Detecting motor function abnormalities in individuals with Autism Spectrum Disorder without intellectual impairment via visual-perceptive computing

Introduction Beside the core symptoms, motor function abnormalities such as dyspraxia and abnormal gait are characteristics of individuals with autism spectrum disorder (ASD). However, detailed behavioral characterization of motor function abnormalities in adults with ASD is sparse. In this pilot study, we aim at more objectively assessing motor function abnormalities in adults with ASD without intellectual impairment using visual-perceptive motion capture and explore their association with symptom severity.

Methods 37 individuals with ASD and 45 healthy subjects with an IQ > 85 and aged 18 to 65 years were matched for sex, age, verbal IQ, height, and BMI. While performing nine movement tasks, participants were filmed by a 3D-infrared camera (Microsoft Kinect for Xbox One). Assessed anatomical models were quantified using custom-made software and resulting kinematic parameters were compared between individuals with ASD and HCs using independent t-tests and Cohenʼs D. Furthermore, the association between motor function abnormalities and severity of autistic symptoms assessed with the Autism Diagnostic Observation Schedule (ADOS-2) score module 4 severity was explored in adults with ASD via Pearsonʼs correlation.

Results ASD individuals showed a reduced walking speed and cadence, and a greater mediolateral deviation while walking, greater sway during stance, tandem stance and single leg stance, a greater arrhythmicity during jumping jack tasks and an impaired manual dexterity during finger tapping tasks when compared to HCs (p < .1 and |D| > .48). Furthermore, some motor abnormalities correlated moderately to the ADOS score severity (Pearsonʼs r up to .45).

Conclusion Our results show that adults with ASD display motor patterns different from HC. Adults with ASD moved slower, had a greater postural instability and reduced manual dexterity. The data reinforce knowledge of motor abnormalities shown in children and adolescents with ASD. Also, visual-perceptive computing analysis appears to be a feasible instrument for detection of subtle motor abnormalities in ASD and may be useful in the diagnosis of ASD.