Responding to and Learning Joint Attention: A Comparison of Infant Siblings of Children with Autism and Typically Developing Infants

Background: Regardless of whether they are subsequently diagnosed with autism, infant siblings of children with autism (IS) may attend differently to faces and objects, and to social cues linking faces and objects, than typically developing (TD) infants (McCleery et al., 2009). The ability to follow another's gaze to an object, or responsiveness to joint attention (RJA), is often impaired in autistic children and may be atypical in infant sibs by 15 months of age (Presmanes et al., 2007; Sigman et al., 1999). However, autistic preschoolers who demonstrated limited spontaneous RJA exhibited more joint attention after a training period when a model's gaze was predictive of the activation of a toy (Leekam et al., 2000).

Objectives: We examined responsiveness to social cues and spontaneous and learned RJA in a cross-sectional sample of IS and TD infants.

Methods: Each infant watched a video of a model turning to 1 of 2 objects on either side of her. Eye movements were recorded with an eye tracker. Each trial consisted of a baseline phase, an infant directed greeting and smile, and the model turning and fixating an object for 5 seconds. Direction of turn was counterbalanced across trials. There were 4 “spontaneous RJA” trials with identical legos, 4 “spontaneous RJA” trials with non-identical objects, and 6 “learned RJA” trials in which the attention getter moved and emitted music for the last second of the model's fixation. Proportion of time attending to the face relative to the objects was calculated and RJA was indexed by proportion of time the infant attended to the object the model was looking at relative to the other object. After the eye tracking session, infant sibs also participated in the ESCS. Once ESCS scores have been calculated, we will correlate RJA scores on the ESCS to those on the eye tracking measure to test the external validity of the eye tracking measure.

Results: At 6 months, TD infants spent proportionally more time looking at the face relative to the objects than IS infants, t(36)= 2.21, p=.046 (TD n=26, M=.92, SD= .16; IS n=12, M=.72, SD= .31). No differences in this measure were observed at 12 months, and no differences in spontaneous RJA were observed at either age. At 6 months IS infants exhibited more learned RJA when the model's gaze was predictive of the subsequent motion of the rattle, t(16)=- 2.48, p=.024 (IS n=7, M=.82, SD=.37; TD n=11, M=.36, SD=.39). At 12 months the reverse pattern was observed: TD infants demonstrated more learned RJA than IS infants, t(14)=2.88, p=.012 (TD n=9, M=.73, SD=.39; IS n=7, M=.22, SD=.29).

Conclusions: These preliminary results suggest that early in development infant sibs may be more interested in objects and predictive contingencies than typically developing infants. This is consistent with previous work demonstrating that infant sibs looked away from their mother's faces during a still face paradigm more than typically developing infants at 6 months and showed quicker neural responses to objects than faces at 10 months (Ibanez et al., 2008; McCleery et al., 2009).