The Role of Audiovisual Synchrony in Modulating Attention to Biological Motion in Infancy

Background: Preferential attention to biological motion is present in human infants from the first days of life and plays a critical role in facilitating filial attachment and in guiding social interaction (Simion et al., 2007). In humans, biological motion includes gesture, gait, and facial expressions, actions that contain both rich social information and non-social audio-visual synchronies (AVS). By 24-months, these two features hold different significance for TD and ASD toddlers: while TD toddlers show a preference for biological motion regardless of underlying AVS, ASD toddlers preferentially attend to AVS, suggesting a disruption to normative processes of social development (Klin et al., 2009). Mapping the trajectory of how AVS influences preferential attention in typical development will establish a much needed framework for examining when deviations from this normative trajectory are first observed in ASD and how such deviations may impact developmental outcomes.

Objectives: The goal of this study is to measure longitudinal change in the association between AVS and preferential attention to biological motion in children at low-risk (LR) and high-risk (HR) for ASD.

Methods: 88 LR (49 male) and 95 HR (67 male, 50%>12 months) were shown point-light biological motion animations as in Klin et al. (2009). An upright point-light animation was presented on one half of the screen, with the inverted version playing on the opposite side in reverse order. Presentation was counterbalanced to control for side bias. Eye-tracking data, collected at months 2, 3, 4, 5, 9, 15, and 24, were used to calculate percentage of fixation time during each trial. Levels of AVS were quantified in all animations by measuring synchronous change in motion and sound.

Results: Spearman correlations revealed that preferential attention to audiovisual synchrony was developmentally modulated in LR infants: no association between preferential attention and AVS was seen at months 2, 3, 4, or 24, with the latter time point replicating the cross-sectional results of Klin et al (2009). Significant associations were observed at months 5 (r=.26, p<.001), 9 (r=.23, p<.05), and 15 (r=.36, p<.001). Data from HR infants showed increased variability at all time points. 

Conclusions: The present findings reveal a period in typical development when preferential attention is associated with AVS. Previous studies have identified strong effects of AVS on selective attention to inanimate objects during early infancy (Bahrick, Lickliter & Flom, 2004), but in the context of socially-rich biological motion, the present study shows a peak in preferential attention to AVS at 15 months. Intriguingly, this time point coincides in typical development with a period of increased visual attention to the mouth (Jones & Klin, 2013), an area of high AVS, and also to the onset of the oft-cited “vocabulary spurt” (Benedict, 1979), consistent with accounts of how language learning affects attention to faces (Lewkowicz & Hansen-Tift, 2012). Future analyses will examine associations between attention to AVS and to the mouth in the current cohort, and will focus on parsing the heterogeneity of HR infants to better understand the increased variability in their looking patterns.