Neural Basis of Interlimb Coordination during Walking in Children

People swing their arms unconsciously in a reciprocal manner during walking. This feature seems quite useless for locomotion at first sight. After all, people are able to walk without the arms swinging, or when the arms are making other (complex) movements (such as waving the arms or juggling). Even though the arm movements are unnecessary to walk, they do have some beneficial effects on the walking pattern. For instance, when the arms are swinging during walking, less energy is consumed compared to when they are not swinging. This positive effect of arm swinging on gait, however, does not explain what causes people to swing their arms without having to think about them. Therefore, in the current doctoral thesis, we wanted to gain more insight in the neural basis of the natural reciprocal arm swing. It is described in literature that arm muscle activation may come about by the way the nervous system is built, with interconnected Central Pattern Generators (CPGs) generating locomotion patterns. It is suggested that bipedal and quadrupedal locomotion share common spinal neuronal control mechanisms, which is based on the assumption thatduring evolution, man started to walk bipedally, and the circuitry, previously used for the arms during locomotion, remained operational. TheseCPGs generating locomotion patterns are located in the spinal cord, where they are interconnected and controlled by brainstem and cortical circuits. Some authors, however, have emphasized the direct control from the cortex. If arm swinging during gait would primarily originate fromcortical contributions, one would expect this to be reflected in deteriorated or altered arm swing in persons with a cortical deficit. To this end, the studies in the current doctoral thesis have focused on childrenwith spastic hemiplegic and diplegic Cerebral Palsy. These are the two most common types of Cerebral Palsy (CP), in which non-progressive impairment to the brain resulted in abnormal limb strength, control, and/or muscle tone. In children with hemiplegia, one side of the body is more affected than the other, while in children with diplegia the lower extremities are more affected than the upper extremities. In childrenwith CP, systematic studies about the arm movements during walking are very scarce. Therefore in the first phase of the current doctoral thesis, different aspects of the arm behavior during walking in these childrenwere examined.
The different aspects under investigation were arm swing amplitude, arm posture, and the interlimb coordination. We found that, overall, children with CP moved their arms differently during walkingand this was reflected in all three evaluated aspects. In particular, children with CP were unable to further increase their arm swing amplitude to the same extent as typically developing (TD) children when walking faster. Moreover, children with hemiplegia specifically showed increasedarm swing on the non-hemiplegic side and decreased arm swing on the hemiplegic side. With respect to arm posture, both CP groups presented withand altered posture on both sides of the body. They held their hands higher and more in front of the body with their upper arm was rotated more to the posterior. Again, children with hemiplegia showed a clear asymmetry. Their hemiplegic arm was held in an even higher position. As expected, these alterations in arm swing and posture affected interlimb coordination during walking as well. Specifically, in children with hemiplegia,the hemiplegic arm impaired coordinative stability and constrained the synchronization of the limbs. In contrast, in children with diplegia, the legs limited the ability to coordinate the limb pairs, but it were thearms that affected coordinative stability.
Overall these alterationsand deficits seemed to be related to secondary causes, such as spasticity, muscle weakness, and compensations (for stability or angular momentum), rather than they are directly related to the primary cortical deficit. Thus, this newly obtained knowledge is crucial in order to know whichaspect of the arm behavior should or should not be adapted in order to improve the overall walking pattern of a patient with CP.
From the experiments of phase one, it already appeared that arm swing does not entirely depend on cortical control because, despite the difference between children with CP and TD children, the basic pattern was maintained. In the second phase of the current doctoral thesis we further explored the role of the cortex in the neural basis of the arm movements during gait. To this end, we used the forward walking (FW) and backward walking (BW).This paradigm allowed us to infer about the similarity in the neural mechanisms controlling the limbs for the different directions of walking. This has been done in earlier studies for the leg kinematics, which werefound to be similar between FW and BW reversed in time, but this has not been done for the arm kinematics. Hence, this led us to investigate whether, as for leg movements during walking, the kinematical patterns of the arm movements during FW and BW would be equivalent but reversed in time. The results, indeed, demonstrated this similarity in healthy participants (i.e. TD children), and supported the idea that the neural control of the locomotor arm movements is organized in a similar way as for the leg movements. Further, in order to differentiate whether the neural control of locomotor limb movements primarily arises from a cortical source or from peripherally located networks, we determined whether an intact cortex is needed to sustain the simple kinematical reversal from FW toBW. To this end, we investigated whether in children with CP the same kinematical reversal for the arm and leg movements took place from FW to BW as in TD children. It turned out that also in children with CP the degree of similarity between the limb kinematics of FW and BW was considerable, which indicated that the neural mechanism of interlimb coordination during walking does not depend mainly on a cortical source.
In summary, the above described findings have mapped the uncharted impairments of arm behavior (i.e. arm swing, arm posture and interlimb coordination deficits) arising during walking in children with CP. This knowledge canbe used to aid in gait rehabilitation when attempting to implement arm movements in gait training programs. Furthermore, fundamental knowledge is gained about the possible causes (i.e. spasticity, muscle tone, compensation strategy) of the altered arm behavior in children with CP. Studying the arm behavior during FW and BW allowed us to acquire further insights in the neural mechanisms controlling locomotor arm movements (i.e. the neural mechanisms for the lower and upper limbs are organized in a similar manner). Furthermore, comparing the results of TD children with children with CP, increased our understanding where the neural control oflocomotor limb movements predominantly originates from (i.e. from sitesmore peripherally located than the cortex such as the brain stem or thespinal cord). The insights provided by the current doctoral thesishave opened the way for further research on the implementation of arm movements in the gait rehabilitation in children with CP and on the relative contribution of the different neural areas/networks in control of locomotor limb movements.

Keywords:Forward and backward gait, Split-belt treadmill, Overground walking, Primitive reflexes, MRI, Cerebral palsy, Interlimb coordination, Arm swing

Disciplines:Education curriculum

Project type:PhD project