By Arnaud Gouelle, PhD, and Patrick Roscher, MS
How do we use gait analysis?
Have you ever asked yourself this simple question: why do we measure and analyze gait? Overall, the answers will revolve around the same ideas: to gauge the functional status of a person; to follow-up the natural history of a disease; to determine immediate or long-term treatment requirement and effects. Now, if the question is what main information are you taking from a gait analysis, answers will surely vary throughout different clinical expertise (see “Of Gait Analyses and Elephants,” page 9). An orthopedic surgeon will be interested in deviations in joint kinematics and kinetics; a physiotherapist will look at a more generic level of the motor performance, considering the whole speed or the step length asymmetry; a geriatrician will be interested in quantifying the fall risk or the impact of a cognitive load on the locomotion; a podiatrist will investigate gait at a foot level through the distribution of the weight during the roll-on. Similarly, a person’s academic or professional backgrounds guide the perception that a person will have about the use and the interpretation of a clinical gait analysis.
On one hand, this diversity illustrates the recognized value and wealth of gait analysis. While walking seems to be a simple movement, it is in fact a particularly complex function, combining the musculoskeletal and nervous systems with a multifaceted relationship between the automated and voluntary parts of control. From the brain to the effectors, most disturbances in the physical or cognitive status can have an effect which becomes noticeable in gait features. Even tiny details such as chewing or being in a bad mood change the way you walk, if you know where to look!
On the other hand, these multiple ways to examine gait also highlight a critical fact: whatever the promises, neither a unique system nor a single parameter can address the globality of the gait. With today’s technology, there is no way to get a full picture and using any analysis in isolation creates the risk of assessing a patient from a single, incomplete perspective. For example, for individuals diagnosed with hamstring triceps spasticity and/or bones misalignment, treatments can contribute to changes in the joint motion making it closer to healthy references. However, what is the consequence on the dynamic stability for a person whose locomotion was built on these so-called defects, especially in the first months after surgery? In another example, gait speed is one of the most commonly used parameters and is even suggested as a sixth vital sign of health. However, as the product of step length and cadence, it is possible to produce the same speed by multiple configurations ranging from quick, small steps to slow, long steps. Walking with higher speed than 1 m/s (value often used as a reference) does not mean that the configuration between step length and cadence is optimal. If walking at this pace causes the patient to produce multiple small steps at a fast rate, they could be at higher risk for falls.
The motor control, or how a movement is produced and regulated, is an essential component of gait interpretation; unfortunately, it is often ignored. While it is true that healthy gait is a hallmark, normative references are obviously needed. However, walking with outcomes closer to those seen in control populations does not necessarily imply a better performance. It must be kept in mind that what a patient does can be related to the disease as well as being the product of coping adaptations, all according to physiological possibilities and limitations.
Who’s In Control?
Different Approaches of Motor Control Lead to Different Interpretations of Spatiotemporal Parameters and Variability. In a classical view, the central nervous system (CNS) controls all movements of the body. Coordinated movement patterns, such as walking, require motor planning and constant feedback from the surroundings to regulate the movement by a retroactive control. Consequently, any disorder could affect the production and the regulation of the movement planned by the CNS. Seen under this perspective, increased difference from “normal” walking parameters (for example, a slow walking speed) is seen as a linearly worsening clinical status. Therefore, the invariance of the movement is interpreted as the natural order of the biological system and the variability (for example, variation in the step-to-step values for step length, stride width or step time) only seen as a reflection of a physiological inability to produce the consistent movement as prescribed by the CNS. Clinically, gait variability is often only interpreted as a “noise” which must be removed, disregarding its potential for regulation.
Another approach considers further the possibility for a self-organization of the movement, to respond to the constraints as much as possible. Outside periods when a direct cognitive intervention is required to take control of the movement (eg, to choose which direction to turn at next street corner), it is postulated that the walking pattern emerges spontaneously from constraints related either to the task (eg, walking under time pressure), the organism (eg, pathological limitations), and/or the environment (eg, walking within a crowd). This implies that the way a person walks and the variability represent much more than the result of physical inabilities. Clinical assessment of gait is standardly administered in straight, unobstructed walking pathway, at self-pace. In this situation, there is no reason to see excessive variability in the gait parameters if there is not an underlying issue. However, if gait variability underlines the existence of an issue, it also demonstrates that the person can regulate the perturbations to walk. Another person walking without variability could be at higher risk for falls than the first patient when faced with a more complex situation, as the low variability could be hiding an inability to adapt.
Overall, an asymptomatic system must demonstrate control that allows it to be both robust (preserve the homeostasis) and flexible (ability to vary in order to adapt to the constraints).
A method to look at organization and variability
In a recent pilot study, Gouelle et al1 built on these concepts to introduce a scoring methodology based on the walking speed and the relationships between step length and cadence. One method, named the Organization Score, is proposed for looking at the mean gait characteristics and to quantify how much a patient’s gait pattern deviates from healthy references. This distance is not seen ultimately as a bad thing, instead it is thought to reflect the biomechanical adaptation and, indirectly, the residual capacity of adaptation which could be mobilizable for the management of internal or external perturbations. Therefore, a second component, the Variability Score, is needed to quantify the level of variability. By applying this method to a cohort of patients with Friedreich ataxia who walked independently, with one cane, or with a rollator, the authors demonstrated that equal levels of organization do not imply a same level of variability and vice-versa. Overall, the Organization Scores demonstrated a longitudinal deterioration in the gait characteristics from independent ambulators to those who ambulated with a rollator. Variability Scores mostly reflected dynamic instability, which became greater as the requirement of an ambulation aid or the switch from a cane to a rollator was imminent. In order to consider both components and the requirement to use an assistive device, the authors finally introduce the Global Ambulation Score (GAS) as a way to summarize the whole performance as a single number. The GAS might be a valuable outcome measure for longitudinal disease progression as it showed statistical correlation with the clinical status assessed by a standard notation scale for gait ability in Friedreich ataxia.
Gait analysis can be performed using numerous methods that vary depending on purpose for the patient and the background of the clinician or researcher administering the analysis. Due to gait’s complexity, it is essential that appropriate analysis tools, protocols, and outcome variables are used, as there is no global test or factor to answer all questions surrounding gait function. The most important process is to define the clinical question that you want to address. Always keep in mind that gait analysis is highly contextual, don’t be too quick to extrapolate because the “same” results can have different interpretations in different individuals/under different circumstances. Under standardized conditions in a gait laboratory, the variability mainly reflects disturbances. But, in real-world assessments, with wearable sensors for example, if one is navigating in a crowd, an appropriate ability to regulate gait is necessary for safety.
Arnaud Gouelle, PhD, is Director of Research at ProtoKinetics. A Gait Disorder Specialist, he previously served as head of the Gait & Motion Analysis Unit at a pediatric hospital in France. His publications focus mainly on stability and motor variability in functional movements.
Patrick Roscher, MS, is Chief Technical Support Engineer for ProtoKinetics. His background includes graduate education in biomedical engineering with a concentration in human biomechanics and working in the clinical gait laboratory at the former Rehabilitation Institute of Chicago.
This recurring column in LER will seek to present methods, protocols, and outcomes that are appropriate for answering a variety of questions that surround the analysis of gait. LER is proud to partner with the Gait and Balance Academy of ProtoKinetics to provide this technical information. Questions are welcome; please direct them to email@example.com.
- Gouelle A, Norman S, Sharot B, Salabarria S, Subramony S, Corti M. Gauging Gait Disorders with a Method Inspired by Motor Control Theories: A Pilot Study in Friedreich’s Ataxia. Sensors (Basel). 2021;21(4):1144.