Examination of Motor Sequence Learning Over Multiple Sessions in Children with Autism

Background: Autism is a developmental disorder characterized by abnormalities in the acquisition of social, communicative and motor skills. Given the developmental context of autism spectrum disorders, careful examination of mechanisms underlying learning and memory may be critical to understanding the neural basis of autism and to refining treatment. Motor sequence and other forms of procedural learning have thus far been examined in autism over a single session. However, acquisition of skill through procedural means tends to be a gradual process that results in acquisition of skills that are very stable over time.

Objectives: To examine motor sequence learning in children with high-functioning autism (HFA) over the course of multiple sessions.

Methods: 5 children with HFA and typically developing (TD) 5 children completed a Rotary Pursuit (RP) task over two consecutive days with two sessions a day (sessions 1 to 4). Each session consisted of four blocks and each block consisted of three 30-second trails. Improving performance on the RP task involves learning a sequence of complex movements that anticipate the motion of a target in a novel pattern. Each subject performed one of two motor patterns (either circle or square), with the first session of day 2 (session 3, block 3) also including a single interference block, so that children performing the circle pattern were exposed to a block of square trails and children performing the pattern square pattern were exposed to a block of circle trials. For both experiments, learning was assessed using repeated measures ANOVA (RM-ANOVA) to examine change in time-on-target across blocks.

Results: Analysis across all sessions revealed a significant main effect of diagnosis (F1,8=6.9, p<0.03), with children with HFA showing less overall time-on-target than did TD children.  Across both groups, there was a significant main effect of sessions (F3,24=4.6, p<0.02), with an increase in time-on-target  Both groups showed increased performance over blocks and sessions, however children with HFA showed less time on target than TD children in all the 4 sessions. Analysis of the interference effect in session 3 (change from block 2 to 3) revealed that while TD children showed a near significant decrease in time-on-target (F1,5=4.6 p=0.08), children with HFA (p>0.2) did not.

Conclusions: The RP findings over multiple sessions indicated that while children with autism showed less time-on-target than do TD children, both groups showed improved performance consistent with motor learning.  There is suggestion, however, that for children with autism the mechanism used to acquire the motor sequence may have differed from TD children; children with HFA showed show less decrement in performance during the interference block compared to TD children.  The lesser interference effect of a contrary visual pattern may be indicative of decreased reliance on visual feedback to guide acquisition of novel sequences of movement.  The findings have implications for therapies aimed at helping children with autism to acquire new skills, motor skills as well as aspects of social skills that depend on learning novel motor gestures.