The Role of Joint Attention in Predicting Gains in Language and Social Interaction Skills in Very Young, Minimally Verbal Children with ASD

Background:  Joint attention (JA) skills, and particularly the ability to initiate joint attention (IJA), have been shown to predict receptive and expressive language development in young children with ASD (Charman et al., 2003; Mundy, Sigman, & Kasari, 1990). Fewer studies have focused specifically on JA indices as predictors of language and social interaction skill development in very young (<3 years) minimally verbal children with ASD. 

Objectives: We explored whether one-year gains in language and social interaction skills were associated with the following baseline joint attention indices in toddlers with ASD who were minimally verbal: initiation of, and response to, joint attention, behavioral requests, and reciprocal social interactions. 

Methods:  Participants were 62 children with ASD assessed first at 18 to 33 months (Time 1, T1; M=28±4) and again approximately one year later (T2). All participants were minimally verbal at time 1, defined by Mullen Scales of Early Learning expressive language t-scores of 20. Social interaction skills were assessed with the ADOS Reciprocal Social Interaction scores and mother-reported Vineland Adaptive Behavior Scales (VABS-I) Socialization standard scores. Language abilities were assessed with the Mullen Scales of Early Learning Expressive and Receptive Language t-scores, and JA indices assessed with the Early Social Communication Scales. Regression analyses examined whether T1 JA indices independently predicted improvements in receptive and expressive language and social interaction skills, represented by T1-T2 difference scores. T1 Mullen nonverbal developmental quotient (NVDQ) and/or T1 age were covaried in regression models when correlated with the dependent variable. 

Results:  Prior to examining regression models, pairwise comparisons established significant improvements from T1 to T2 on all outcome measures (ts > |3.5|, ps < .001). Younger age at T1 and higher T1 NVDQ scores were associated with greater gains in receptive language. Children who were younger at time 1 showed greater improvement in ADOS Reciprocal Social Interaction score, whereas the reverse was true for the VABS-I Socialization scores. T1 IJA and response to JA (RJA) significantly predicted gains in receptive language skills (β = .29, p = .02; β = .47, p < .001). JA indices were unrelated to gains in expressive language. T1 response to social interaction significantly predicted greater social interaction symptom reduction on the ADOS and larger gains in VABS-I Socialization score (β = .26, p = .04; β = .33, p = .01). 

Conclusions:  Previous literature highlighted early initiation of, and response to, joint attention as predictors of both receptive and expressive language gains. Our results show that for these young minimally verbal children, early IJA and RJA skills were associated with gains in receptive language and were unrelated to gains in expressive language. Improvements in social interaction skills and reductions in social interaction symptoms were associated with responsiveness to other social situations (e.g., reciprocal social games).