The Early Developmental Trajectory of Initiating Joint Attention, but Not Expressive Vocabulary, Predicts Later ASD Severity in ASD-Sibs

Background:

Joint attention and expressive vocabulary are two key facets of social communication that undergo rapid change early in life. These two communicative competencies have been examined in the at-risk siblings of children with an Autism Spectrum Disorder (ASD; ASD-sibs) as compared to the infant siblings of typically developing children (COMP-sibs). It remains unclear, however, when differences between the two groups emerge as the developmental trajectories of these behaviors across multiple time points have not been evaluated.

Objectives:

To describe the developmental trajectories of initiating joint attention (IJA) and expressive vocabulary in ASD-sibs and COMP-sibs between 12-18 months, and the extent to which they predict later ASD symptom severity at 24 months.

Methods:  

IJA and expressive vocabulary were examined in ASD-sibs (n=50) and COMP-sibs (n=39) at 12, 15, and 18 months. IJA was measured as the number of Directing Attention items passed on the Screening Tool for Autism in Toddlers (STAT). Expressive vocabulary was measured as the Understands and Says score on the MacArthur Communicative Development Inventory. ASD severity was measured as the calibrated severity score on the Autism Diagnostic Observation Schedule at 24 months in 28 ASD-sibs and 24 COMP-sibs for whom data were available.

Results:

Hierarchical Linear Models were used to model IJA and expressive vocabulary. IJA had an intercept significantly different from 0 that varied between infants, β₀₀=1.41, p<.01, and significant fixed linear change, β₁₀=.15, p<.01. Group status had a significant effect on the intercept, β₀₁=-.48, p<.01; ASD-sibs had lower overall IJA than COMP-sibs.

Expressive vocabulary had a fixed intercept significantly different from 0, β₀₀ =5.79, p<.01, fixed non-significant linear change, β₁₀=-.13, p=.89, and significant quadratic change that varied between infants, β₂₀=1.85, p< .01. IJA was a significant time-varying predictor of expressive vocabulary, β₃₀ =1.57, p=.04. Group status had a significant effect on the quadratic change, β₂₁=-.83, p<.02, as ASD-sibs had a slower upward curvature (increase) in expressive vocabulary than COMP-sibs between 12 and18 months.

The ordinary least squares estimates of the intercept of IJA and the quadratic change of expressive vocabulary (modeled independently of IJA) were examined as predictors of ASD severity. Among ASD-sibs, individual IJA intercepts predicted ASD severity, r(26)=-.75, p<.01; this association was not significant for COMP-sibs, r(22)=-.17, p=.42. Quadratic change in expressive language was not associated with ASD severity in either group, ps<.53.

Conclusions:

Differences between ASD-sibs and COMP-sibs in the developmental trajectories of joint attention and expressive vocabulary emerged within the first 18 months. Relative to COMP-sibs, ASD-sibs demonstrated slower growth in expressive vocabulary and showed lower levels of IJA at baseline and through 18 months (due to similar growth rates). Among ASD-sibs, higher baseline levels of IJA predicted lower ASD severity. Higher levels of IJA also predicted higher levels of expressive vocabulary for both groups. Overall, difficulties with IJA evident by 12 months (baseline) may represent an early marker of later ASD impairment.