Objectives: Since autism is not typically diagnosed until approximately 2-3 years of age, our primary goal was to investigate immediate imitation in infancy to assess its potential role in early identification. To maximize the likelihood that the sample would contain children eventually receiving an autism diagnosis and to evaluate the initiation of imitative behavior in infants at heightened risk for autism, we observed younger siblings of children already diagnosed with autism (Heightened-Risk Group, HR) and compared them to later-born infants with no such family history (Low-Risk Group, LR).
Methods: Twenty-one HR and 18 LR infants at 18-months of age were videotaped for approximately 45-minutes at home in naturalistic interaction and semi-structured play with caregivers. At 36-months, HR infants were administered the ADOS-G. At that time, three HR infants (Heightened-Risk-Diagnosed, HR-D) scored above the cutoff for an autism diagnosis. Using a coding scheme developed specifically for this project, videotapes were coded for instances of Spontaneous Imitation (i.e., imitation occurring without an interlocutor’s bid for imitation), Elicited Imitation (i.e., imitation occurring following an interlocutor’s Bid for Imitation), and Bids for Imitation (i.e., interlocutors’ attempts to elicit imitation by the infant). Spontaneous and Elicited Imitations were further coded for Type of behavior imitated (Vocalizations, Actions on Objects, Gestures, Other Bodily Movements); and Interlocutor Bids were also coded for Type (i.e., Verbal, Action, Verbal-Action Combination), whether the infant attended to the bid, and whether the bid successfully elicited imitation.
Results: Relative to LR infants, HR infants exhibited significantly lower rates (per 10 minutes) of Spontaneous Imitation (p = .016) and also received Bids for Imitation at a significantly lower rate (p = .035). There were, however, no significant group differences in percentages of bids to which infants attended or percentages of bids that were successful in eliciting imitation. Furthermore, no significant differences were found between HR and LR infants in types of behaviors imitated either spontaneously or when elicited. Finally, the three HR-D infants produced rates of imitation which fell below the HR group medians for both Spontaneous Imitation and percentage of bids leading to Elicited Imitation.
Conclusions: Results are in line with previous research indicating reduction in the rate of spontaneous initiation of behavior (e.g., communicative behavior) in HR infants. While those HR infants eventually receiving an autism diagnosis were toward the bottom of the HR distribution in imitative behavior, they did imitate; and it does not appear, based on these data, that reduced spontaneous imitation at 18 months has strong predictive value for a later autism diagnosis.