Joint Attention at 22 Months As a Predictor of Communication Skills in Preschool

Background:  Early joint attention (JA) skills are known to predict later language skills in children with ASD (Charman, 2003; Poon et al., 2012). However, little is known about the extent to which these skills may be variably related to general language outcomes versus more pragmatic aspects of language use. We aimed to address this gap in the literature via analyses using a sample of children (n=46) who were identified as at-risk for ASD at 12 months of age using a community screener (FYI; Baranek et al., 2003). Specifically, we sought to examine the extent to which general language outcomes versus pragmatic language outcomes at preschool age were predicted by JA skills at 22 months.

Objectives:  Our aim was to determine if JA skills at 22 months variably predict general language outcomes versus pragmatic aspects of communicative competence, specifically reciprocity, initiation, and topic maintenance during conversations and narrative retell ability.

Methods:  Items on the ADOS Module 1 targeting JA skills (Showing, RJI, and Spontaneous IJA) collected at 20-24 months (m=22, sd=.8) were used in regression models to predict general language skills on the Preschool Language Scales-5^th Edition (PLS-5; Zimmerman et al., 2011). Next, JA skills at 22 months were used to predict specific pragmatic skills on the Pragmatic Rating Scales-School Age (PRS-SA; Landa, 2011; Reciprocity, Initiation, and Topic Maintenance), and narrative retell abilities using the Narrative Scoring Scheme (NSS; Heilmann et al., 2010) for a sub-sample of 4-5 year olds (m=59.72 months, sd=6.7), whose language levels were advanced enough to code using the PRS-SA.

Results:  In this sample of children identified at 12 months as at-risk for ASD, the three items targeting JA on the ADOS at 22 months accounted for a significant amount of the variance in Expressive Communication (R²= .21, p=.02) but not Auditory Comprehension on the PLS-5 (R²=13.98, p=.09) in preschool. RJA was the only significant predictor of Expressive Communication, accounting for most of the variance in preschool expressive language skills. None of the JA items were significant predictors of Reciprocity or Initiation on the PRS-SA, but JA items at 22 months accounted for significant amounts of variance in PRS-SA Topic Maintenance (R²=.34, p=.03) and total narrative retell skills on the NSS (R²=.31, p=.02) in preschool. Showing (p=.06) and RJA (p=.06) were the biggest contributors of variance in Topic Maintenance, while Showing (p=.01) was the only significant predictor of narrative retell. The contributions of the JA items in all of these models accounted for significant variance above and beyond the ADOS total algorithm score, which was not a significant predictor in any of the models.

Conclusions:  In this at-risk sample, JA skills at 22 months significantly predicted expressive language ability in preschool. However, early JA skills accounted for a larger amount of variance in the children’s maintenance of another person’s topic and narrative retell ability than they did in general expressive language skills at preschool age. Findings support the importance of interventions targeting JA skills in toddlers at-risk for ASD to improve later pragmatic functioning.