The Development of Pretend Play in Very Young Children with Autism Spectrum Disorder (ASD)

Background: Our ability to identify and diagnose ASD in toddlerhood allows study of critical developmental phenomenon such as the emergence of pretend play. However, there remains little understanding of the very early development of pretence abilities in ASD compared to children without ASD, and whether individual differences make some children with ASD more likely to have intact pretend play. Longitudinal studies are needed to examine causal factors in the early emergence and development of pretence. While it is well established that pretence and cognition are interrelated, there remains no evidence on whether this relationship is bidirectional or unidirectional in children with ASD.

Objectives: We examined the development of spontaneous pretend play in children with ASD compared to children with Developmental and/or Language Delay (DD/LD) from 24- to 48-months of age. A further aim was to investigate whether the relationship between pretend play and cognition (verbal and non-verbal) was bidirectional within the ASD group using longitudinal data.

Methods: The participants comprised 48 children with ASD and 20 children with DD/LD who were assessed at both 24- and 48-months. All children were assessed with the Mullen Scales of Early Learning (MSEL) at each age, with the target group being significantly lower on Verbal Mental age (VMA), Non-VMA and overall Developmental Quotient compared to the DD/LD group at both 24- and 48-months. Spontaneous pretend play was coded continuously (frequency and duration) off videotapes by a coder blind to group status when participants were 24- and 48-months of age.

Results: Very little pretence was observed at 24-months in both groups. A two way repeated measures ANOVA indicated a significant increase in spontaneous pretend play in both groups over time, with no group differences at either 24- or 48-months.

Significant concurrent and cross-lag correlations were found between cognition (age equivalent scores on visual reception, fine motor, expressive language and receptive language) and spontaneous pretend play at 24-and 48-months in the ASD group. Hierarchical multiple regression showed that pretend play at 24-months did not significantly predict pretend play at 48-months. With the addition of the cognitive variables, the model significantly predicted pretend play at 48-months, accounting for 27.4 % of the variance. The cognitive variables, on their own, predicted 20.6% of the variance. However, pretend play at 24-months was not a significant predictor of verbal or non-verbal cognition at 48-months, indicating a unidirectional relationship between cognition and pretend play in young children with ASD.

Conclusions: Our findings indicate that young mixed ability children with ASD engaged in equivalent levels of spontaneous pretend play at 24- and 48-months compared to DD/LD children, despite having significantly lower cognitive abilities overall. The findings support previous work with older and more able children with ASD indicating they can engage in pretend play even when it is not elicited; these findings, together, support neither a competence nor performance deficit in pretence in ASD. The findings also suggest that cognitive abilities facilitate the development of early pretend play in children with ASD and that this relationship is unidirectional.