Elsevier

Gait & Posture

Volume 43, January 2016, Pages 192-197
Gait & Posture

Reliability and minimal detectable difference in multisegment foot kinematics during shod walking and running

https://doi.org/10.1016/j.gaitpost.2015.09.022Get rights and content

Highlights

  • The Oxford multisegment foot model is used in both walking and running studies.

  • Reliability of the marker set during shod gait had not been reported.

  • Minimum detectable change in joint angles between days and within session was found.

  • Excursion and range of motion values tended to be more reliable than peaks.

  • Findings will aid interpretation of longitudinal and single session gait studies.

Abstract

There has been increased interest recently in measuring kinematics within the foot during gait. While several multisegment foot models have appeared in the literature, the Oxford foot model has been used frequently for both walking and running. Several studies have reported the reliability for the Oxford foot model, but most studies to date have reported reliability for barefoot walking. The purpose of this study was to determine between-day (intra-rater) and within-session (inter-trial) reliability of the modified Oxford foot model during shod walking and running and calculate minimum detectable difference for common variables of interest. Healthy adult male runners participated. Participants ran and walked in the gait laboratory for five trials of each. Three-dimensional gait analysis was conducted and foot and ankle joint angle time series data were calculated. Participants returned for a second gait analysis at least 5 days later. Intraclass correlation coefficients and minimum detectable difference were determined for walking and for running, to indicate both within-session and between-day reliability. Overall, relative variables were more reliable than absolute variables, and within-session reliability was greater than between-day reliability. Between-day intraclass correlation coefficients were comparable to those reported previously for adults walking barefoot. It is an extension in the use of the Oxford foot model to incorporate wearing a shoe while maintaining marker placement directly on the skin for each segment. These reliability data for walking and running will aid in the determination of meaningful differences in studies which use this model during shod gait.

Introduction

There has been increased interest recently in measuring kinematics within the foot during gait. While several multisegment foot models have appeared in the literature, one in particular has been used frequently for both walking and running. The two segment plus hallux Oxford foot model was originally developed for use during barefoot walking [1]. This model was subsequently modified, which improved reliability of rearfoot motion when used with children walking barefoot [2]. This model has since been used to measure multisegment foot kinematics during both walking and running. Given this model's wide adoption in the field, this study will focus on reliability and minimum detectable difference (MDD) of the modified Oxford foot model.

Reliability of a marker set is of particular concern in longitudinal studies where changes must exceed the MDD to be considered meaningful [3]. Low reliability leads to large MDDs. Several studies have reported the reliability of the Oxford foot model or modified Oxford foot model. However, MDDs have not been reported. There is a need to provide MDDs for commonly studied variables in both walking and running to aid the interpretation of longitudinal studies. In particular, studies of different foot strike patterns during running have stimulated interest in kinematics within the foot. Similarly, investigations of walking in people with osteoarthritis or other pathologies indicate a need to track walking biomechanics longitudinally.

It is considered good practice in marker-based studies for a single investigator to place markers on all participants at all visits. This reduces the between subject variability by removing the influence of inter-rater differences. In interpreting findings for between-day reliability, it is important to note that there is a minimum MDD that reflects the inherent variability of human subjects. This within-session or inter-trial reliability is a floor in terms of the smallest MDD that can be achieved. Thus, within-session MDDs are useful for interpreting the effects of different conditions within a session, such as changing velocity or stride length. They are also useful in interpreting group differences in cross-sectional studies.

While several studies have reported the reliability for the Oxford foot model, most have reported reliability for barefoot walking [4], [5], [6], [7]. One study reported reliability for a multisegment foot using shoe mounted markers [8]. However, this model tracks shoe rather than foot motion and cannot provide insight into foot movement as foot and shoe likely move differently [9]. Thus, it is unknown whether similar reliability would be found for a model using skin-mounted markers during shod walking or for running. Since excursions and peak magnitudes of joint angles are larger in running than walking, MDDs should be reported separately for each. The purpose of this study was to determine between-day (intra-rater) and within-session (inter-trial) reliability of the modified Oxford foot model during shod walking and running. The focus was on variables which are commonly reported in the gait literature. In addition, MDDs were determined to aid the interpretation of the findings of gait studies.

Section snippets

Participants

Healthy male runners between 18 and 50 years of age were recruited for this study. All procedures were approved by the Institutional Review Board prior to the start of the study. All participants provided written informed consent to participate. Participants ran at least 10 miles per week and had been doing so for a year or more. Participants who answered ‘yes’ to any of the questions on the Physical Activity Readiness Questionnaire [10] were excluded. Those reporting a history of major lower

Results

The between-day reliability of multisegment foot variables during running ranged from an ICC of 0.644 to 0.993 (Table 1). Seven of 20 variables had excellent between-day reliability, and 5 more had good between-day reliability. The relative measures, excursion and range of motion, tended to have higher reliability than the absolute measures of angle at foot strike and peak angle. The between-day reliability of the remaining six variables could not be reported, as the assumptions for ICC

Discussion

The purpose of this study was to determine the reliability of the modified Oxford foot model during shod gait. Reliability during walking and running between days was determined using ICCs, and the MDD for absolute and relative variables of interest in the foot and ankle was calculated. Between-day reliability depends on both participant consistency from day to day and investigator consistency in placing markers on anatomical landmarks. Thus, within-session reliability and MDD were also

Conclusion

The reliability of the modified Oxford foot model was determined for shod walking and running in healthy adult men. Overall, between-day ICCs were comparable to those reported previously for adults walking barefoot. Our study is an extension in the use of the model to incorporate wearing a shoe while maintaining marker placement directly on the skin for each segment. Furthermore, these reliability data for walking and running will aid in the determination of meaningful differences in studies

Conflict of interest

The authors report no conflicts of interest.

Cited by (16)

  • Multisegment Foot Models

    2022, Foot and Ankle Biomechanics
  • Oxford foot model kinematics in landings: A comparison between professional dancers and non-dancers

    2020, Journal of Science and Medicine in Sport
    Citation Excerpt :

    The Oxford Foot Model provides kinematic data of a three-segment foot (hindfoot, forefoot and hallux).6,7 Several studies6,7,17 have reported the reliability of this model in adults and children, healthy and pathological populations, in different tasks (e.g. walking, running). However, there is meager scientific information related to the multi-segmented foot in dancers during single-leg multidirectional jump-landings.

  • Can foot orthoses impose different gait features based on geometrical design in healthy subjects? A systematic review and meta-analysis

    2020, Foot
    Citation Excerpt :

    The quality of walking condition was determined based on four parameters: walking with shoes (1) or barefoot (0); walking along walkway (1) or on treadmill (0); the number of cycles with more than or equal to 5 cycles (1), otherwise (0); monitoring and controlling the speed of walking (1), otherwise (0). Those experimental aspects are known to affect the quality of outcome measures [34–39] and also the gait pattern [40–43]. The score was expressed in percentage and the biomechanical quality was classified similar to the methodological quality.

  • Multisegmented ankle-foot kinematics during gait initiation in ankle sprains and chronic ankle instability

    2019, Clinical Biomechanics
    Citation Excerpt :

    The level of significance was p ≤ .05 for all analyses. A priori sample size estimation of 19 participants were needed to demonstrate large effects based on variance of multisegmented rearfoot eversion, an α=0.05, and β=0.20 (Milner and Brindle, 2016). Group mean joint angle curves and standard deviations were calculated and assessed for differences using an omnibus Statistical Parametric Mapping (SPM) one-way analysis of variance SPMANOVA (Pataky, 2012).

  • Effect of foot orthosis design on lower limb joint kinematics and kinetics during walking in flexible pes planovalgus: A systematic review and meta-analysis

    2018, Clinical Biomechanics
    Citation Excerpt :

    Except question 5, scored as 0 (“barefoot”), 1 (“sandals”), 2 (“shoes”), all the questions were scored as 1 (“yes”) or 0 (“no”). These experimental aspects were assessed because they are shown to affect lower limb biomechanics (Menant et al., 2009; Morio et al., 2009; Pothrat et al., 2015; Shultz et al., 2011; Stoquart et al., 2008; Xu et al., 2017; Zhang et al., 2013) and the quality of biomechanical parameters (Almarwani et al., 2016; Geerse et al., 2017; Laroche et al., 2011; Milner and Brindle, 2016; Monnet et al., 2010; Riva et al., 2014). As previously discussed for the methodological quality, scores were expressed in percentage of the maximum score (8) and studies were classified as having high, moderate or low biomechanical quality.

View all citing articles on Scopus
View full text