Assessment of Social Communication in Infants At High Risk for Autism Spectrum Disorders: A Comparison of Contexts

Background: In addition to demonstrating impairment in gesture production and use of eye gaze to coordinate attention, children with autism spectrum disorders (ASD) have difficulty producing these behaviors in combination (e.g, Wetherby et al., 2004). However, communicative performance in these children appears to be strongly influenced by contextual variations (e.g., Lewy & Dawson, 1992; Sigman et al., 1986). Therefore, it may be important to measure social communication behaviors in multiple contexts when evaluating for ASD in infancy. This study examined whether infants at heightened ASD risk (later-born siblings of children with autism; HR) demonstrate variability in communication skills across unstructured (naturalistic and toy play) and structured (Early Social Communication Scales; ESCS) contexts.

Objectives: To assess the extent to which production and temporal coordination of social communicative behaviors in HR infants vary as a function of interactional context.

Methods: Fifty HR infants (44% male) were observed at home with a primary caregiver at 8, 10, 12, 14, and 18 months of age. At 24 and/or 36 months, all infants were given a diagnostic evaluation (i.e., ADOS and clinical judgment using DSM-IV-TR criteria by a trained clinician blind to all previous study data); eight infants received a diagnosis of ASD. All infant-initiated gestures, eye gaze, and gesture+eye gaze combinations were coded during a) naturalistic interaction consisting of everyday household activities and toy play with caregiver, and b) semi-structured assessment designed to elicit social communication behaviors (i.e., ESCS).  

Results: Preliminary data were subjected to a two-way repeated-measures ANOVA with Age and Context as within-subjects factors. With regard to the mean number of gestures produced at each session, the Age x Context interaction was significant, F( 3,54) = 5.285, p = .003. The frequency of gestures increased steadily with age in the ESCS context; however, the mean frequency of gestures produced in the naturalistic context was lower than in the ESCS context and remained so over time. Regarding the mean frequency of gestures coordinated with eye contact, again, there was a significant Age x Context interaction, F(3, 54) = 6.717, p = .001. Specifically, gestures coordinated with eye contact occurred at a greater frequency in the naturalistic context at 8 and 10 months but not after 12 months. Beyond 12 months of age, gesture + eye gaze coordinations were more frequently observed in the highly structured ESCS context. In both analyses, the difference between contexts was most dramatic at the 14 month observation. Results suggest that observed frequency of communication in HR infants is highly susceptible to differences in the degree of structure afforded by the measurement context, particularly after 12 months of age. 

Conclusions: Contexts that offer more structure may provide a better indication of overall communicative competence, while familiar settings with familiar social partners may offer a more accurate picture of infants' day-to-day behavior. Results of this study suggest that each methodology provides unique and valuable information about the development of social communication skills, and that naturalistic communication samples should be utilized in concert with standardized assessments in evaluating children at risk for ASD.