Gesturing during Conversation and Free Play in Children with ASD

Background:  Non-verbal communication has long been recognized as a core deficit in Autism Spectrum Disorder (ASD), and has recently been acknowledged as a diagnostic criterion. Gesturing, including conventional, informational, descriptive and emotional gestures, forms a main modality for non-verbal communication. Individuals with ASD have been argued to show deficits both in the overall use of gestures and, more specifically, in the use of descriptive and emotional gestures. However, research on gestures is sparse and inconclusive: whereas some point to a lifelong deficit in gesture use among individuals with ASD, others support a developmental delay that may later be diminished. Yet, most studies focus on toddlers with ASD, with little evaluation of older children.
Objectives:  To compare gesturing in children with ASD to that of Typically Developing (TD) children, during a mother-child interaction. The effect of activity-type was also explored, by assessing the children with their mothers both during conversation, in which more gesturing was expected, and during free-play which was expected to elicit less gesturing.
Methods:  Forty children with ASD (clinically diagnosed, and verified using ADOS-2), aged 5-10 years, and 35 typically developing (TD) controls, matched on gender and cognitive abilities, were video-recorded with their mothers during a free-play, and during a conversation. Videos were micro-coded for use of two types of gestures: conventional and descriptive. Calculated measures included total duration of gesturing, mean duration of each gesture, and onset time for initial gesturing. Levels of autistic symptoms was parent reported, using the SRS-II.
Results:  As predicted, children produced longer conventional gestures, on overage, during conversation, compared to free play. This difference was not found for descriptive gestures. Compared to the ASD group, TD children spent more time performing gestures, and, on average, had longer durations for gestures. These differences were found for conversation but not for free play. Onset time for initial gesturing during conversation was shorter for the TD group, compared to the ASD group.
In the TD group, the level of autistic mannerisms was negatively associated with mean duration of conventional and of descriptive gesturing during conversation, but not during free-play. However, during free-play the mean duration of descriptive gesturing was positively associated with autism symptoms.
Conclusions:  The study is among the first to investigate gesture use in older children with and without ASD. Results show the effect of activity type on gesture performance in ecological parent-child interactions. Our findings support the notion that children with ASD gesture less, use shorter, less elaborate gestures, and take longer to initiate gestures, compared to matched TD children, specifically during conversation. For TD children, it was found that gesturing less during conversation is associated with elevated autism symptoms, but also that gesturing more during free-play (instead of playing and handling toys) is related to elevated autistic symptoms. Using gestures appropriately in different contexts may be used as a marker for TD children examined for ASD, though more research on the use of gestures during conversation and free play with peers and other figures, in addition to mothers, is needed.