Defining Social Engagement Deficits in Young Children with ASD: Similarities Between Retrospective and Prospective Parental Reports

Background: Social-communicative impairments are essential features of autism spectrum disorders (ASD) and are thought to appear before three years of age. However, methodological challenges have hampered investigations into the nature of these early deficits, particularly in infants/toddlers who may not receive diagnoses until the preschool years. Moreover, social engagement in young children with ASD is typically not absent, as social behaviors may be present but qualitatively different from those seen in typically-developing peers. Measures such as the Detection of Autism by Infant Sociability Interview (DAISI) (Wimpory et al., 2000) show promise in identifying deficits in social engagement associated with ASD that may aid in early identification and intervention.

Objectives:   To identify specific social engagement deficits in young children with and at risk for ASD.

Methods:  Measure: Social engagement was measured using the Detection of Autism by Infant Sociability Interview (DAISI) which is a 19-item parental interview assessing social behaviors typically occurring within the first 24 months of life.  Each behavior is scored as present or absent. The DAISI was used with two samples to identify the most robust behaviors differentiating children from ASD from other comparison groups.  Samples:  Sample 1 comprised 31 children with ASD (23 autism, 8 PDD-NOS) individually matched on CA and MA to 31 children with Developmental Delay (DD) (mean CAs = 34 months and 33 months, respectively). Parents were interviewed about their child’s past behavior in the first 2 years during a research evaluation.  Sample 2 comprised 49 later-born siblings of children with autism (Sibs-ASD) (mean CA = 15 months) participating in a longitudinal follow-up study. For this sample, parents were interviewed about their child’s current behavior. Results for children with a follow-up diagnosis of ASD (n=6) or Broader Autism Phenotype (BAP; n=8) were compared with those who received no diagnosis at follow-up (n=35) (mean CA = 34 months).

Results:  For Sample 1, ASD-DD comparisons revealed significant differences in the proportion of parents endorsing 9 DAISI items: frequency and use of eye contact, giving and showing objects, pointing and following a point, teasing, and preverbal turn-taking and communication (all ps < .05). Seven of the 9 behaviors were identical to those differentiating ASD and DD groups in the original study (Wimpory et al., 2000). For Sample 2 (Sib-ASD), significant group differences for 4 items (i.e., use of eye contact, showing, pointing, and waving) were found between the siblings who later received an ASD or BAP diagnosis and those with no diagnosis. In both samples, differentiating items received a lower percentage of positive responses in the autism-related group.

Conclusions: In both retrospective and prospective samples, young children with - or at risk for - ASD showed a distinct pattern of social engagement relative to nonautistic samples. This study replicates and extends Wimpory et al.'s (2000) findings with a larger sample and demonstrates the potential utility of the DAISI for early identification of ASD in young, high risk samples.