Autistic kinematics diverge from the power laws that typically govern movement

Abstract

Extant work reliably demonstrates that autistic individuals move with increased jerk (where jerk concerns change in acceleration). Although it follows that autistic movement may therefore diverge from fundamental power laws that govern movement, this hypothesis has not been directly tested to date. This lack of insight holds back progress in understanding the mechanisms underpinning differences in autism in motor control particularly with respect to movement jerkiness. Here we investigated whether movements executed by autistic adults diverged from the typical power law relationship that links movement speed and curvature. x and y position of the stylus tip was recorded at 133 Hz while 21 autistic and 19 non-autistic age-, intelligence quotient- and sex-matched adults traced, on a tablet device, a range of shapes that varied in angular frequency from 2/33 (spiral-like shapes) to 4 (square-like shapes). The gradient of the relationship between speed and curvature for each angular frequency-defined shape is reliably predicted by a set of mathematical equations often referred to as fundamental power laws thus, to assess deviations from power laws, we compared autistic and non-autistic participants in terms of speed-curvature gradients. To gain insight into potential mechanisms underpinning any differences we also used fast Fourier transform to explore amplitude spectral density across all angular frequencies. Compared to non-autistic adults, autistic adults exhibited significantly steeper speed-curvature gradients. Fast Fourier transform further revealed that non-autistic participants exhibited highly precise modulation of speed oscillations around the target frequency. For example, when drawing an ellipse their speed profile was dominated by speed changes in a band centred around the angular frequency 2 with minimal changes in other bands. Autistic adults, in contrast, exhibited less precise modulation of speed oscillations around the target frequency, a result that is reminiscent of a literature reporting broader auditory filters in autistic individuals. These results evidence, in autistic adults, a deviation from the power laws that typically govern movement and suggest differences in motor cortical control policies and/or biomechanical constraints.