with infant gestures at 12-months only for HRA+ infants (p=0.009, r=0.60, Figure 2A), and at 18-months only for LRC infants (p=0.002, r=0.54, Figure 2B). Maternal gestures at 12-months were positively correlated with child gestures at 18-months only for LRC infants (p=0.016, r=0.44; Figure 2C). Total maternal gestures were not significantly correlated with concurrent or later infant language scores for any of the three groups (p>0.1).

Conclusions: These results demonstrate that mothers of high-risk infants have similar gestural communication qualities as mothers of low-risk infants. Maternal gestures at 12- and 18-months correlated with LRC infant gestures at 18-months, suggesting a more robust relationship between mothers and LRC infants. However, high-risk 18-month infants had no such relationships, suggesting a lack of synchrony during communication. Understanding the nature of communication synchrony between mother-child informs the development of quality early interventions. Future analyses will focus on the interconnection and communicative synchrony between mother and child, and child contributions to conversation.