Elsevier

Gait & Posture

Volume 50, October 2016, Pages 126-130
Gait & Posture

Full length article
Accelerometer-based determination of gait variability in older adults with knee osteoarthritis

https://doi.org/10.1016/j.gaitpost.2016.08.024Get rights and content

Highlights

  • Accelerometry provides an effective tool to analyze the temporal aspects of gait variability.

  • Knee osteoarthritis may increase gait variability.

  • Gait variability is inversely related to walking speed.

Abstract

Knee osteoarthritis (KOA) can affect the spatiotemporal (ST) aspects of gait as well as the variability of select ST parameters based on standard linear measures of variability (e.g., standard deviation (SD) and coefficient of variation). Non-linear measures (e.g., fractal scaling index (FSI) and sample entropy) can be more sensitive to changes in gait variability, and have been used to quantify differences in the stride patterns of patients with Parkinson’s disease and the motion of ACL-deficient knees. However, the effect of KOA on the dynamic complexity of the stride pattern has not been investigated. Therefore, the purpose of this study was to investigate the effect of KOA on gait variability (linear and non-linear measures) in a group of older adults, and to compare these results to a healthy control group. Participants walked for 10 min with a tri-axial accelerometer placed at the lower back. Mean and SDs of stride time and step time as well as the FSI for the entire series of stride times were calculated for each participant. Participants with KOA had significantly greater mean stride time (p = 0.031) and step time (p = 0.024) than control group participants. While stride and step time variability (SD) were greater in the KOA group, the differences were not significant, nor was the difference in the FSI. Low statistical power (β = 0.40 and 0.30 for stride and step time SD, respectively) combined with the confounding effects of walking speed and heterogeneous KOA severity likely prevented significant differences from being found.

Introduction

As individuals age, changes in gait patterns become apparent, and are often influenced by the effects of aging itself. However, knee osteoarthritis (KOA) can have an even more pronounced effect on gait alterations with a subsequent negative impact on an older adult’s level of independent living. The degenerative effects of KOA have been shown to impact several of the age-related changes in the mean spatiotemporal parameters of gait (e.g., shorter stride lengths, longer stance phases, reduced speed, longer stride time and increased double support time) [1], [2], [3], [4], which may reflect compensation strategies to minimize joint pain and protect the knee [4].

These compensations can also affect the variability of gait [5], [6], [7], and recently the subject of gait variability (i.e., the magnitude of stride-to-stride fluctuations and their changes with respect to time) has become a more thoroughly investigated aspect of gait analysis. It provides important information about the rhythmic pattern of the gait cycle that is overlooked when more traditional measures such as the average speed, mean stride time, and mean step length are determined [8]. In non-linear biological systems, variability is related to complexity, wherein a certain amount of variability in the gait cycle is considered to be healthy and represents adaptability and efficient gait control in response to unstable environmental conditions [9]. However, as people age and/or experience different pathologies, the behavioural complexity of the neuromusculoskeletal system can change. Too much or too little variability can affect the stability of the system, as has been shown by studies that have found increased values for the standard deviation (SD) of stride time [10], [11].

Normal gait cycles also possess fractal properties that can be quantified using the fractal-scaling index (FSI), which provides a measure of the long-range, self-similar stride patterns that are associated with gait stability8. Studies show that the FSI can discriminate between healthy gait patterns and those affected by disease and aging, such that lower values (i.e., more random uncorrelated patterns of strides) are associated with impaired health [8], [12], [13]. More random gait patterns are often associated with rhythmic impairments in the neuromuscular control system, but the literature also indicates that increased randomness (less complexity) and gait instability are also associated with joint pathology [5], [6], [7]. This suggests that patients with KOA may express decreased complexity as demonstrated by lower FSI values. Although it is evident that KOA can negatively affect the basic parameters of gait compared to healthy individuals, there has been limited research on the effects of KOA on gait variability. Kiss [14] and Kiss et al. [15] have shown a relationship between KOA and increased gait variability for parameters such as step length and double-support phase duration. However, it has been suggested that different gait variability parameters represent different constructs [16], and as such additional research is needed to fully understand the gait variability characteristics that are associated with KOA. Determining the effects of KOA on stride pattern variability and the fractal structure of the gait cycle will provide a better understanding of the extent to which changes in complexity are associated with the degenerative nature of KOA. To our knowledge, the effect of KOA on stride time variability and the fractal dynamics of the gait cycle has not previously been investigated.

Therefore, the purpose of this study was to investigate the effect of KOA on gait variability (linear and non-linear measures) in a group of older adults, and to compare these results to a healthy, age- and sex-matched control group. It was hypothesized that gait variability (i.e., stride time SD and step time SD) would be greater and the pattern of strides would be more random/uncorrelated (i.e., lower stride time FSI) in the KOA group compared to the healthy older adult control group.

Section snippets

Participants

Thirty older adults (KOA: n = 15; Control: n = 15) matched for age and sex participated in the study (see Table 1). Test participants were admitted to the study if they received a medical diagnosis of KOA, were ≥55 years of age, and could walk for ten minutes without the use of an assistive device and without pain. All participants in the KOA group had bilateral KOA. Based on radiographic assessments of the participants’ knees, the Kellgren-Lawrence (K-L) grading scale was used to categorize the

Results

Demographics, KOOS score, and gait speed results for both groups are shown in Table 1. The two groups were matched according to age as closely as possible, and the results show that the small difference was not significant. KOA group participants had a significantly greater mass, BMI, and KOOS score, and walked at a significantly slower self-selected walking speed than control group participants.

There was an average of 1081 ± 76.07 and 1030.8 ± 65.48 steps taken by the control group and KOA group,

Discussion

This study investigated gait variability in older adults with and without KOA and we hypothesized that gait variability would be significantly different between age and sex-matched participants in the KOA group and control group. The results showed that the participants with KOA walked significantly slower than the healthy control group participants, which is consistent with findings in the literature [1], [2], [3], [26]. The KOA group also had a significantly greater mean stride time and mean

Conclusion

The complex gait characteristics of patients with KOA can make it difficult for clinicians to administer meaningful, objective assessments of their KOA patients’ mobility patterns. This is the first study that has investigated the effect of KOA on certain temporal aspects (i.e., stride time and step time) of the variability and fractal dynamics of the gait pattern. Although the differences in gait variability for this study were not statistically significant, KOA was associated with increased

Acknowledgements

We would like to acknowledge Dr. Megan Dash, Dr. Jeremy Reed, the Centre on Aging and Health, and the University of Regina Fitness and Lifestyle Centre for their assistance with this project.

References (33)

  • R.M. Kiss

    Effect of walking speed and severity of hip osteoarthritis on gait variability

    J. Electromyogr. Kinesiol.

    (2010)
  • T.M. Owings et al.

    Measuring step kinematic variability on an instrumented treadmill: how many steps are enough?

    J. Biomech.

    (2003)
  • J.A. Zeni et al.

    Knee osteoarthritis affects the distribution of joint moments during gait

    Knee

    (2011)
  • S. Lord et al.

    Gait variability in older adults: a structured review of testing protocol and clinimetric properties

    Gait Posture

    (2011)
  • K.K. Patterson et al.

    Evaluation of gait symmetry after stroke: a comparison of current methods and recommendations for standardization

    Gait Posture

    (2010)
  • K.S. Al-Zahrani et al.

    A study of the gait characteristics of patients with chronic osteoarthritis of the knee

    Disabil. Rehabil.

    (2004)
  • Cited by (38)

    • Objective gait assessment in individuals with knee osteoarthritis using inertial sensors: A systematic review and meta-analysis

      2022, Gait and Posture
      Citation Excerpt :

      The length/duration of these walking trials was 500 m [17] in one study and 11 min in the other study [38]. Two studies used a 200 m oval track that subjects were asked to walk for a duration of 9 min at a self-selected speed [20,29]. In another study subjects walked continuously for 10 min [22].

    View all citing articles on Scopus
    View full text