The Influence of Goals On Movement Kinematics and Eye Movements During Imitation in Autism

Background: It has been reported that individuals with autism have reduced imitation abilities, specifically when the action to be imitated does not have a goal (Hamilton 2008). Imitation is a process which is important in learning how to communicate, and understand others, and therefore may have implications for understanding autism. It has been proposed that reduced imitation abilities in autism may be a result of impaired sensory motor integration (Gowen and Miall 2005), which may have a greater effect on the direct visuomotor mapping of novel, non-goal actions, than on actions directed towards a goal, which may rely more heavily on stored representations.

Objectives: To investigate the influence of visual goals on movement kinematics and eye movements, during imitation in autism.

Methods: Sixteen adults (mean age 30.63 ± 7.9 yrs, 5 females) diagnosed with Autism or Asperger Syndrome, along with sixteen age, sex and IQ matched controls took part in an imitation experiment. Participants were required to observe, and then imitate, movie clips of hand movements, which were either directed towards goals, or towards nothing (no-goals). Goal and non-goal movements were displayed in separate blocks, each made up of movements which varied in speed (slow/fast), size (normal/short) and trajectory (flat/elevated). Hand movements were recorded using a magnetic Polhemus motion sensor, which was attached to the index finger of the dominant hand, recorded at 120 Hz. Eye movements were recorded using an Eyelink II head-free eye tracker,  and were recorded at 250 Hz. The degree to which participants modulated their own movement parameters with changes in the observed movement was investigated.

Results: Movement duration was not adjusted when goals were present; both groups performed movements of similar duration during slow and fast trials. When goals were absent, however, the control group exhibited longer movement durations in slow trials compared to fast trials, but the autism group did not differ. The same pattern was found for the average peak velocity of the movement. In measures of movement amplitude, the autism group performed better than the control group, showing a larger difference in movement size between short and normal trials. Eye movement data showed that the autism group spent a significantly longer time looking at target areas, and significantly less time between targets, in both goal and no goal trials.

Conclusions: The results show that people with autism are able to successfully imitate aspects of goal-less actions, namely the start and end-points of movement, which are perhaps attributed goal status in order to be imitated. In contrast to the control group, kinematics inherent in the movement, such as duration and velocity, were not successfully modulated in the autism group during goal-less imitation. Eye data indicates that the autism group spent more time looking at goal areas, even in the non-goal condition. This implies that people with autism do not switch to the direct visuomotor pathway for the imitation of goal-less action, supporting the idea that the route may be impaired as a result of deficient integration of visual information with the motor system.