Visuomotor Impairment and Underlying Cortico-Cerebellar Dysfunctions in Individuals with Autism

Background:  Dyspraxia is common in individuals with autism, but it remains unclear whether sensorimotor abnormalities reflect impaired processing of sensory feedback for action planning, or fundamental deficits in motor control.

Objectives:  To clarify the integrity of sensorimotor and motor control systems in autism. 

Methods: 26 individuals with autism and 26 matched healthy controls (ages 8-35) participated in two laboratory studies of motor skills; 15 individuals from each group also completed an fMRI study.  Subjects performed sustained precision grip force tasks in which the amplitude of the target force and the precision of visual feedback were varied independently.  They viewed a white bar that moved upwards with increased force toward a fixed green target bar.  Subjects were instructed to sustain a constant force in order to stabilize the white bar at the level of the green bar for 15 sec trials.  During the “motor” manipulation, the green target bar was set to 5, 25, 45, 65 or 85% of individual subjects’ maximum force contraction.  During the “sensory feedback” manipulation, the vertical distance the white bar moved per Newton of force applied by the participant was set to visual angles of .02, .19, and 2.02 deg. When the visual angle was large, the white bar moved a larger distance for every Newton of force applied, increasing the precision of sensory feedback. 

Results:  Subjects with autism did not show atypical mean force levels, but they did show reduced control of their motor output as demonstrated by increased variability of sustained force during the course of trials (F=14.87; p<.001).  This impairment was more robust at greater force amplitudes (Group x Amplitude: F=9.47; p<.001).  Force variability was increased in individuals with ASD across varying precision of sensory feedback (F=7.58; p<.01), but particularly when feedback was less precise (Group x Visual Angle: F=4.28; p=.02).  In fMRI studies, when sensory feedback was less precise, subjects with autism showed reduced activation in left motor cortex, left dorsal premotor cortex, superior parietal lobule and cerebellar lobules IV-VI.  These brain abnormalities were not evident when sensory feedback was precise.  Increased variability of sustained force was associated with clinical ratings of communication impairment and motor stereotypies in individuals with autism.

Conclusions:  Our results provide evidence that visuomotor impairments in autism reflect deficits in controlling sustained motor output and in transforming sensory input for use in precise motor control.  Further, motor impairments appear to be related to communication deficits and stereotypies, indicating a relationship with core clinical features of the disorder.  Our fMRI studies highlight dysfunction in cortico-cerebellar circuitry which integrates sensory input to adjust movement plans organized by frontoparietal motor systems.  Combined with findings from postmortem studies of autism documenting reduced Purkinje cell size and density, these results suggest that abnormal output from the cerebellum to neocortical systems may underlie the dyspraxia and poor fine motor control that are present in the majority of individuals with this disorder.