Abnormal Walking Patterns in Children with Autism Spectrum Disorder Associated with Social Impairments

Background: Children with autism spectrum disorder (ASD) have various degrees of motor impairments. Researchers have been striving to establish the association between their atypical motor patterns (e.g., motor coordination) in early development with their social function, and to use motor deficits as early behavioral markers of ASD. Previous studies have found that ASD children tend to walk with gait abnormalities, but the evidence are inconsistent. For example, retrospective video analyses showed asymmetry in arm movement in infants, but movements of lower limbs measured by motion capture system and force plate failed to identify similar asymmetry. Furthermore, gait abnormality, demonstrated by reduced stride length, increased stride width and reduced range of motion of ankle and knee angles, failed to correlate with social impairments. We postulate that kinematic measures of lower limb movement during walking vary widely across individuals and idiosyncrasy might mask ASD-related changes in gait.

Objectives: We aimed to find gait abnormity and its association with social impairments might be revealed by examining foot pressures during foot-ground contact.

Methods: Twenty-five low-functioning autism (LF, age: 5.28±0.84 yrs, IQ: 84.12±18.41), 33 high-functioning autism (HF, age: 5.50±0.61 yrs, IQ:113.36±12.03 ) and 29 age- and IQ-matched TD children (age: 5.69±0.62 yrs, IQ: 109.69±10.96, n.s. from HF) were recruited. Children were instructed to walk naturally, self-paced and looking straight ahead, for 5 meters for 10 trials. A high-density, high-frequency pressure mat (RSScan, Inc) was placed in the middle of the walkway to collect the pressure distribution of foot contacts (Fig1). The parents completed Chinese version of the Autism Spectrum Quotient: Children’s Version (AQ) and the Social Responsiveness Scale (SRS) before the experiment.

Results: We found no evidence of tendency to toe walk in ASD groups since majority of trials involved heel contact (Table1). The frequency of initial foot-contact in the toe and fore-foot regions were similar between three groups. However, during the foot roll-off, the center of pressure (CoP) started more anteriorly on the foot (initial contact of CoP) and traveled less (distance of CoP), even after removing the individual difference in foot length (Fig1). This indicates that they walk with less foot roll-off. Their stride-to-stride variability was higher as shown by the standard deviation of these two CoP measures. We also found more left-right asymmetry among ASD children, shown by significantly smaller correlation between left and right-foot CoP movements. Examining the force profiles of the metatarsal region, the predominant region for force production during foot contact, revealed a similar reduction in left-right correlation. These converging evidences indicate that ASD children have larger left-right asymmetry in foot function. Importantly, most gait variables were significantly correlated to social function (SRS and AQ scores) but not to IQ. The more abnormal the gait deficit the more severe the social impairment.

Conclusions: In conclusion, we find that foot pressure is a sensitive measurement for quantifying autism-related motor deficits and its correlation with social impairments support the proposition that motor impairments might contribute to the development of core symptoms of ASD.