Computerized Quantification of Hand Stereotypies and Postural Control

Background: Motor abnormalities are highly prevalent in Autism Spectrum Disorders (ASD) yet they are poorly understood and have not been studied with the necessary fine-grained methodologies. These anomalies encompass different aspects of the motor system including gait, apraxia, hypotonia or stereotypies. Among these motor signs, postural control (PC) defined as dynamic body stability is, in fact the earliest diagnostic marker that differentiates ASD high-risk siblings from typically developing (TD) infants. Body stability is an essential component of motor development that requires attention and integration of multisensory inputs such as visual and somatosensory cues. In view of recent findings on sensory integration processes in ASD we sought to use virtual reality to assess PC and compared the performances of children with ASD with their matched TD peers. We used motion capture technologies to examine how the brain integrates sensory information during execution of movements. Another prevalent and highly stigmatizing behavior in ASD is repetitive purposeless hand movements called stereotypies. The unknown pathophysiology and the lack of objective measurements have made their treatment largely unsuccessful. Here we expand our previous work on characterization of stereotypies, using wireless child-friendly accelerometery technologies to quantify repetitive behaviors. We examined the effects of age, IQ, autism status, occluded vision and stereotypies on postural control.

Objectives: (1) assess motor development using a clinical battery (2) apply our experimental PC protocol (3) quantify stereotypies at the kinematic and physiological levels (4) test the hypothesis that ASD children with stereotypies show significant decrease in postural stability (larger sway area) and lower motor scores compared to matched TD children.

Methods: A battery of motor, cognitive and diagnostic assessments was administered to 6 children with ASD (mean age, 5 years 3 months) and their TD peers matched on chronological age and NVIQ.  (a) Variability in sway area during PC tasks was recorded for a range of challenging postures through Optitrack and Microsoft kinect motion capture systems. Fourteen postures of 20 sec each alternating between eyes closed and eyes open were portrayed by a cartoon character and projected on a wall facing the child. (b) Hand stereotypies were video-recorded during a series of standardized activities and quantified using wireless wrist accelerometers.

Results: Preliminary analysis shows that TD children exhibit less body sway than children with ASD. Postural performance (e.g., standing on heels vs. standing on toes) differed between TD and ASD groups. The more challenging postures resulted in larger differences between the groups with eyes-closed postures resulting in larger sway in ASD than TD children. Developmental motor assessments revealed lower scores in the ASD group for timed manual dexterity and complex motor coordination tasks. The 3-axis accelerometery data point to episodes of variable length of predetermined patterned movements of abnormally high frequency.   

Conclusions: Preliminary results of reduced postural stability with occluded vision in ASD are consistent with reported underdeveloped postural control system. The quantifiable motor performances in children with ASD are in line with a neuronal-based model suggesting abnormalities in underlying cerebellar and basal ganglia networks and calls for further investigations.