Early Play Behaviors in Infants at Risk for Autism

Background: Few studies have found early markers of autism that are associated with later diagnosis. Various studies examined early play behaviors after nine months of age but none have looked at early play behaviors during mid-infancy. Early play behaviors include the ability to move and explore one’s environment, play with objects, and attend and interact with a caregiver. These play abilities may provide valuable information about sensory, motor, cognitive, and social functioning that may predict risk factors for ASD. 

Objectives: This study examined the relation between early play behaviors at age 6 to 7 months and 36-month outcome classification of ASD in high and low risk infants. 

Methods: A retrospective analysis of longitudinal data was collected from a cohort of 58 infant participants by coders blinded to child’s diagnostic category. Play skills were rated from videotapes of administration of the Mullen Early Learning Scale and a Peek a boo task. The Revised Knox Preschool Play Scale was used to code these dimensions of play: gross motor skills, fine motor skills, early imitation and early social and communication behaviors (e.g. participation) during interactions (Knox, 1997, 2008). Researchers operationally defined play categories and established reliability in their scoring. Social and communication behaviors associated with a diagnosis of ASD were determined at age 3 by expert clinical researchers using the DSM IV TR criteria (2000) for autistic disorder and pervasive developmental disorder not otherwise specified along with performance on the following diagnositic assessments: The Autism Diagnostic Observation Schedule, Mullen Scales of Early Learning Receptive and Expressive Language Scales, and Preschool Language Scale.

Results: A Mann-Whitney U test was run to determine if there were differences in play scores between the infants with ASD and infants without an ASD diagnosis. Distributions of the scores were not similar, assessed by visual inspection. Play score for infants later diagnosed with ASD at age 3 had lower play abilities at 6 months compared to infants without an ASD diagnosis (U = 130, z = -2.29, p < 0.05, r = 0.31). Analysis of the play dimensions revealed that infants with ASD had a lower imitation score  (U = 145.5, z = -2.26, p < 0.05, r = 0.31) and participation score (U = 138, z = -2.30,  p < 0.05, r = 0.31) compared to the NonASD group.

Limitations: Play skills were observed retrospectively during the administration of a standardized assessment rather than in a naturalistic play context. In addition, there were only a small number of infants diagnosed with ASD in the sample.

Conclusions: Play behaviors may be early indicators of developmental disruption in infants later diagnosed with ASD. Findings suggest the need to move beyond standard assessments that focus on the accomplishment of gross or fine motor tasks to look also at the quality of interaction with a caregiver during early play, which may serve as potential indicators of ASD.