Stability and Change in Language Development from 4 to 11 Years in Verbal Children with Autism Spectrum Disorder: A Community-Based Study

Background: Language difficulties are a common feature of autism spectrum disorder (ASD) and are associated with adverse outcomes including literacy difficulties, challenging behaviour and aggression (Mawhood 2000; Baghdadli 2003; Hartley 2008). Yet we understand little about how language develops for children with ASD compared with other children, particularly beyond the preschool years.

Objectives: We compared language ability at ages 4, 5, 7 and 11 in four groups of children: ASD with language disorder (ASD+LD; n=17), ASD without language disorder (ASD-LD; n=30), developmental language disorder (DLD; n=107) and typical language (TD; n=872). We also investigated predictors of language outcome in ASD.

Methods: Participants were selected from a large, prospective community-based cohort study of child language. A comprehensive, standardised language assessment (The Clinical Evaluation of Language Fundamentals- Fourth Edition; CELF-4) was used to assess receptive and expressive language across four time points. Mean scores on the CELF-4 and slopes (rate of language growth) were estimated from 4 to 11 years using generalised estimating equations. We used linear regression to analyse predictors of language outcome at 11 years.

Results: There was individual variability in scores and rate of growth for children in all four groups. For expressive language, children in the ASD-LD group had estimated mean scores that were 1.09 units lower than the TD group indicating similar ability at 4 years. By contrast the ASD+LD and DLD groups both had substantially lower estimated mean scores than the TD group (33.45 and 31.84 units lower, respectively). The estimated mean difference in slopes was similar for the ASD-LD and TD groups (p=0.905) indicating comparable rate of growth in language from 4 to 11 years. There was, however, a significant difference between the DLD and ASD+LD groups compared to the TD group (p=0.001 and p=0.003, respectively) indicating mean standard scores increased more quickly for the DLD and ASD-LD groups relative to the TD group from 4 to 11 years. This increased rate of growth was particularly evident for the ASD+LD group (Figure 1). The overall findings for receptive language were comparable to those for expressive language although rate of growth for the ASD+LD group was less rapid (Figure 2). Language at 4 years was the only consistent predictor of language at 11 years in the ASD group (p=0.001 for both receptive and expressive language).

Conclusions: For language ability and rate of growth, children with ASD-LD had similar profiles to children with TD, and children with ASD+LD had similar profiles to those with DLD. While rate of growth followed a predictable pattern (based on norms) for the TD and ASD-LD groups, those with DLD and ASD+LD demonstrated some developmental ‘catch up’ between 4 and 11 years. Language at 4 years predicted language at 11 years for ASD. These data can assist parents to better understand their child’s language prognosis and inform intervention and service planning. The findings also contribute to our understanding of critical time periods for growth and development in children with ASD and how this may be similar or different to other groups of children.