Full length articleInter-session agreement and reliability of the Global Gait Asymmetry index in healthy adults
Introduction
Gait symmetry is generally defined as the identical behaviour of the left and right limbs during gait [1]. Pronounced asymmetry levels have been associated with pathological conditions such as stroke [2], lower limb amputations [3], osteoarthritis [4] and arthroplasties [5], as well as anterior cruciate ligament injury [6]. Consequently, there has been a growing effort in restoring/increasing gait symmetry in several clinical conditions [7] back to those found in asymptomatic individuals.
Researchers have been studying the efficacy of gait retraining programs in improving gait symmetry [7]. A prerequisite to restore gait symmetry is the ability of the chosen approach to impose changes that exceed the between-session variability [8]. Hence, knowledge of the measurement error and reliability of symmetry indices is required. However, to the best of the authors’ knowledge only two studies [8], [9] have analysed the repeatability (i.e. agreement and reliability under identical conditions [10]) of symmetry indices. Lewek and Randall [8] analysed the repeatability of the Symmetry Ratio in stance time, swing time and step length for stroke patients. Accordingly, symmetry was calculated by dividing each of these parameters from the paretic limb by the corresponding parameter in the non-paretic limb (i.e. a step length ratio of 1 indicates perfect symmetry; whereas a ratio of 0.5 means that the step length from the paretic limb is half the step length from the non-paretic limb). These authors reported inter-session Intraclass Correlation Coefficient (ICC) (2,1) values between 0.925 and 0.976, and minimal detectable change (MDC) percentages between 8% and 18%. One of the limitations in this study is the wide range of days between the two visits (between 3 and 36 days). Secondly, although the ICC and MDC values were very good, the calculation of the standard error of measurement (SEM), and consequently the MDC are questionable. The SEM cannot be calculated using a model of ICC that includes systematic error as this is not reflected in the pooled standard deviation. Moreover, the calculation of SEM from the ICC is not recommended as it highly depends on the sample’s heterogeneity [11]. Senden et al. [9] examined the repeatability of step time asymmetry in healthy individuals, where step time asymmetry was the difference between the durations of successive left and right steps, divided by the mean duration between sides. These authors reported intra-session variation coefficients and ICC values from 29.25% to 47.88% and 0.509 to 0.787, respectively, and inter-session/observer ICC values between 0.010 and 0.351. These results suggest very poor inter-session reliability. Furthermore, the authors did not specify which ICC model was used, nor the measurement error, which are essential to interpret the ICC reported [12]. The interpretation of ICC values is also incomplete without its confidence intervals, which are lacking in both studies.
Besides these limitations, these studies assessed the repeatability of local symmetry indices. This type of indices quantify symmetry based on a single discrete metric (i.e. vertical ground reaction force, step length or knee sagittal angle, among others) that only relates to a portion of the gait cycle (i.e. peak value), therefore neglecting the temporal information of gait waveforms [13] as well as compensation strategies that affect other local parameters [14]. Consequently, some patients may be considered symmetric based on their step length symmetry, for example, even though their overall gait is highly asymmetric [15]. Alternatively, several local scores could be used, but it would be much harder for the clinician to assess the effectiveness of the intervention this way [16]. To overcome these limitations, global symmetry indices have been developed [16], [17], [18]. Contrarily to local indices, global symmetry indices can reduce the information from various continuous gait waveforms (for example, the joint angles from the entire lower limbs, at each percentage of the gait cycle) into a single numeric score. Hoerzer et al. [16] developed a global symmetry index which uses information from the three components of the ground reaction force, as well as the hip, knee and ankle joint angles, moments and velocities in all three planes of motion. Instead, the global symmetry index developed by Nigg et [17] includes data from the vertical and anterior/posterior components of the ground reaction force as well as the angular positions and velocities at the hip (in all three planes of motion), the knee (sagittal plane only) and ankle (sagittal and frontal planes). However, these two indices only include information from stance phase, therefore neglecting toe clearance and the preparation for the heel strike. Additionally, the combination of quantities with different physical units requires normalization, which may mask or artificially inflate the level of symmetry [19]. The AsymGPS, a global symmetry index proposed by Lundh et al. [18], overcomes these limitations by including data from the entire gait cycle and by limiting the analysis to the lower limb joint angles. However, this index does not include certain angles in the frontal and transverse planes which can be valuable to clinicians. Hence, the Global Gait Asymmetry (GGA) index was recently developed to include all three components of the joint angles at the lower limbs and trunk throughout the gait cycle.
Global symmetry indices may be more appropriate for gait retraining purposes as they provide information on overall gait symmetry in a single score, thus facilitating the assessment of a patient’s progress and therefore of the effectiveness of a chosen intervention [16]. However, their repeatability has never been studied. The aim of the present study was to assess the inter-session agreement and reliability of the GGA [19] in healthy adults.
Section snippets
Study design
This study followed a prospective test-retest (inter-session, within examiner) design.
Participants
The data used in this study was collected from the sample used in Fernandes et al. [20]. Sample size was calculated for a predefined 5% level of significance with 80% power, using the formula of Kraemer and Thiemann [21]. For a desired reliability coefficient of 0.90 with a minimum reliability of 0.70, a sample of 17 participants was required. Thus, to allow for non-attenders and increased precision 23
Results
The participants’ walking speed remained consistent between visits (0.71 ± 0.08 vs 0.71 ± 0.09 statures/s; t(22) = 0.706, P = 0.488). The GGA scores obtained in the first and second visits were 619 ± 177 (intra-individual SD: 6–79) and 605 ± 119 (intra-individual SD: 13 to 94), respectively. The mean difference in GGA (1st visit minus 2nd visit) was 14 ± 194 (95% CI: −70 to 98), with 95% LOA of −366 (−60%) and 394 (64%). As seen in the Bland and Altman Plot (Fig. 2), the standard deviation of these
Discussion
This is the first study to assess the repeatability of a global symmetry index in healthy individuals. The present results show that the GGA has very poor agreement and reliability between sessions.
The presence of outliers, in terms of the differences between visits, may have contributed to the poor repeatability of this index. However, the time between visits was very similar among participants. Moreover, the participants maintained their healthy condition and had both gait analyses performed
Conclusion
The values obtained for reliability and measurement error prevent the application of the GGA as a measure of symmetry for monitoring progress, both in a research and clinical context. These results highlight the importance of testing the repeatability of other global symmetry indices before being used to assess the outcome of gait interventions. In addition, this study suggests that global indices composed of kinematics that are less sensitive to marker placement precision and soft tissue
Conflict of interest statement
W. Scott Selbie is the CEO and Director of Research at C-Motion, the company which commercialises Visual 3D. The other authors have no conflicts of interest to declare.
Acknowledgements
This study was funded by the Portuguese Foundation for Science and Technology and CIPER (Research Project: PEst-OE/SAU/UI447/2014; PHD Grant: SFRH/BD/69424/2010).
References (30)
- et al.
Symmetry and limb dominance in able-bodied gait: a review
Gait Posture
(2000) - et al.
Step length asymmetry is representative of compensatory mechanisms used in post-stroke hemiparetic walking
Gait Posture
(2011) - et al.
Between-limb kinematic asymmetry during gait in unilateral and bilateral mild to moderate knee osteoarthritis
Arch. Phys. Med. Rehabil.
(2013) - et al.
Asymmetric hip kinematics during gait in patients with unilateral total hip arthroplasty: in vivo 3-dimensional motion analysis
J. Biomech.
(2015) - et al.
Altering asymmetric limb loading after hip arthroplasty using real-time dynamic feedback when walking
Arch. Phys. Med. Rehabil.
(2005) - et al.
Acceleration-based gait test for healthy subjects: reliability and reference data
Gait Posture
(2009) - et al.
When to use agreement versus reliability measures
J. Clin. Epidemiol.
(2006) - et al.
Biomechanical features of gait waveform data associated with knee osteoarthritis: an application of principal component analysis
Gait Posture
(2007) - et al.
Effects of unloading bracing on knee and hip joints for patients with medial compartment knee osteoarthritis
Clin. Biomech.
(2011) - et al.
Evaluating asymmetry in prosthetic gait with step-length asymmetry alone is flawed
Gait Posture
(2012)
Development of a symmetry index using discrete variables
Gait Posture
Three dimensional multi-segmental trunk kinematics and kinetics during gait: test-retest reliability and minimal detectable change
Gait Posture
Position and orientation in space of bones during movement: anatomical frame definition and determination
Clin. Biomech.
Human movement analysis using stereophotogrammetry Part 4: Assessment of anatomical landmark misplacement and its effects on joint kinematics
Gait Posture
Spinal, pelvic, and hip movement asymmetries in people with lower-limb amputation: systematic review
J. Rehabil. Res. Dev.
Cited by (13)
Overnight unilateral withdrawal of thalamic deep brain stimulation to identify reversibility of gait disturbances
2022, Experimental NeurologyCitation Excerpt :Spatial and temporal parameters obtained from Mobility Lab software were: stride length, stride velocity (corresponding to gait speed), stride time, cadence, double support, swing and stance time and ranges of motion (RoM) of knees, shanks and trunk. Parameters reflecting asymmetry and peak velocities, that are reported to have a low reliability and agreement between different measurements(Cabral et al., 2017) were not included in further analysis. Coefficients of Variation (CoV) were calculated for gait parameters in all trials as standard deviation divided by the respective mean.
Differences in body positional bilateral symmetry between stance and supine positions, and the impact of attention and awareness on postural symmetry
2019, Gait and PostureCitation Excerpt :Structural and functional asymmetry may cause pathological posture and impaired movement pattern [3–5]. Although it is believed that the ideal anterior postural alignment in healthy adults is characterized by bilateral symmetry [2], studies have shown small postural asymmetries and asymmetrical movement patterns on healthy adults [6,7]. These asymmetries increase among subjects with Adolescent Idiopathic Scoliosis (AIS) [8], Cerebral Palsy [9], and Stroke [3].
Comparison of plugin and redundant marker sets to analyze gait kinematics between different populations
2023, BioMedical Engineering OnlineGait characteristics related to fall risk in patients with cerebral small vessel disease
2023, Frontiers in NeurologyReliability of a Global Gait Symmetry Index Based on Linear Joint Displacements
2022, Applied Sciences (Switzerland)Gait symmetry-A valid parameter for pre and post planning for total knee arthroplasty
2022, Journal of Musculoskeletal Neuronal Interactions