The effect of limb selection methods on gait analysis in Parkinson?s disease

Introduction: Asymmetry of motor symptoms is a common characteristic of Parkinson's disease (PD), yet gait outcomes are often reported as limb averages or authors fail to report which limb is being analyzed. This study aimed to investigate how varying limb selection methods may impact statistical comparisons of common gait measures amongst fallers and non-fallers with PD. Methods: Overground walking data was collected on 53 fallers and 117 non-fallers during routine clinical visits. The relationship between limb selection method (left, right, most-affected, least-affected, and limbs averaged) and faller status (faller vs non-faller) on spatiotemporal gait parameters was analyzed using a mixed linear model. Results: Significant interactions between limb selection method and faller status were found for step time variability and swing time variability. Regardless of selection method, it was possible to discern significant differences between fallers and non-fallers. Yet, if researchers only analyze the least-affected limb during gait analysis, the differences between fallers and non-fallers are less apparent. Conclusion: In individuals experiencing uneven laterality of symptoms that affect gait, limb averaging may alter interpretation of statistical findings and mask compensation patterns. This study promotes a refined gait analysis process, particularly in individuals that present with possible asymmetric walking. Including limb selection methods in future studies encourages holistic and transparent analyses within the literature.