There is a growing body of research suggesting that young children with ASD have imitation difficulties. However, the prevalence of early imitation problems and its risk factor for ASD at preschool age is unknown.
To establish imitation retardation in preschoolers suspected of ASD using a cohort type diagnostic accuracy study design.
86 children (aged 1.9-4.5 years) suspected of autism and consecutively referred to University Autism Clinics in Flanders (
A multidisciplinary clinical consensus classification revealed 68 children with ASD and 18 children with Non-Spectrum Developmental Disorders (NS-DD). The two groups did not differ significantly in chronological and nonverbal mental age.
Differences between imitation, language, motor age-equivalents and nonverbal mental age were used to predict the diagnosis of autism.
Bodily (gestures and facial expressions) and procedural (actions with objects) imitation levels were determined with the Preschool Imitation and Praxis Scale (PIPS; Vanvuchelen, 2009).
The PIPS consists of 30 actions with different effects (salient environmental and internal), representational levels (meaningful and non-meaningful, goal directed and non-goal directed), temporal complexities (single and sequential) and visual monitoring possibilities (transparent and opaque).
Bodily and procedural imitation age-equivalent scores were derived from PIPS scores of 654 typically developing children between 1 and 4.9 years of age.
Receptive and expressive language levels were examined with the Dutch version of the MacArthur-Bates Communicative Development Inventories (N-CDI; Zink & Lejaeghere, 2002) and the Reynell Developmental Language Scales (RTOS; Schaerlaekens, Zink, & Van Ommeslaeghe, 2003).
Gross and fine motor levels were determined with the Peabody Developmental Motor Scales-2 (PDMS-2; Folio & Fewell, 2000).
The nonverbal mental level was examined with the Dutch version of the Bayley Scales of Infant Development (BSID-II-NL; Van der Meulen, Ruiter, Lutje Spelberg, & Smrkovskı, 2000) and the Revised version of the Snijders-Oomen Nonverbal Intelligence Test for Children (SON-R 2.5-7; Tellegen, Winkel, Wijnberg-Williams, & Laros, 1998).
There is evidence for bodily and procedural imitation retardation in relation to nonverbal mental age in 72%, respectively 69% of the preschoolers with ASD.
Two factors were found to be significantly associated with ASD using simple logistic regression analyses: procedural imitation [OR 1.07 (95% CI: 1.01-1.13)] and receptive language retardation [OR 1.08 (95% CI: 0.99-1.16)].
Using the optimal cut-off for the procedural imitation retardation score based on the Receiver Operating Characteristic (ROC) curve yields a sensitivity of 82% (95% CI: 71%; 90%) and specificity of 50% (95% CI: 26%; 74%).
In a multivariate model, only procedural imitation retardation remained a significant predictor of ASD [OR 1.21 (95% CI: 1.05-1.40)].
Findings suggest that retardation in procedural imitation that goes beyond the nonverbal mental retardation may predict the diagnosis of autism.
Results are new to the literature and therefore require replications in other settings.