Gait Analysis and Motor Performance in Children with Autism Spectrum Disorders during Discrete Gait Perturbation

Background:  It has been hypothesized that studies on motor function could have significant potential in providing critical insights into the neurobiological basis of Autism Spectrum Disorder (ASD) and in improving its diagnostic characterization (Minshew et al., 2004). On the basis of decades of anatomical and imaging data (Becker & Stoodley, 2013), it has been suggested that the cerebellum could be primarily implicated in ASD. Moreover, the cerebellum has been known to control and adjust gait. Quantitative methods, such as gait analysis, could offer insight on a possible motor signature of ASD that may potentially identify a well-defined motor phenotype within the spectrum.

Objectives:  To describe gait patterns and motor performance during discrete gait perturbation in children with ASD compared to typically developing peers.

Methods:  Five children with ASD (mean age: 9.2 years) and eight typically developing children (TD) matched by gender and age were enrolled. Gait analysis was carried out in an immersive virtual environment using a 3D motion analysis system with a dual-belt, instrumented treadmill (GRAIL). After 6 min of walking at a comfortable pace, a 20-steps cycle was recorded as baseline. Then, each participant was exposed to 20 trials with a discrete gait perturbation: after a random number of steps, a single perturbation was applied to the dominant side at toe-off, using a split-belt acceleration. Immediately afterwards, a 20-steps cycle was acquired. Finally, at the end of the perturbed trials, we recorded a 20-steps cycle as post perturbation trial. Gait parameters were extracted as previously done (Biffi et al.,2015). Data recorded in perturbation trials were linearly interpolated and R² value was used to estimate the goodness of responses to perturbation. Paired and non-paired non-parametric tests were performed within and between groups.

Results:  At baseline, children with ASD had significantly increased stance time, reduced range of motion, peak extension, and peak of power at the ankle, increased time of maximum knee flexion with a more negative minimum moment, increased intrarotation of hip and increased pelvic tilt at the initial contact (all p<0.05). With respect to perturbation trials, R² values for parameters of the perturbed step were significantly higher for TD children (p<0.001). Finally, a baseline vs. post perturbation comparison showed in TD children significant increased stance time, maximum extension at hip, abadduction ROM of hip, maximum moment at hip and ROM of pelvic obliquity. In contrast, ASD showed only a reduced time of maximum flexion at hip.

Conclusions:  These preliminary findings extend an earlier investigation of our group, depicting gait abnormalities in children with ASD, as stiffer gait and difficulty in balance control (Nobile et al., 2011). With respect to response to discrete perturbation, our results indicate a small but significant adaptation in both groups; however, TD children showed a better rate of adaptation compared to children with ASD. Finally, the observation of greater aftereffects in TD children seems to confirm that they were more able to adapt and store an effective response to perturbation than children with ASD.