Level of autistic traits in neurotypical adults predicts kinematic idiosyncrasies in their biological movements

Abstract

Introduction: Over the last decade of research, a notable connection between autism spectrum disorder (ASD) and unique motor system characteristics has been identified, which may influence social communication through distinct movement patterns. In this study, we investigated the potential for features of the broader autism phenotype to account for kinematic idiosyncrasies in social movements expressed by neurotypical individuals.

Methods: Fifty-eight participants provided recordings of point-light displays expressing three basic emotions and completed the Autism Spectrum Quotient (AQ). We extracted kinematic metrics from the biological movements using computer vision and applied linear mixed-effects modeling to analyze the relationship between these kinematic metrics and AQ scores.

Results: Our results revealed that individual differences in the total AQ scores, and the sub-scale scores, significantly predicted variations in kinematic metrics representing order, volume, and magnitude.

Discussion: The results of this study suggest that autistic traits may intricately influence the movement expressions at the microlevel, highlighting the need for a more nuanced understanding of the potential endophenotypic characteristics associated with social movements in neurotypical individuals.

Keywords: Autism Spectrum Quotient (AQ); autistic traits; broad autism phenotype (BAP); broader autism phenotype; kinematics; micromovements; movement kinematics; movement production.

Figure 1

Visualization of the 4 regions of interest (ROIs) defined based on the fiducial points labeled in the KBC toolbox: lower extremities (LE), right and left hand (RH and LH), and the head.

Figure 2

Flow diagram for extraction of kinematic features from each point light display (PLD).

Figure 3

Mean absolute value of acceleration in each 7-frame window across the expression of anger by subjects with low (left) and high (right) AQ scores. Acceleration profiles are based on the center of mass of all fiducial points. Note the periodic nature of the low-AQ exemplar (left), which corresponds to a lower entropy level due to the repeated patterning of movement and the pulsatile and abrupt nature of the high-AQ exemplar’s (right) movement which demonstrates lower AUC for a similar level of Max acceleration.