Reliability of centre of pressure, plantar pressure, and plantar-flexion isometric strength measures: A systematic review

Gait & Posture 75 (2020) 46–62

Contents lists available at ScienceDirect

Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost

Review

Reliability of centre of pressure, plantar pressure, and plantar-flexion isometric strength measures: A systematic review

T

Kim Hébert-Losier , Lauralee Murray ⁎

University of Waikato, Division of Health, Engineering, Computing and Science, School of Health, Adams Centre for High Performance, 52 Miro Street, Mount Maunganui, 3116, Tauranga, New Zealand

ARTICLE INFO

ABSTRACT

Keywords: Balance Centre of pressure Plantar pressure Psychometric properties Triceps surae

Background: Centre of pressure (COP), plantar pressure (PP), and plantar-flexion isometric strength (PFisom) are often examined in relation to postural control and gait. Research question: Our aim was to systematically review and quality appraise articles addressing the reliability of COP and PP measures in static stance and PFisom measures. Methods: Three electronic databases (SCOPUS®, SportDISCUS™, and PubMed) were searched and supplemented by a manual search. Peer-reviewed original research on the reliability of COP, PP, and PFisom in healthy adults (≥18 years) was included. Quality appraisal was done according to the updated COnsensus-based Standards for the selection of health Measurement INstruments reliability checklist. Data regarding study characteristics, test protocols, outcome measures, and reliability metrics were extracted. Results: Forty articles met inclusion and were assessed for their methodological quality. Only four articles (10%) obtained uppermost quality scores. From the reviewed studies, the most reliable measures were: COP sway area and path length; PP mean pressure, percentage body weight distribution, and contact area; and PFisom peak torque and force. Although these measures generally exhibited good-to-excellent relative reliability based on correlation coefficients, absolute reliability based on typical errors were not always optimal (variation > 10%). Literature on PP reliability was scarce (n = 2). Significance: Our findings highlight the need for better quality methodological reliability studies to be undertaken to make stronger inferences about the reliability of COP, PP, and PFisom measures. The most reliable measures based on the current review are: COP sway area and path length; PP mean pressure, percentage of body weight distribution, and contact area; and PFisom peak torque and peak force. These measures are the ones that should be selected preferentially in clinical settings, bearing in mind that their typical errors might be suboptimal despite exhibiting strong relative reliability.

1. Introduction Postural control regulates our ability to maintain upright stance and is necessary to perform activities of daily living [26]. As such, clinicians and scientists often examine postural control. Tracking centre of pressure (COP) measures from force plates in quiet stance is standard in examining postural control in both healthy and patient populations [41,63]. Previous studies indicate that force plate COP measures from balance tests are useful in predicting falls in community-dwelling older adults presenting in good general health [56] and are altered after the completion of an ultra-marathon race in runners [19]. These findings indicate that COP measures have the potential to identify alterations in postural control in cohorts of adults identified as being in good general health. ⁎

Pedobarography is the study of pressure between the plantar surface of the foot and supporting surface. Pedobarography has been used to investigate foot-ground interaction, foot posture [27], and screen for possible pressure abnormalities in healthy individuals that are likely to lead to pathologies [6]. Plantar pressure (PP) distribution in static stance has been associated with pain and pathological profiles in older adults [27] and runners [13], as well as obesity status in children [20]. The plantar-flexor muscles are recognised to play a critical role in posture regulation [30] and locomotion [55]. With plantar-flexor muscle fatigue, postural control has been shown to worsen and the relative contribution of the vestibular, somatosensory, and proprioceptive systems to postural control are reported to change [34]. Handheld [16,68] and isokinetic dynamometers are the main tools used in practice and research to measure the strength of the plantar-flexors

Corresponding author. E-mail addresses: [email protected] (K. Hébert-Losier), [email protected] (L. Murray).

https://doi.org/10.1016/j.gaitpost.2019.09.027 Received 27 May 2019; Received in revised form 20 September 2019; Accepted 24 September 2019 0966-6362/ © 2019 Elsevier B.V. All rights reserved.

Gait & Posture 75 (2020) 46–62

K. Hébert-Losier and L. Murray

objectively, with isometric tests useful for assessing neuromuscular function [25]. Given that COP, PP, and isometric plantar-flexion strength (PFisom) are commonly used in clinical and scientific enquiry of postural control, it is vital to understand the reliability of these measures. For example, knowledge on reliability is important before interpreting changes in measures and using them to inform practice. The aims of this combined systematic review of the literature was to critically appraise and summarise research that has investigated the reliability of COP measures in static stance from force plates, PP distribution measures in static stance from plantar pressure mats, and PFisom measures from dynamometers in cohorts of healthy participants.

3 Include participants 18 years or over; 4 Be an original research article published in a peer-reviewed journal; 5 Be available in the English language. Articles that solely reported the reliability of measures in a patient population (i.e., individuals with pathologies or injuries) were excluded to narrow the scope of the review and considering that health status can influence the reliability of measures [29]. Reliability data reported in symposium reports or conference abstracts were not considered. For the three independent searches, articles were excluded if: 1 COP: static stance condition was not examined (i.e., only dynamic conditions were assessed) or when force plates were not used. 2 PP: static stance condition was not examined (i.e., only dynamic conditions were assessed) or when in-sole plantar pressure sensors were used rather than plantar pressure mats. 3 PFisom: isometric condition was not examined (e.g., only isokinetic or isotonic conditions were assessed).

2. Methods 2.1. Systematic search This combined systematic review adhered to the structures and reporting requirements of the PRISMA statement. Given that the reliability of measures were reviewed rather than health-related outcomes, this systematic review was not registered in the PROSPERO database. The SCOPUS®, SportDISCUS™, and PubMed electronic databases were systematically searched on the 6th of September 2017. Three independent searches were conducted to address the reliability of COP, PP, and PFisom measures (Fig. 1). No explicit limiters were applied in any of the electronic databases searched. The search syntaxes shown in Fig. 1 can be copied and pasted into the generic Scopus® (Article title, Abstract, Keywords), SPORTDiscus™ (EBSCOhost basic search platform), and PubMed search bars to reproduce the systematic search. To be included, articles needed to:

One reviewer conducted the database search (LM) and compiled all articles in a reference manager software (Endnote™, version X8, Clarivate Analytics, Philadelphia, PA, USA). Duplicate articles were removed before screening the titles, abstracts, and full-text articles in that order for inclusion and exclusion. Results from the screening process were verified by a second independent reviewer (KHL). The process was repeated for the reference list of all articles meeting inclusion until no additional article of relevance could be found (Fig. 1). 2.2. Quality assessment The updated COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) Risk of Bias Checklist (Box 6. Reliability) was used to assess the methodological quality of articles

1 Address test-retest or rater reliability; 2 Derive reliability metrics from a cohort of “healthy” individuals;

Fig. 1. Flow diagram of the article selection process of studies addressing the reliability of (A) centre of pressure, (B) plantar pressure, and (C) plantar-flexion isometric strength measures. 47

Gait & Posture 75 (2020) 46–62

K. Hébert-Losier and L. Murray

identified [51]. The 8-item COSMIN Box 6. Reliability checklist was chosen given that it is one of the few checklists available that targets reliability studies and suitable to assess test-retest, inter-rater, and intra-rater reliability studies [51]. The original COSMIN checklist has demonstrated excellent inter-rater agreement (percentage agreement: 94%; intraclass kappa: 0.77) [52]. An item was scored positively (√) when an article was allocated a “very good” or “adequate” rating, and negatively (X) when allocated a “doubtful” or “inadequate” score. Items 5–7 were not applicable across studies, and hence not reported here. The number of positive scores across the remaining 5 items were summed to allocate a final quality score to articles (range: 0–5), with greater scores indicating articles of superior methodological quality according to COSMIN [51]. Worth noting is that the COSMIN Box 6. Reliability checklist awards a positive score for the statistical use of intraclass correlation coefficient (ICC) in presence of continuous measures [51], with Prescott (2019) advising against the use of ICC to interpret the reliability of measures. Two reviewers (LM and KHL) met to agree on how to score each COSMIN item before independently quality assessing the articles. Both reviewers then quality appraised all articles meeting inclusion (n = 40). After independent assessments, any inconsistencies in scoring were discussed until a consensus was achieved on the remaining items. A third reviewer was identified to reconcile differences in opinion, but was not needed.

3. Results 3.1. Literature search and quality appraisal A total of 40 articles were reviewed: 27 for COP, two for PP, and 11 for PFisom (Fig. 1). Quality assessment scores ranged from 0 to 5 (Table 1). Only 4 studies (n = 10%) achieved uppermost quality score ratings of 5 [14,43,57,67]. Failure to provide evidence that participants were stable in the interim period or failure to report key methodological information (e.g., rest intervals between trials, sampling frequencies of testing devices, order of experimental conditions, and rater information) were the main reasons for negative quality ratings. A summary of the extracted data from each study is presented in Table 2. 3.2. Centre of pressure The median quality score of the 27 COP reliability studies [2, 3,9,11,12,15,17,18,21,24,26,33,42–46,48,50,57,59,61,64–66,69,70] was 4 out of 5 (range 0–5). Two studies [43,57] achieved the uppermost quality rating of 5 out of 5. A total of 929 healthy participants were included across these 27 studies [mean sample size of 34 (41) individuals, range 7–220]. Weighted mean values of participant characteristics for the appraised studies were 42.7 (23.9) y, 168.5 (6.7) cm, and 66.3 (7.6) kg. The studies assessed the reliability of various COP related measures, with the most common being “sway area”, “velocity”, and “total path length”. Most of the articles were test-retest in design, with time between trials ranging from 2 min to 9 months. Of the 27 articles assessed, three articles [11,18,69] investigated both test-retest and rater reliability. Most often, studies examined four different balance tasks and three trials per task, with trials typically lasting 30 s. Reliability was poor [9,21,26,33,43,48,70] to excellent [2,17,18,26,42,45,46] for the various measures reported, with ICCs and CVs ranging from 0.06 to 0.97 and 17 to 28%. Relative reliability based on correlation measures was the smallest for sway area [48] and the greatest for path length [42] in an eyes-open condition. The absolute reliability of measures was generally suboptimal, as would indicate the relatively large CV values.

2.3. Data extraction Information concerning study aims, design, population, equipment, protocol, outcome measures, statistical analysis, and results were extracted from each article using a standardised template by one reviewer (LM). Completeness of data extraction was verified by a second reviewer (KHL). Study design was classified as test-retest (i.e., intra-session or inter-session) or rater (intra-rater or inter-rater) reliability. 2.4. Data analysis Data were managed and analysed using Microsoft Excel 2016 (Microsoft Corporation, Redmont, WA, USA). Descriptive statistics were expressed using means and standard deviations (SD), median and interquartile ranges (median [lower quartile, upper quartile]), mode, ranges (minimum to maximum), counts (n), or percentages (%) depending on data type. When possible, weighted means based on sample size were calculated for age, height, and mass of cohorts. Meta-analysis was not attempted due to the heterogeneity of outcome measures and reliability statistics used across studies. Studies quantify reliability using a range of statistical measures [1,36,37,58]. Both absolute and relative reliability metrics are important to clinicians and scientists [35,36]. Typical errors expressed either in raw units or coefficients of variations (CV) provide an indication of the absolute reliability of measures and the variation within each individual [35,36]. Absolute reliability is typically deemed acceptable when CV values are < 10% and suboptimal when ≥ 10% [1]. Correlation-based measures, such as intraclass (ICC) and Pearson correlation coefficients, indicate the relative reliability of measures and the reproducibility of the rank order of individuals on a retest [35,36]. Relative reliability of measures based on ICC are commonly interpreted as poor ICC < 0.40, fair 0.40 ≥ ICC < 0.75, good 0.75 ≥ ICC < 0.90, and excellent ICC ≥ 0.90 [62]. Even though the use of ICC as reliability measure is ‘standard’ practice [58] and supported by statistical experts [35,36], others argue against the use of this particular measure [58] and are preponderates of limits of agreements [1,58]. Despite the fact ICC and CV measures were selected to structure our results, other reliability metrics were extracted (e.g., limits of agreements and minimal detectable change values) to inform readers and favour an inclusive approach to our systematic review.

3.3. Plantar pressure The two PP reliability studies [6,38] were of relatively poor methodological quality (range 2–3). A total of 96 healthy participants were included. Weighted mean values of participant characteristics for these two studies were 39.8 (9.4) y, 167.4 (0.5) cm, and 70.2 (1.7) kg. Several measures were investigated within the two studies, including relative pressure-load bilaterally and unilaterally (% body weight), and mean and peak values for PP and contact area. Both studies had a test-retest design, with time between testing sessions ranging from 3 to 27 days. Static PP distribution was assessed using 5 trials of 30 s in duration in one study [6], whilst the other did not state the length of trials [38]. Overall, inter-session relative reliability was greater than intra-session, with good to excellent relative reliability reported for relative pressure (ICC: 0.95 to 0.97), mean pressure (ICC: 0.93 to 0.98), contact area (ICC: 0.93 to 0.97), and peak pressure (ICC: 0.86). Across both studies [6,38], intra-session relative reliability was slightly better in the second than the first testing session for all aforementioned outcome measures (ICC range: 0.92 to 0.99 versus 0.87 to 0.98), except for peak pressure where relative reliability was similar (ICC: 0.93 versus 0.92) [38]. Absolute reliability of measures was generally acceptable (i.e., CV < 10%), except for surface area demonstrating larger CV values [6]. 48

Gait & Posture 75 (2020) 46–62

K. Hébert-Losier and L. Murray

Table 1 Quality appraisal scores based on the COnsensus-based Standards for the selection of health Measurement INstruments reliability checklist. Article

Patients stable

Time interval

Test conditions

Important flaws

ICC

Score

Centre of pressure Bauer et al. [2] Bauer et al. [3] Carpenter et al. [9] Chang et al. [11] Chiari et al. [12] Clark et al. [15] Corriveau et al. [17] Corriveau et al. [18] Doyle et al. [21] Geurts et al. [24] Golriz et al. [26] Hill et al. [33] Kitabayashi et al. [42] Lafond et al [43] Letz et al. [44] Levy et al. [45] Lin et al. [46] Mani et al. [47] Mattacola et al. [48] Moghadam et al. [50] Pinsault et al. [57] Raymakers et al. [59] Riley et al. [61] Santos et al. [64] Schmid et al. [65] Swanenburg et al. [69] Takala et al. [70]

√ √ √ X √ X X √ √ X √ X √ √ X X X √ √ X √ X X X X X X

√ √ √ √ √ √ √ √ X √ √ √ √ √ √ √ √ X √ √ √ √ X √ √ √ √

√ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ X √ √ √ √

X X X √ X √ √ X X X X X X √ X X √ √ X √ √ X X √ X √ X

√ √ √ √ √ √ √ √ √ X √ √ √ √ X √ √ √ √ √ √ X X √ √ √ √

4 4 4 4 4 4 4 4 3 2 4 3 4 5 2 3 4 4 4 4 5 2 0 4 3 4 3

Plantar pressure Becerro de Bengoa Vallejo et al. [6] Izquierdo-Renau et al. [38]

X X

X √

√ √

X X

√ √

2 3

Plantar-flexion isometric strength Bemben et al. [7] Clark et al. [14] Clarke et al. [16] Ford-Smith et al. [22] Fouré et al. [23] Joseph et al. [40] Mattes et al. [49] Moraux et al. [54] Sleivert et al. [67] Spink et al. [68] Topp et al. [72]

X √ X X X X X X √ X X

√ √ √ √ √ √ √ X √ √ √

X √ √ √ X X √ √ √ √ √

X √ √ √ X √ X X √ X X

X √ √ √ √ √ √ √ √ √ √

1 5 4 4 2 3 3 2 5 3 3

3.4. Plantar-flexion isometric strength

0.98), whilst inter-rater relative reliability ranged from poor to good (ICC range = 0.15 to 0.82) for peak [16] and mean force [68]. Where reported, CV values were generally deemed acceptable (i.e., below 10%) [7,14,23,68], although absolute reliability was less than optimal for select age groups [7] and when comparing data between raters [68].

The median quality score of the 11 PFisom reliability studies [7,14,16,22,23,40,49,54,67,68,72] was 3 out of 5 (range 1–5). Two studies [14,67] achieved the maximum quality rating of 5 out of 5. A total of 481 healthy participants were examined across studies, with a mean sample size of 44 (43) individuals, range: 14–155. Weighted mean values for participant characteristics were 43.3 (22.0) y, 175.7 (6.2) cm, and 75.7 (6.9) kg. All studies [7,14,16,22,23,40,49,54,67,68,72] measured peak torque, peak force, or rate of torque development. Furthermore, the reliability of several types of measures were also concurrently reported (e.g., electromyography and ultrasound measures). Most of the studies were testretest in design [7,14,22,23,40,54,66,67,72], with a range of 1 h to 12 weeks between testing sessions. Three studies reported intra-rater reliability [16,49,68], with two of these studies also reporting inter-rater reliability [16,68]. Typically, protocols were 3 × 3–5 s maximal contractions with a 30-s rest period (up to 180 s of rest) between contractions. Three studies did not state rest periods [40,67,68]. Overall, intra-session relative reliability for peak torque was excellent [40,72]. For inter-session, relative reliability was generally good to excellent for peak torque and force, good for mean force, and poor for peak rate of torque development (ICC: 0.13 [67]). Intra-rater relative reliability of PFisom measures was fair to excellent (ICC = 0.56 to

4. Discussion Clinicians and researchers often examine centre of pressure (COP), plantar pressure (PP), and plantar-flexion isometric strength (PFisom) measures in relation to normal and pathological postural control and gait. Knowledge on the reliability of these measures is important before interpreting changes in measures and using them to inform practice. This review critically appraised and summarised research addressing the reliability of three different biomechanical measures commonly used in research and clinical settings to inform practice. From the 40 studies examined, only 10% (n = 4) achieved the uppermost quality score of 5 out of 5 based on the COSMIN Box 6. Reliability checklist [51]. Future reliability studies in the area should provide evidence that participants were stable in the interim period and report methodological information in sufficient detail to allow reproducibility of protocols. Reliability of COP measures from force plates in static stance was the most researched (n = 27), with relative reliability ranging from 49

20 M, 29 F

Carpenter et al. [9]

50

T-RT

Chiari et al. [12]

6 F, 6 M

171.9 (8.3) cm 68.8 (11.3) kg

T-RT

Rater

15 M, 15 F 24.4 (3.9) y

Chang et al. [11]

T-RT

T-RT

22 F, 8 M 77.23 (6.81) y

21 M, 42 F 78.74 (6.65) y 1.61 (0.11) m

Participants

Bauer et al. [3]

T-RT

Centre of pressure Bauer et al. [2]

Design

Equipment: Bertec 4060-08 FP 20 Hz Measures: Mean velocity Area ellipse 95% Fractal dimension Centroidal frequency

Protocol: 10 x 50 s, 60 s rest Arms by sides, look at a target at eye level 3 m away;comfortable stance Conditions: EO, EC

Hands on hips; double-leg stance –medial malleoli in contact, EC Conditions: firm and foam, double-leg, single-leg and tandem Time between sessions: 7 days

Equipment: Inverted Wii Balance Board affixed to OR6-7-2000 FP Customised software in LabVIEW Quickcam Pro 5000 3 raters Measures: Path length

Measures: Mean power frequency Mean position

Stand quietly on a FP with feet in a box; dimensions equal to foot length (feet traced); arms by sides, head face-forward position; eyes on target ˜2 m away Protocol: 1 x 20 s

Equipment: 20 Hz

Measures: Trajectory: ML, AP

Conditions: 2 cm heel distance; 30° between feet, EC and EO Narrow stance, EC and EO Protocol: 3 x 120 s, 120 s seated rest

Equipment: SATEL FP and software, 40 Hz

Measures: Area 95 Length ML sway AP sway

Conditions: Heels 2 cm apart, 30˚ between feet (normal stance); Narrow stance; EO, EC Protocol: 3 x 30 s, 2 min rest No shoes, heads erect, arms by sides; instructed to maintain balance.

Equipment: SATEL FP, 40 Hz custom software

Equipment & measures

Protocol: 3 x 30 s, 2 min rest No shoes, heads erect, arms resting at sides, instructions to maintain balance.

Protocol

Table 2 Key characteristics and results from reliability studies on centre of pressure, plantar pressure, and plantar-flexion isometric strength.

ML sway: 0.918 (0.874, 0.948) AP Sway: 0.926 (0.874, 0.956)

Length: 0.885 (0.825, 0.9527) ML sway: 0.899 (0.844, 0.936) AP Sway: 0.843 (0.763, 0.900) EO, narrow stance: Area 95: 0.878 (0.814, 0.922) Length: 0.886 (0.826, 0.927) ML sway: 0.841(0.758, 0.899) AP sway: 0.907 (0.858, 0.941)

Area 95: 0.58

ICC EO: Mean velocity: 0.83

FP: 0.66 Wii Balance board: 0.68 BESS raters: No value EC, inter-rater: No value

ICC [2,1] EC, T-RT:

(continued on next page)

EC: Mean velocity: 0.87 Area 95: 0.70

WBB: 0.99 BESS: No value

Pearson correlation (r) EC, validity (compared to FP):

ML 60 s: 0.84 AP 60 s: 0.89 ML 120 s: 0.85 AP 120 s: 0.91

EC, narrow stance: ML: 0.906 (0.791, 0.962),21% AP: 0.853 (0.675, 0.940), 26%

AP: 0.655 (0.258, 0.864), 24% EO, narrow stance: ML: 0.846 (0.658, 0.937), 20% AP: 0.828 (0.619, 0.930), 19% ICC EO, mean position: ML 15 s: 0.75 AP 15 s: 0.86 ML 30 s: 0.79 AP 30 s: 0.87

EC, normal stance: ML: 0.806 (0.570, 0.921), 25% AP: 0.792 (0.539, 0.914), 19%

ICC [2,1] (99% CI), CV (%) EO, normal stance: ML: 0.706 (0.349, 0.880), 28%

EC, narrow stance: Area 95: 0.710 (0.553, 0.818) Length: 0.945 (0.915, 0.966) ML sway: 0.933 (0.896, 0.958) AP sway: 0.946 (0.915, 0.967)

EC, normal stance: Area 95: 0.945 (0.917, 0.965) Length: 0.933 (0.891, 0.959)

ICC [2,1] (95% CI): EO, normal stance: Area 95: 0.873 (0.803, 0.920)

Reliability of selected measures

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

69.6 (11.3) kg 1.59 (2.50) m

T-RT

51

T-RT

Geurts et al. [24]

T-RT

Doyle et al. [21]

Rater

Group 2: 4 M, 4 F 24.9 (2.4) y

Group 1: 4 M, 4 F 44.3 (19.7) y

71 (12) kg

10 F, 20 M 23 (5) y 1.75 (0.09) m

18 F, 27 M 70.5 (6.0) y

Corriveau et al. [18]

T-RT

4 F, 3 M 68.6 (4.3) y

60 s rest between device or task

6.38 (15.20) kg

Corriveau et al. [17]

3 x 30 s, 15 s rest Hands on hips, as still as possible;

23.7 (5.6) y 1.68 (0.09) m

T-RT

Group 1: 3 x 20 s, 1 min rest Group 2: 2 x 30 s, 1 min rest Feet against a foot frame; (medial heels 8.4 cm apart, toeing-out 9˚); hands clasped lightly behind back Conditions: Group 1: EO, blurred vision, EC (with dark glasses) Group 2: single task, dual task Time between sessions: Biweekly

3 x 10 s CTSIB-M; Feet position based on height; (21, 25 or 30 cm width); arms by sides Conditions: EO, EC, rigid and foam Randomized order of testing

Double leg stance, feet pelvis width (feet traced); max 14˚ hip external rotation; flat-soled shoes; look straight with head erect, arms by sides Conditions: EO, EC Time between sessions: Intra-rater: 30 min Inter-rater and T-RT: 3- 7 days

4 x 120 s, 5 min between trials, 10 min between condition

Stood quietly, look straight ahead, arms comfortable at sides, EO

Protocol: 11 x 120 s, 5 min rest between trials

Conditions: Feet together: EO, EC Single limb, EO, EC Random order of tasks and devices Time between sessions: Within 2 weeks, at least 24 h apart

Protocol:

10 M, 20 F

Clark et al. [15]

Protocol

Participants

Design

Table 2 (continued)

RMS, AP: 37% Mean frequency: 31% Mean frequency: 36% RMS velocity, ML: 35% RMS velocity, AP: 24%

Measures: RMS amplitude: ML, AP

RMS vrelocity: AP, ML Peak-to-peak amplitude: AP, ML

Mean frequency: AP, ML

Mean CV % EO, group 1: RMS, ML: 39%

Range of sway AP: 0.43 Peak sway velocity ML: 0.29 Peak sway velocity AP: 0.12 Total excursion area: 0.49

ICC [2,1] EO: Range of sway ML: 0.71

Rater ML: 0.66 (0.45, 0.80) Rater AP: 0.92 (0.87, 0.96) T-RT ML: 0.72 (0.43, 0.83) T-RT AP: 0.91 (0.85, 0.95)

ICC [2,1] (95% CI) EO, COP-COM, mean:

ICC [2,1] (95% CI), MMDC EO, COP-COM, single trial: ML: 0.64 (0.44, 0.92) AP: 0.79 (0.58, 0.99)

Between device: Day 1: 0.77 (0.46, 0.90) Day 2: 0.78 (0.54, 0.90)

WBB: 0.66 (0.20, 0.85), 4.0, 27.9%

ICC [2,1] (95% CI), SEM (cm), MMDC EO, path length: FP: 0.86 (0.71, 0.93), 2.2, 14.5%

Reliability of selected measures

Equipment: FP: Group 1: 100 Hz, Group 2: 30 Hz

Measures: Range of sway: AP, ML Peak sway velocity Total excursion area: AP, ML Fractal dimension: AP, ML

Equipment: Fitness Technologies FP, 100 Hz

Measures: Root square mean COP-COM: AP, ML

Equipment: 2 x AMTI FP, 20 Hz, MATLAB 5.1, 2 raters

Equipment: 2 x AMTI FP, 20 Hz 3 x OPTOTRAK sensors, 20 Hz Matlab 5.1 Measures: RMS COP-COM: AP, ML

WBB – custom-software Labview 8.5, 40 Hz Measures: Total path length

ATMI Model OR6-5, 40 Hz

Equipment:

Equipment & measures

(continued on next page)

Mean frequency: 30% Mean frequency: 32% RMS velocity, ML: 35% RMS velocity, AP: 20%

EC, group 1: RMS, ML: 36% RMS, AP: 33%

Peak sway velocity ML: 0.19 Peak sway velocity AP: 0.58 Total excursion area: 0.95

EC: Range of sway ML: 0.51 Range of sway AP: 0.65

Rater AP: 0.92 (0.86, 0.96) T-RT ML: 0.72 (0.53, 0.83) T-RT AP: 0.90 (0.83, 0.94)

EC, COP-COM, mean: Rater ML: 0.79 (0.64, 0.88)

EO, COP-COM, mean 4 trials: ML: 0.90, 16 mm AP: 0.94, 10 mm

Between device: Day 1: 0.89 (0.71, 0.95) Day 2: 0.88 (0.67, 0.95)

FP: 0.94 (0.87, 0.97), 4.0, 16.1% WBB: 0.91 (0.80, 0.96), 6.6, 24.5%

EC, path length:

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

52

4 F, 3 M 67.9 (4.3) y

Lafond et al. [43]

T-RT

108 M, 112 F 20.1 (1.6) y, 19.6 (1.4) y 173.3 (5.9) cm, 161.0 (5.8) cm 67.0 (7.9) kg, 54.3 (6.1) kg

Kitabayashi et al. [42]

T-RT

17 participants 69.5 (7.3) y

Hill et al. [33]

Path length Area Velocity: X, Y axis Distribution of amplitude Power spectrum Vector

EO

Equipment: 2 x Model OR6-5 FP, 20 Hz

Measures:

Conditions:

Double-leg stance pelvis width stance (feet traced); EO, look straight ahead, head erect;

Equipment: Anima stabilometer G5500, 20 Hz

Measures: COP

Equipment: Chattecx Balance System

Body weight %: L, R

Sway area

Average COP location

3 x 1 min, 1 min rest Barefoot, arms comfortably at sides

Conditions: EO: Feet apart, feet together; Sharpened Romberg, stable and rotating platform at 50% and 100% speed Time between sessions: 7 days

9 x 25 s, 1 min rest Shoes removed, safety harness at waist; feet 12 cm apart or feet together; looking at picture in front, hands by sides

Measures: COP mean velocity

Time between sessions: 5 min

25.6 (5.5) kg/m2

T-RT

Midot posture scale analyser QPS 200, 200 Hz

Feet shoulder width apart (traced); shoeless, arms by sides, breather normally; distribute weight evenly on both feet;

30.5 (7.2) y

Equipment:

5 x 60 s, 1 min rest (allowed to sit)

16 M, 14 F

Equipment & measures

Golriz et al. [26]

Protocol

Participants

Design

Table 2 (continued)

ICC [2,1] (95% lower bound): EO, sway area:

Velocity X-axis: 0.96 Velocity Y-axis: 0.96

Area circle: 0.90

ICC EO: Path length: 0.97

Feet apart: 0.55, 17% Feet together: 0.27, 19%

ICC [2,1], CV (%) EO:

1 rep: 0.19 (-0.75, 0.62), 4.4, 12.1, -1.9 – 0.7 2 reps: 0.83 (0.65, 0.92), 1.3, 3.5, 0.1 – 0.4 3 reps: 0.95 (0.90, 0.98), 0.7, 2, 0,0 – 0.5 4 reps: 0.97 (0.94, 0.99), 0.6, 1.6, -0.1 – 0.3 5 reps: 0.92 (0.84, 0.96), 0.9, 2.5, -0.3 – 0.3 EO, location (mm): 1 rep: 0.53 (-0.01, 0.78), 14.4, 39.7, -6.2 – 2.6 2 reps: 0.92 (0.84, 0.96), 8.8, 24.4, -4.7 – 1.5 3 reps: 0.91 (0.81, 0.96), 8, 22.1, -3.2 – 2.4 4 reps: 0.93 (0.85, 0.97), 7.7, 21.3, -3.4 – 2.1 5 reps: 0.94 (0.88, 0.97), 7.3, 22, -3.2 – 2.0

ICC [3,k] (95% CI), SEM (units), MMDC (units), LOA (units) EO, velocity (mm/s):

Reliability of selected measures

(continued on next page)

EO, COP mean velocity:

1 rep: 0.06 (-1.02, 0.56), 1556.9, 4302.7, -686.4 – 155.1 2 reps: 0.47(-0.13, 0.75), 763.7, 2116.8, -302.3 – 156.3 3 reps: 0.63 (0.28, 0.82), 612.7, 1698.3, -350.4 – 32.1 4 reps: 0.68 (0.33, 0.85), 504.8, 1399.2, -278.9 – 43.8 5 reps: 0.83 (0.64, 0.92), 380.4, 1054.3, -178.7 – 81.7

EO, sway area (mm2):

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

53

Mani et al. [47]

T-RT

Lin et al. [46]

T-RT

Levy et al. [45]

T-RT

Letz et al. [44]

65.6 (17.5) kg 161 (12) cm

T-RT

36 M 40 (20) y

Older: 8 M, 8 F 65.4 (3.7) y, 60.8 (6.4) y 175.5 (8.1) cm, 160.2 (7.5) cm 88.9 (13.3) kg, 66.2 (15.8) kg

Younger: 8 M, 8 F 20.3 (1.4) y, 21.5 (2.0) y 176.1 (4.6) cm, 166.1 (5.2) cm 74.7 (12.1 kg, 59.6 (5.1) kg

75.8 (7.7) y

16 M, 31 F

2 x 60 s

15 F, 15 M 23 – 60 y

9 x 120 s, 5 min rest

Participants

Design

Table 2 (continued)

Velocity ML: 0.95 (0.91) Sway area: 0.92 (0.84)

Detrended fluctuation analysis: ML, AP Equipment: AMTI model BP2436, 10.5 Hz,

Velocity AP: 0.95 (0.90)

Hurts rescaled analysis: ML, AP

3 x 30 s, no formal rest

Velocity ML: 0.91 (0.81) Velocity AP: 0.86 (0.72) Sway area: 0.79 (0.60) Older:

Mean velocity: ML, AP Median power frequency: ML, AP RMS distance: ML, AP Sway area

Min. 2 days

ICC [2,3] (95% CI) EO, composite scores:

Younger:

Measures:

Time between sessions:

ICC (95% one-sided lower CI) EC, intra-session:

EC, T-RT: BTrackS: 0.83 (0.71, 0.90), 7.0, 19.4

BTrackS: 0.83 (0.71, 0.89), 3.47, 9.6

ICC [2,1] (95% CI), SEM (cm), MMDC 95% EO, T-RT:

Mean Pearson correlation r EO: Sway ML: 0.68 Sway AP: 0.53 Area: 0.85 Speed: 0.92

30 s: 0.22 60 s: 0.47 120 s: 0.41 (0.16) EO, COP range: ML 30 s: 0.44 AP 30 s: 0.29 ML 60 s: 0.57 AP 60 s: 0.38 ML 120 s: 0.62 (0.35) AP 120 s: 0.52 (0.25)

Reliability of selected measures

Equipment: AMTI OR6-7-1000 FP, 100 Hz

Measures: Sway: AP, ML

AMTI OR6-7 2000 FP, BTrackS, 20 Hz, customized LabView

Equipment:

Equipment: AMTI OR6-3 FP Measures: RMS distance: ML, AP Mean sway radius: ML, AP Sway path: ML, AP Area (triangle) Sway speed

Measures: RMS Sway area COP range COP mean velocity Mean power frequency Median power frequency

Equipment & measures

3 x 75 s, 1 min rest Barefoot, feet together (feet traced); as still as possible; EC, arms at side, head straight ahead

Time between sessions: 3 days

Hands on hips, feet shoulder width; instructed to stay as still as possible Conditions: EO, EC

6 x 20 s (3 x EO and 3 x EC), 15 s rest

Time between testing sessions: 6 – 12 days

Conditions: EO, EC

No shoes, feet together

arms by sides

Protocol

(continued on next page)

Velocity AP: 0.92 (0.87) Sway area: 0.90 (0.83)

Velocity ML: 0.91 (0.85)

Older:

Velocity AP: 0.77 (0.65) Sway area: 0.72 (0.59)

Velocity ML: 0.79 (0.67)

Younger:

EC, inter-session:

FP: 0.92 (0.88, 0.95) BTrackS: 0.95 (0.91, 0.97) EC, validity: FP: 0.95 (0.92, 0.97) BTrackS: 0.97 (0.94, 0.98)

EO, validity:

Pearson correlation r (95% CI)

EC: Sway ML: 0.84 Sway AP: 0.83 Area: 0.96 Speed: 0.96

ML 30 s: 0.87 AP 30 s: 0.73 ML 60 s: 0.90 AP 60 s: 0.77 ML 120 s: 0.94 (0.85) AP 120 s: 0.83 (0.64)

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

10 F, 2 M 24.7 (3.3) y 62.2 (7.5) kg 164.8 (7.1) cm

Mattacola et al. [48]

54

161.4 (6.22) cm 68.65 (9.57) kg

5 M, 5 F 24.6 (2.5) y

175.1 (10.1) cm

T-RT

Pinsault et al. [57]

T-RT

68.9 (14.2) kg

10 F, 6 M 69.6 (4.5) y

Moghadam et al. [50]

2 x 10 s

1.80 (0.06) m 79.25 (10.58) kg

T-RT

Rater

Participants

Design

Table 2 (continued)

Area

1h

Range Mean velocity Max velocity

Measures:

Equipment: Equi+ model PF01, 64 Hz

Measures: SD of amplitude: ML, AP SD of velocity: ML, AP Phase plane portrait: AP, ML, total Mean velocity Area 95 ellipse

Time between sessions:

10 x 30 s, 60 s rest Barefoot, EC, natural position; feet traced; (feet abducted at 30°, heels separated by 3 cm); arms ny sides, stand as still as possible

Conditions: EO, EC and foam EC (blindfold), dual task Time between sessions: 7 days

Bertec 4060-10 FP 100 Hz

Equipment: Strain gauge

Measures: Sway

Conditions: Double leg: static and dynamic, EO, EC Single leg: static, dominant and non-dominant leg, EO, EC Single leg: dynamic, dominant and non-dominant, EO Time between raters: 30 min 3 x 30 s, 1 min rest Quiet stance barefoot, EO looking straight ahead; arms by sides, feet 50% hip-to-hip distance, EC blindfolded

Equipment: Chattecx Dynamic Balance system 2 raters

Measures: Mean distance RMS distance Total excursion area Mean velocity Area 95 ellipse Area 95 circle Sway area Mean rotational frequency Composite and ML, AP for each measure

Scilab 5.2.2 software

Equipment & measures

Focus on mark on wall; barefoot, knees slightly flexed (5 to 15˚); arms by sides, stand as still as possible

Time between tests: 2-3 min between each testing period

Shoes, feet in normal manner (feet traced), arms by their sides, EO, looking straight ahead (mark 1.5m, eye level) Conditions: Bipedal, unipedal, limits of stability task, lifting task Counterbalanced

Protocol

EC, 3 trials avg: Area (mm2): 0.94 (0.81, 0.98), -34.9 – 28.3 Velocity (mm/s): 0.84 (0.55, 0.95), -15.5 – 13.7 Max velocity (mm/s): 0.80 (0.29, 0.94), -4.9 – 6.5

Area (mm2): 0.61 (0.08, 0.89), -85.2 – 68.0 Velocity (mm/s): 0.82 (0.57, 0.92), -2.5 – 1.3 Max velocity (mm/s): 0.79 (0.45, 0.94), -10.0 – 15.1

ICC [2,1] (95% CI), LOA (units) EC, 1 trial:

Velocity: 0.89 (0.58, 0.97) Area 95: 0.86 (0.44, 0.96)

ICC [2,3] (95% CI) EO:

ICC [2,1], SEM (cm) EO: 0.06, 0.26

Distance: 0.84 (0.71, 0.91) Excursion: 0.95 (0.91, 0.97) Velocity: 0.95 (0.91, 0.97) Ellipse area: 0.84 (0.71, 0.91) Sway area: 0.86 (0.75, 0.92)

Reliability of selected measures

(continued on next page)

EC, 10 trial avg: Area (mm2): 0.91 (0.72, 0.95), -47.7 – 49.1 Velocity (mm/s): 0.89 (0.64, 0.97), -1.9 – 1.2 Max velocity (mm/s): 0.81 (0.29, 0.95), -5.8 – 5.7

Area 95: 0.80 (0.18, 0.95)

EC: Velocity: 0.70 (0.00, 0.92)

EC: 0.75, 0.06

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

7 M, 4 F 50.25 (22.63) y 1.71 (0.09) m

Riley et al. [61]

55 Barefoot (feet traced), double-leg stance; arms by side, looking straight ahead

18 F, 8 M 71 (6) y 69 (11) kg

166 (8) cm

T-RT

4 x 20 s, 20 s rest, 2 min in between tasks

Time between sessions: 1 - 3 days

Swanenburg et al. [69]

T-RT

3 trials, 10 min rest EO looking 3 m in front, feet hip width apart; arms by side

EO, EC counterbalanced Time between sessions: No later than 7 days

Barefoot, feet parallel on both sides of a 5.1 cm T-shape separator; arms by sides looking 2 m ahead Conditions:

8 x 60 s (4 x EO, 4 x EC)

Equipment: AMTI Accusway, 50 Hz SWAYWIN software

Equipment: Bertec 4060-08 FP, 400 Hz Step PC software Measures: Mean velocity Mean amplitude Sway area Mean power frequency Centroidal frequency

Measures: RMS distance Mean velocity COP range Mean frequency Median power frequency Sway area Area 95 ellipse

Equipment: AMTI BP900900 FP, 100 Hz In-house C++, MATLAB

Measures: Centre of gravity: ML, AP Centre of pressure: ML, AP

Conditions: Wide base (heels 30 cm apart): EO Narrow base: EO, EC Semitandem (1 cm apart)

4 M, 4 F 24 – 32 y

74.9 (13.1) kg

12 M 26.9 (4.7) y 1.75 (0.07) m

Equipment: Kistler FPs SELSPOT II/TRACK acquisition, 153 Hz

Measures: Sway range: AP, ML Mean displacement velocity Mean velocity

Equipment: Kistler FP, 10 Hz Bioware software and customised program

Equipment & measures

2x7s One foot on each FP, feet in one of three stances

Time between sessions: 7 days

EO looking at wall 150 cm in front Conditions: EO, cognitive, foam, EC

Barefoot, feet parallel to a 4 cm T-shaped separator;

60 – 120 s

Protocol

Schmid et al. [65]

T-RT

Santos et al. [64]

69.17 (11.29) kg 23.40 (2.09) kg/m2

45 young, 38 older 21 – 45 y, 61 – 78 y

Raymakers et al. [59]

T-RT

Participants

Design

Table 2 (continued)

ICC [3,1] (95% CI), MMDC, LOA EO, T-RT: Max-ML (cm): 0.75 (0.52, 0.88), 0.37, 0.00 (0.37) Max-AP (cm): 0.43 (0.06, 0.70), 0.55, 0.01 (0.55)

ICC EO, 0.8 Hz: Velocity: 0.75 Sway area: 0.62

Velocity ML: 0.46 Range AP: 0.55 Range ML: 0.48 Sway: 0.55 Area 95: 0.40

ICC EO, one trial: Velocity AP: 0.44

COP ML: 0.91 COP AP: 0.78

Pearson correlation r COG ML: 0.75 COG AP: 0.50

Range ML: 19% Velocity: 14% Area: 26%

Standardised CV% EO: Range AP: 28%

Reliability of selected measures

(continued on next page)

ICC [2,1] EO, rater: Max-ML (cm): 0.80 (0.60, 0.90), 0.34, 0.04 (0.34) Max-AP (cm): 0.56 (0.24, 0.72), 0.53, 0.01 (0.53)

EO, 10 Hz: Velocity: 0.71 Sway area: 0.55

Range AP: 0.19 Range ML: 0.36 Sway: 0.38 Area 95: 0.43

EC, one trial: Velocity AP: 0.32 Velocity ML: 0.41

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

56

Izquierdo-Renau et al. [38]

T-RT

Simulate gait (walk in place) 15 s, Stand in a natural manner; Look straight ahead;

47.7 (16.7) y 71.6 (13.5) kg 167 (8.1) cm

5 trials Stand centre of platform; look ahead, arms naturally side of body; simulate gait (walk in place) for 15 s

23 F, 17 M

28.78 (11.43) y

Arms close to body Time between sessions: Range: 3-20 days 9.90 (4.11) days

5 x 30 s

Conditions: Two-feet: EO, EC One-foot: each foot, EO Stepping responses: both feet Time between sessions: 1 day and 9 months

3 (maximal) x 30 s Two-feet stance (4 cm apart), arms crossed; facing wall 150 cm in front, no shoes or thick socks

36 M, 20 F

1.73 (0.10) m 69.5 (9.3) kg

T-RT

Plantar pressure Becerro de Bengoa Vallejo et al. [6]

9 M, 9 F 38.7 (10.9) y

RMS: ML, AP Mean velocity Area 95

Time between sessions: 7 days

Takala et al. [70]

Max sway: ML, AP

EO, EC, no task, with task

29.4 (3) cm (base of support width)

S-Plate platform, max 100 Hz

Equipment:

Measures: BW% Surface area %BW bilateral fore-foot %BW bilateral rear-foot %BW fore-foot %BW rear-foot Mean pressure

EPS-Platform 60 Hz Foot Checker v.40 for Windows

Equipment:

Measures: Max sway: AP, ML Mean amplitude Sway velocity Mean sway frequency Sway area

Equipment: Custom made FP, 40 Hz

Measures:

Conditions:

34.4 (2) cm (hip width)

Equipment & measures

Rater

Protocol

Participants

Design

Table 2 (continued)

Intra-session, session 1:

ICC [1,1] and [1,k], CV (%)

Intra-session, session 2: %BW L: 0.70 and 0.92. 4.9% %BW R: 0.70 and 0.92, 4.2% Mean pressure L: 0.87 and 0.97, 7.1% Mean pressure R: 0.88 and 0.97, 6.4% Surface area L: 0.74 and 0.94, 10.3% Surface area R: 0.74 and 0.94, 9.1%

Intra-session, session 1: %BW L: 0.68 and 0.92, 4.8% %BW R: 0.68 and 0.92, 4.0% Mean pressure L: 0.78 and 0.95, 8.5% Mean pressure R: 0.85 and 0.97, 6.9% Surface area L: 0.67 and 0.91, 12.2% Surface area R: 0.57 and 0.87, 12.3%

ICC [1,1] and [1,k], CV (%)

Sway velocity: 0.64 Sway area: 0.57 EC, consecutive days: Sway velocity: 0.56 Sway area: 0.31

ICC EO, consecutive days:

EC, T-RT: Max-ML (cm): 0.83 (0.65, 0.92), 0.37, 0.05 (0.37) Max-AP (cm): 0.83 (0.65, 0.92), 0.70, 0.07 (0.70) Velocity (cm/s): 0.87 (0.74, 0.94), 0.98, 0.18 (0.90) Area 95 (cm2): 0.73 (0.49, 0.87), 2.40, 0.36 (2.40)

Velocity (cm/s): 0.84 (0.68, 0.93), 0.48, 0.03 (0.48) Area 95 (cm2): 0.62 (0.32, 0.81), 1.48, 0.06 (1.48)

Reliability of selected measures

(continued on next page)

Inter-session:

ICC [1,k]

ICC [1,k], SEM (%, kPa or cm2), error (%) Inter-session %BW L: 0.95, 0.9, 3.9% %BW R: 0.95, 0.9, 3.3% Mean pressure L: 0.98, 2.4, 6.0% Mean pressure R: 0.98, 2.5, 5.7% Surface area L: 0.95, 8.2, 14.9% Surface area R: 0.93, 8.6, 14.8%

EC, 9 months later: Sway velocity: 0.77 Sway area: 0.54

Sway area: 0.64

EO, 9 months later: Sway velocity: 0.86

EC, rater: Max-ML (cm): 0.78 (0.57, 0.90), 0.35, 0.03 (0.35) Max-AP (cm): 0.84 (0.67, 0.92), 0.48, 0.03 (0.48) Velocity (cm/s): 0.89 (0.77, 0.95), 0.93, -0.09 (0.89) Area 95 (cm2): 0.76 (0.54, 0.89). 2.11, -0.12 (2.11)

Velocity (cm/s): 0.81 (0.57, 0.87), 0.55, 0.08 (0.55) Area 95 (cm/s2): 0.65 (0.35, 0.83), 1.61, 0.12 (1.61)

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

68.24 (13.53) kg 1.68 (0.09) m

T-RT

57 3 x 5 s MVIC, 30 s rest, 3 m rest between testers Warm up, leg set by coin toss; Position: prone lying on plinth, neutral ankle, hands by side, palms up Conditions: Constant order: Knee extension, hip extension, ankle plantarflexion Time between ratings: 7 days

20 M, 18 F

21.8 (2.4) y

Clarke et al. [16]

Rater

Seated, L keg, hip, knee, and ankle joints secured at 90˚

20.9 (0.72) y 165.9 (2.4) cm 64.4 (2.4) kg

T-RT Time between sessions: 4 weeks

Min 4 x ˜3-4 s MVIC, 1-2 min rest

12 F, 5 M

Time between sessions: 24 hr

Measures: Peak force (N)

MicroFET3 HHD

Equipment:

Custom-modified Parabody 826, MLP-300-T force transducer, 625 Hz Other: EMG recordings, electrical stimulation, mechanical recording, MRI Measures: Peak force (N)

Equipment:

−61.7 – 45.3

−51.7 – 72.4 Rater 2: 0.88 (0.78, 0.94), 18.91,

ICC [3,1] (95% CI), SEM (N), 95% LOA (N) Intra-rater: Rater 1: 0.56 (0.29, 0.74), 21.89,

ICC [2,1] (95% CI), mean CV%, ratio LOA Peak force: 0.97 (0.92, 0.99), 4.19%, 15.15

Total impulse: 0.91

Measures: Peak force (N) Time to peak force Peak force rate (60ms) Total impulse

Plantar-flexion: R knee 180˚ and ankle 90˚ Conditions: Finger flexors, thumb abductors, forearm extensors; dorsiflexors, plantar-flexors (random order)

Pearson correlationr, CV% All age groups: Peak force: 0.99

Peak pressure: 0.73, 8.85% Mean pressure: 0.82, 6.37% BW L: 0.82, 9.18% BW R: 0.80, 8.62% Total contact area: 0.91, 6.28% Contact area L: 0.85, 8.88% Contact area R: 0.85, 8.43% Intra-session, session 2: Peak pressure: 0.69, 8.43% Mean pressure: 0.86, 4.88% BW L: 0.85, 8.40% BW R: 0.82, 8.10% Total contact area: 0.95, 4.77% Contact area L: 0.90, 7.68% Contact area R: 0.88, 7.64%

Reliability of selected measures

Peak rate: 0.98

Equipment: Daytronic model 9130 or model 300

Measures: Peak pressure: L, R Mean pressure: L, R Weight each foot Total foot contact Foot contact: L, R

Equipment & measures

L leg extended, knees over edge

3 x MVIC, 1 min rest, 5 min between muscle groups Force testing table, semi-reclined; Hands on hips;

Time between sessions: 7 days

Protocol

Clark et al. [14]

Plantar-flexion isometric strength Bemben et al. [7] 155 M, 12 age groups 22.2 (1.7) y, 77.0 (1.4) y T-RT 76.0 (7.3) kg, 74.7 (2.8) kg 177.1 (6.1) cm, 175.9 (8.2) cm

Participants

Design

Table 2 (continued)

y: y: y: y: y: y: y: y: y:

0.84, 0.80, 0.93. 0.90, 0.77, 0.87, 0.66, 0.62, 0.37,

7.8% 9.9% 6.6% 6.3% 11.7% 9.7% 12.8% 15.1% 18.9%

(continued on next page)

approx. -150 – 50 N Day 2: 0.15 (0.04, 0.37)

Inter-rater: Day 1:0.23 (0.03, 0.45),

ICC [2,1] (95% CI), LOA (N)

35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79

30-34 y: 0.62, 12.2%

Peak force: 20-24 y: 0.90, 6.9% 25-29 y: 0.77, 10.6%

Peak pressure: 0.86 Mean pressure: 0.93 BW L: 0.97 BW R: 0.97 Total contact area: 0.97 Contact area L: 0.96 Contact area R: 0.96

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

58

Sleivert et al. [67]

20 M, 3 F 24.7 (3.6) y

76 participants

Moraux et al. [54]

T-RT

181.4 (4.7) cm 79.4 (9.3) kg

3 x 3 s at each speed: 0, 1.05, 2.10. 3.14 and 4.19rad/s

Conditions: Dorsiflexion and plantar-flexion: L, R Time between sessions: At least 1 h up to 30 days

2 x 2-4 s, 30 s rest; Seated, 90° hip, knee and foot; foot flat; pull strap

Conditions: Isometric strength then Isokinetic fatigue (10 x 6 contractions, angular velocity 60˚/s, 10 s rest) Time between sessions: 3 – 7 days

Warm up: 10 min at 9km/h on treadmill

Equipment: Cybex UBXT dynamometer, 2000 Hz

Measures: Maximal torque (N.m)

Equipment: Home-made dynamometer

Measures: Intra-rater Maximum torque (N.m)

IsoMed 2000 dynamometer, 20 Hz

Equipment:

Measures: Peak torque (N.m)

rest for 15 min then ultrasound measures and MVIC Time between sessions: 12 weeks

26.6 (4.3) y

Rater

Mattes et al. [49]

Biodex System 4

Equipment: Phillips HD11 ultrasound

Seated, hip flexed, knee extended, ankle neutral;

3 s MVIC, ramp up and ramp down of 5 s each

Measures: Peak torque

Time between sessions: 2 days

3 x 5 s, 3 min rest

5 M, 5 F 180 (4.9) cm, 165.2 (7.1) cm 97 (14.3) kg, 67.8 (13.8) kg 24 (1.4) y, 23.6 (0.9) y

Joseph et al. [40]

Biodex research toolkit

Equipment: Biodex dynamometer

Measures: Peak force (kg) Composite force (kg)

Equipment: AccuForce II Digital Force Gage

Equipment & measures

Seated: hip 70˚ flexion, R knee 0˚, L leg flexed

2 MVIC, 2 min rest Warm up: 3 min sub-MVIC;

29 M

179.6 (9.1) cm, 166.2 (7.5) cm 74.3 (10.8) kg, 58.0 (8.6) kg

T-RT

T-RT

8 M, 6 F 24.1 (2.2), 20.7 (1.6) y

Fouré et al. [23]

Time between sessions: 7 days

Conditions: Flexor and extensor muscle groups for ankle, knee, and hip

3 x 3 s, approx. 30 s rest

T-RT

Supine on table with ankles over edge; ankle passively placed in neutral;

17 F, 8 M 74.5 (4.5) y

Ford-Smith et al. [22]

Protocol

Participants

Design

Table 2 (continued)

ICC, SEM (N.m) Peak torque: 0.72, 15

ICC [2,1], SEM (N.m), LOA (N.m) Max torque, plantar-flexion: 0.88, 11.0, 30.6

ICC [3,1] (95% CI), SEM (N.m), MMDC (N.m), LOA (N.m), Max torque: 0.98 (0.96, 0.99), 4.0, 7.8, -10.2 – 10.3

Inter-session: Peak torque: 0.95, 7.77

Peak torque: 0.99, 3.52

ICC, SEM (N.m) Intra-session:

ICC [2,k] (CI), CV%, SEM (N.m) Peak torque: 0.91 (0.74, 0.97), 5.4%, 6.7

Composite force: 0.71 (0.50, 0.84)

ICC [3,1] (90% CI) Peak force R: 0.77 (0.59, 0.88) Peak force L: 0.61 (0.36, 0.78)

Reliability of selected measures

(continued on next page)

ICC, SEM (%) % peak torque at peak RTD: 0.02, 8

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

59

9 M, 13 F 72.8 (5.1) y

77.1 (5.7) y 164.4 (10.3) cm 73.8 (14.0) kg 27.2 (3.7) kg/m2

latereal malleous aligned with edge of platform Conditions: Isometric: dorsiflexion, plantar-flexion Isokinetic concentric/eccentric: dorsiflexion, plantar-flexion Time between sessions: 7 days

3 MVIC, 30 s rest Seated, both knees fully extended; no back support, hip flexion ˜90˚; Measures: Peak torque (N.m)

Equipment: Microfet HHD 2-inch diameter concave pressure distributing plate

Measures: Mean force (N)

Equipment: Citec HHD

T-RT: 0.76

Session 2: 0.92

Pearson correlation r Intra-session: Session 1: 0.93

Rater 2: 0.84 (0.76, 0.90), 14.1%, 85.8

Rater 1: 0.89 (0.83, 0.93), 7.9%, 52.0

ICC [3,1] (95% CI), CV%, MDC (N) Intra-rater:

Mean RTD: 0.63, 54 Peak RTD: 0.13, 278

ICC, SEM (N.m.s-1)

Reliability of selected measures

Inter-rater: Session 1: 0.80 (0.70, 0.87), 13.6%,77.9 Session 2: 0.82 (0.72, 0.88), 12.8%, 72.1

Abbreviations: AP, anterior-posterior; BESS, Balance Error Scoring System; BW, body weight; COP-COM, centre of pressure – centre of mass; CTSIB-M, Modified Clinical Test of Sensory Interaction in Balance; EC, eyes closed; EO, eyes open; F, female; FP, force plate; HHD, handheld dynamometer; ICC, intraclass correlation coefficient; L, left; LOA, limits of agreement; M, male; ML, medial-lateral; MMDC, minimal metrically detectable change; MVIC, maximal voluntary isometric contraction; R, right; RMS, root mean square; SEM, standard error of measurement; T-RT, test-restest.

T-RT

Topp et al. [72]

dynamometer on the plantar surface proximal to metatarsal heads Conditions: Ankle: dorsiflexion, plantar-flexion, inversion, eversion, lesser toe plantar-flexion, hallux plantar-flexion Time between sessions: 7 days

67.2 (11.3) kg

22.4 (2.6) kg/m2 17 M, 19 F

lower limb stabilized proximal to the ankle;

172.7 (9.1) cm

Rater

3 x 3-5 s contractions Supine, hips and knees extended;

Nerve conduction velocity, EMG Measures: Peak torque (N.m) Mean RTD (N.m.s-1) Peak RTD (N.m.s-1) % of peak torque at peak RTD

48 hrs

17 M, 19 F 23.2 (4.3) y

Other:

Time between sessions:

Spink et al. [68]

ATCODAS software

Equipment & measures

Supine, upper body secured; knee and ankle at 100˚

75.8 (9.6) kg 184.1 (6.3) cm Sum of 8 skinfolds 80.0 (32.9) mm

T-RT

Protocol

Participants

Design

Table 2 (continued)

K. Hébert-Losier and L. Murray

Gait & Posture 75 (2020) 46–62

Gait & Posture 75 (2020) 46–62

K. Hébert-Losier and L. Murray

poor to excellent and absolute reliability generally less than optimal. Research on the reliability PP measures from plantar pressure mats in static stance was scarce, with the available evidence indicating good to excellent relative reliability and adequate absolute reliability for relative load (i.e., percentage body weight), mean pressure, and contact area. PFisom measures from healthy participants assessed via dynamometers were generally reliable, except for rate of torque development.

findings indicate that multiple trials are often needed to obtain stable measures and improve their reliability [71]. Currently, it is unclear whether a balance task of longer duration than 30 s is associated with superior reliability of measures. 4.2. Plantar pressure There were only two studies [6,38] that assessed the reliability of PP measures in a static stance, both of which obtained poor quality scores. Static foot measures are often used to make inferences on dynamic plantar pressure measures [10,53,60,74] and screen for possible pressure abnormalities likely to lead to pathologies [6]. Therefore, the lack of strong-quality reliability studies on static PP is of considerable concern. The two studies on PP included in this review indicate that mean pressure, percentage body weight distribution, and total and individual foot contact area are reliable between sessions [6,38], with peak pressure demonstrating inferior reliability than mean pressure [38]. As stated above for COP measures, superior reliability of a mean rather than a peak measure is expected given the tendency of measures to regress towards the mean [5]. It could be recommended to clinicians and researchers to make inferences on changes in an individual’s condition based on the values of mean pressure instead of peak pressure given its superior reliability. The two PP studies included for review were undertaken barefoot [6,38]. However, the majority of daily and gait activities are completed wearing shoes. Research has shown differences in the way feet and shoes interact with the ground [8] and in gait patterns barefoot compared to shod [28]; hence, it might be relevant to undertake PP measurements in shod conditions as well, with currently limited information on the reliability of such PP assessments. With only two studies available at the time of this review, the dearth of information on the reliability of static PP measures is clear and needs to be addressed to inform both scientists and clinicians.

4.1. Centre of pressure Concisely summarising or making firm inferences about the reliability of COP measures proved difficult due to the heterogeneity in the age of participants, equipment, protocols, and outcome measures examined. The two studies that achieved the uppermost quality rating had relatively small (i.e., n ≤ 10) sample sizes [43,57]. Sway area, path length, and velocity of the COP during balance tasks were most commonly reported, and were associated with poor to excellent reliability levels (ICC range: 0.06 – 0.95). Although it has been proposed that other COP measures may provide a more in depth understanding of postural balance [59], these measures tend to be harder to comprehend and thus may be harder to integrate in clinical practice than COP sway area, path length, and velocity measures. Results from some of the articles of superior methodological quality (COSMIN score ≥4) indicated that sway area [2,57,69] and path length [2,15,42] exhibit good to excellent relative reliability within and between sessions in both eyes open and eyes closed conditions. Measures relating to velocity of the COP from eyes open balance tasks also demonstrated good to excellent test-retest reliability when 2 trials or more were preformed, with 4 trials proving to be the most reliable [26]. Based on the better-quality studies [2,3,15,18,26,42,57,69], clinicians and scientists can be confident that the relative reliability of sway area and path length measures from balance tests in both eyes closed and eyes open conditions is good. These measures may be useful to monitor changes over time in healthy populations. Velocity measures, however, should probably be derived from a minimum of 2 trials to be deemed reliable. Furthermore, the absolute reliability of COP measures is not as strong and demonstrates greater variability. These findings suggest that the reproducibility of the rank order of individuals on a balance task is superior to that of the reproducibility of an outcome for a given individual, which needs to be considered in the interpretation of outcomes from balance assessments in both research and practice. The large range of reliability values for some of the COP measures is likely due to several underlying factors, such as differences in trial duration, balance protocols, force plate sampling frequencies, filtering methods, number of trials assessed, and population groups. For example, data were acquired at sampling frequencies ranging from 10 to 400 Hz (Table 2), with significant differences shown to exist between COP measures sampled at 10 versus 50 Hz [59] and their reliability according to filtering frequencies [65]. Foot placement, for instance, is another factor that may affect reliability. Of the three studies that assessed different foot placements [2,3,33], the reliability of COP measures in a narrow and normal stance was found to be similar in two of the three studies, both in eyes open [2,3] and eyes closed conditions [2], but worse in a narrow stance in one of the studies [33]. The majority of studies [2,3,9,11,18,21,24,26,33,43,46,47,57,59,64,69,70] standardised foot placement by either tracing or implementing a predetermined foot width or angle. These practices likely improve reliability of measures, although they may not reflect habitual foot placement of individuals. Overall, results from the reliability studies reviewed show good to excellent relative reliability of COP measures from 3 × 30 s assessments. A few studies [17,26,57] demonstrated that averaging performance across several trials produced more reliable outcomes than a single trial. An enhanced reliability in mean rather than singular performance measures is expected given the tendency of measures to regress towards the mean when repeated, resulting in a reduced random variation or noise [5]. Indeed, previous research

4.3. Plantar-flexion isometric strength Overall, PFisom measures demonstrated good to excellent (ICC: 0.77 to 0.99) reliability for peak force and torque. The studies involved participants being assessed seated with either a handheld or isokinetic dynamometer. Although the measures of peak and mean torque were found to be reliable in these studies, plantar-flexors are functionally activated when in stance or during locomotion [55]. Assessing plantarflexion strength in stance would be a more valid measure reflecting plantar-flexor function. Measuring muscle force implies direct measurement of muscle output, which is not possible in vivo in clinical practice. The studies reviewed herein reported PFisom outputs either in terms of units of force (Newtons or kilograms) or torque (Nm), the latter considering moment arms and rotating effects of output forces on the measuring apparatus. Even though strength is often quantified in various fields of human movement as either force or torque, strength assessed in force rather than torque units requires a different allometric scaling to adjust for differences in body size [39]. Worth noting, however, are the findings from Baxter and Piazza [4] of no significant relationship between plantar flexor moment arm and body size, despite body size exhibiting a strong relationship with muscle volume. These findings suggest that the influence of moment arm on plantar flexor torque is more complex than a simple consideration of body size. Research findings from PFisom reported in force units should not be liberally compared to findings expressed in units of torque. Age of participants was reported to affect the reliability of maximal PFisom force measures [7], with generally good to excellent reliability in individuals aged 20–64 y and poor to fair reliability in individuals aged 65–79 y. The poorer reliability in older participants could be the result of sarcopenia, age-related muscle weakness, and lessened physical 60

Gait & Posture 75 (2020) 46–62

K. Hébert-Losier and L. Murray

activity levels with aging. Indeed, the maximal number of single-legged heel rises has been shown to decrease with age and physical activity levels in both males and females [32]. Rate of torque development from PFisom exhibited poor reliability [67,73]. These results are consistent with findings of more functional type exercises (i.e., squat, countermovement, and long jump) reflecting lower-limb strength, wherein the reliability of rate of torque development was suboptimal and lower than other metrics [31]. Care should be taken when interpreting changes in rate of torque development measures, as they are less reliable than other strength measures.

[13] T.H. Chow, Y.S. Chen, J.C. Wang, Characteristics of plantar pressures and related pain profiles in elite sprinters and recreational runners, J. Am. Podiatr. Med. Assoc. 108 (2018) 33–44. [14] B.C. Clark, S.B. Cook, L.L. Ploutz-Snyder, Reliability of techniques to assess human neuromuscular function in vivo, J. Electromyogr. Kinesiol. 17 (2007) 90–101. [15] R.A. Clark, A.L. Bryant, Y. Pua, P. McCrory, K. Bennell, M. Hunt, Validity and reliability of the Nintendo Wii Balance Board for assessment of standing balance, Gait Posture 31 (2010) 307–310. [16] M.N. Clarke, D.A. Ni Mhuircheartaigh, G.M. Walsh, J.M. Walsh, D. Meldrum, Intratester and inter-tester reliability of the MicroFET 3 hand-held dynamometer, Physiother. Pract. Res. 31 (2011) 13–18. [17] H. Corriveau, R. Hebert, F. Prince, M. Raiche, Intrasession reliability of the “center of pressure minus center of mass” variable of postural control in the healthy elderly, Arch. Phys. Med. Rehabil. 81 (2000) 45–48. [18] H. Corriveau, R. Hebert, F. Prince, M. Raiche, Postural control in the elderly: an analysis of test-retest and interrater reliability of the COP-COM variable, Arch. Phys. Med. Rehabil. 82 (2001) 80–85. [19] F. Degache, J. Van Zaen, L. Oehen, K. Guex, P. Trabucchi, G. Millet, Alterations in postural control during the world’s most challenging mountain ultra-marathon, PLoS One 9 (2014) e84554. [20] A.M. Dowling, J.R. Steele, L.A. Baur, What are the effects of obesity in children on plantar pressure distributions? Int. J. Obes. Relat. Metab. Disord. 28 (2004) 1514–1519. [21] T.L. Doyle, R.U. Newton, A.F. Burnett, Reliability of traditional and fractal dimension measures of quiet stance center of pressure in young, healthy people, Arch. Phys. Med. Rehabil. 86 (2005) 2034–2040. [22] C.D. Ford-Smith, J.F. Wyman, R. Elswick, T. Fernandez, Reliability of stationary dynamometer muscle strength testing in community-dwelling older adults, Arch. Phys. Med. Rehabil. 82 (2001) 1128–1132. [23] A. Fouré, A. Nordez, C. Cornu, In vivo assessment of both active and passive parts of the plantarflexors series elastic component stiffness using the alpha method: a reliability study, Int. J. Sports Med. 31 (2010) 51–57. [24] A.C. Geurts, B. Nienhaus, T.W. Mulder, Intrasubject variability of selected forceplatform parameters in the quantification of postural control, Arch. Phys. Med. Rehabil. 74 (1993) 1144–1150. [25] O. Girard, G. Millet, M. Jean-Paul, S. Racinais, Alteration in neuromuscular function after a 5 km running time trial, Eur. J. Appl. Physiol. 112 (2012) 2323–2330. [26] S. Golriz, J.J. Hebert, K.B. Foreman, B.F. Walker, The reliability of a portable clinical force plate used for the assessment of static postural control: repeated measures reliability study, Chiropr. Man. Therap. 20 (2012) 1–6. [27] K.D. Gross, D.T. Felson, J. Niu, D.J. Hunter, A. Guermazi, F.W. Roemer, A.B. Dufour, R.H. Gensure, M.T. Hannan, Association of flat feet with knee pain and cartilage damage in older adults, Arthritis Care Res. (Hoboken) 63 (2011) 937–944. [28] J.L. Hall, C. Barton, P. Jones, D. Morrissey, The biomechanical differences between barefoot and shod distance running: a systematic review and preliminary metaanalysis, Sports Med. (2013) 1–19. [29] M.L. Harringe, K. Halvorsen, P. Renström, S. Werner, Postural control measured as the center of pressure excursion in young female gymnasts with low back pain or lower extremity injury, Gait Posture 28 (2008) 38–45. [30] R. Hashish, S.D. Samarawickrame, M.Y. Wang, S.S. Yu, G.J. Salem, The association between unilateral heel-rise performance with static and dynamic balance in community dwelling older adults, Geriatr. Nurs. 36 (2015) 30–34. [31] K. Hébert-Losier, C.M. Beaven, The MARS for squat, countermovement, and standing long jump performance analyses: are measures reproducible? J. Strength Cond. Res. 28 (2014) 1849–1857. [32] K. Hébert-Losier, C. Wessman, M. Alricsson, U. Svantesson, Updated reliability and normative values for the standing heel-rise test in healthy adults, Physiotherapy 103 (2017) 446–452. [33] K. Hill, S. Carroll, A. Kalogeropoulos, J. Schwarz, Retest reliability of centre of pressure measures of standing balance in healthy older women, Australas. J. Ageing 14 (1995) 76–80. [34] P. Hlavackova, N. Vuillerme, Do somatosensory conditions from the foot and ankle affect postural responses to plantar-flexor muscles fatigue during bipedal quiet stance? Gait Posture 36 (2012) 16–19. [35] W.G. Hopkins, Measures of reliability in sports medicine and science, Sports Med. 30 (2000) 1–15. [36] W.G. Hopkins, S.W. Marshall, A.M. Batterham, J. Hanin, Progressive statistics for studies in sports medicine and exercise science, Med. Sci. Sports Exerc. 41 (2009) 3–13. [37] W.G. Hopkins, E.J. Schabort, J.A. Hawley, Reliability of power in physical performance tests, Sports Med. 31 (2001) 211–234. [38] M. Izquierdo-Renau, P. Pérez-Soriano, V. Ribas-García, A. Queralt, Intra and intersession repeatability and reliability of the S-Plate® pressure platform, Gait Posture 52 (2017) 224–226. [39] S. Jaric, S. Radosavljevic-Jaric, H. Johansson, Muscle force and muscle torque in humans require different methods when adjusting for differences in body size, Eur. J. Appl. Physiol. 87 (2002) 304–307. [40] M.F. Joseph, K.R. Lillie, D.J. Bergeron, C.R. Denegar, Measuring Achilles tendon mechanical properties: a reliable, noninvasive method, J. Strength Cond. Res. 26 (2012) 2017–2020. [41] A. Kamieniarz, J. Michalska, A. Brachman, M. Pawlowski, K.J. Slomka, G. Juras, A posturographic procedure assessing balance disorders in Parkinson’s disease: a systematic review, Clin. Interv. Aging 13 (2018) 2301–2316. [42] T. Kitabayashi, S. Demura, M. Noda, Examination of the factor structure of center of foot pressure movement and cross-validity, J. Physiol. Anthropol. Appl. Human Sci. 22 (2003) 265–272.

5. Limitations Our review purposely focused on healthy cohorts, static COP and PP measures, and PFisom to limit the scope of the review, minimise potential confounders from pathologies, and focus on assessments that are used in both practical and research settings. Generalisation of findings from this systematic review to cohorts with pathologies or injuries or cohorts that are younger than 18 years of age; dynamic COP and PP measures; and isokinetic and isotonic plantar-flexion strength measures is not certain. 6. Conclusion The results from this systematic review highlight the need for better quality methodological reliability studies to make strong inferences about the reliability of measure of COP, PP, and PFisom. Based on the current evidence, the most reliable measures are COP sway area and path length; PP mean pressure, percentage of body weight distribution, and contact area; and PFisom peak torque and peak force. Clinicians are encouraged to select these measures preferentially in clinical settings, bearing in mind that the absolute reliability of certain of these measures and their typical errors might be suboptimal despite exhibiting strong relative reliability based on correlation coefficients. Declaration of Competing Interest None. References [1] G. Atkinson, A.M. Nevill, Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine, Sports Med. 26 (1998) 217–238. [2] C. Bauer, I. Gröger, R. Rupprecht, K.G. Gaßmann, Intrasession reliability of force platform parameters in community-dwelling older adults, Arch. Phys. Med. Rehabil. 89 (2008) 1977–1982. [3] C. Bauer, I. Gröger, R. Rupprecht, A. Meichtry, C.O. Tibesku, K.-G. Gaßmann, Reliability analysis of time series force plate data of community dwelling older adults, Arch. Gerontol. Geriatr. 51 (2010) e100–e105. [4] J.R. Baxter, S.J. Piazza, Plantar flexor moment arm and muscle volume predict torque-generating capacity in young men, J. Appl. Physiol. 116 (2013) 538–544. [5] M.L. Beach, J. Baron, Regression to the mean: introduction, in: N. Balakrishnan, T. Colton, B. Everitt, W. Piegorsch, F. Ruggeri, J. Teugels (Eds.), Wiley StatsRef: Statistics Reference Online, John Wiley & Sons, Ltd., 2015, pp. 1–4. [6] R. Becerro de Bengoa Vallejo, M.E. Losa Iglesias, J. Zeni, S. Thomas, Reliability and repeatability of the portable EPS-platform digital pressure-plate system, J. Am. Podiatr. Med. Assoc. 103 (2013) 197–203. [7] M.G. Bemben, B.H. Massey, R.A. Boileau, J.E. Misner, Reliability of isometric forcetime curve parameters for men aged 20 to 79 years, J. Strength Cond. Res. 6 (1992) 158–164. [8] M. Bishop, P. Fiolkowski, B. Conrad, D. Brunt, M. Horodyski, Athletic footwear, leg stiffness, and running kinematics, J. Athl. Train. 41 (2006) 387–392. [9] M.G. Carpenter, J.S. Frank, D.A. Winter, G.W. Peysar, Sampling duration effects on centre of pressure summary measures, Gait Posture 13 (2001) 35–40. [10] P. Cavanagh, E. Morag, A. Boulton, M. Young, K. Deffner, S. Pammer, The relationship of static foot structure to dynamic foot function, J. Biomech. 30 (1997) 243–250. [11] J.O. Chang, S.S. Levy, S.W. Seay, D.J. Goble, An alternative to the balance error scoring system: using a low-cost balance board to improve the validity/reliability of sports-related concussion balance testing, Clin. J. Sport Med. 24 (2014) 256–262. [12] L. Chiari, A. Cappello, D. Lenzi, U. Della Croce, An improved technique for the extraction of stochastic parameters from stabilograms, Gait Posture 12 (2000) 225–234.

61

Gait & Posture 75 (2020) 46–62

K. Hébert-Losier and L. Murray [43] D. Lafond, H. Corriveau, R. Hébert, F. Prince, Intrasession reliability of center of pressure measures of postural steadiness in healthy elderly people, Arch. Phys. Med. Rehabil. 85 (2004) 896–901. [44] R. Letz, F. Gerr, Standing steadiness measurements: empirical selection of testing protocol and outcome measures, Neurotoxicol. Teratol. 17 (1995) 611–616. [45] S.S. Levy, K.J. Thralls, S.A. Kviatkovsky, Validity and reliability of a portable balance tracking system, btracks, in older adults, J. Geriatr. Phys. Ther. 41 (2018) 102–107. [46] D. Lin, H. Seol, M.A. Nussbaum, M.L. Madigan, Reliability of COP-based postural sway measures and age-related differences, Gait Posture 28 (2008) 337–342. [47] R. Mani, S. Milosavljevic, S.J. Sullivan, Control of posture during tasks representing common work-related postures–a reliability study, Ergonomics 58 (2015) 980–989. [48] C.G. Mattacola, D.A. Lebsack, D.H. Perrin, Intertester reliability of assessing postural sway using the Chattecx Balance System, J. Athl. Train. 30 (1995) 237–242. [49] K. Mattes, A.H.W. Eldin, N. Schaffert, S. Manzer, Local concentric muscle fatigue of the ankle dorsiflexors and plantar flexors: a reproducibility study, Isokinet. Exerc. Sci. 23 (2015) 87–92. [50] M. Moghadam, H. Ashayeri, M. Salavati, J. Sarafzadeh, K.D. Taghipoor, A. Saeedi, R. Salehi, Reliability of center of pressure measures of postural stability in healthy older adults: effects of postural task difficulty and cognitive load, Gait Posture 33 (2011) 651–655. [51] L.B. Mokkink, H.C.W. de Vet, C.A.C. Prinsen, D.L. Patrick, J. Alonso, L.M. Bouter, C.B. Terwee, COSMIN risk of Bias checklist for systematic reviews of patient-reported outcome measures, Qual. Life Res. 27 (2018) 1171–1179. [52] L.B. Mokkink, C.B. Terwee, E. Gibbons, P.W. Stratford, J. Alonso, D.L. Patrick, D.L. Knol, L.M. Bouter, H.C. De Vet, Inter-rater agreement and reliability of the COSMIN (COnsensus-based Standards for the selection of health status Measurement INstruments) checklist, BMC Med. Res. Methodol. 10 (2010) 1–11. [53] E. Morag, P. Cavanagh, Structural and functional predictors of regional peak pressures under the foot during walking, J. Biomech. 32 (1999) 359–370. [54] A. Moraux, A. Canal, G. Ollivier, I. Ledoux, V. Doppler, C. Payan, J.Y. Hogrel, Ankle dorsi- and plantar-flexion torques measured by dynamometry in healthy subjects from 5 to 80 years, BMC Musculoskelet. Disord. 14 (2013) 1–10. [55] R.R. Neptune, S.A. Kautz, F.E. Zajac, Contributions of the individual ankle plantar flexors to support, forward progression and swing initiation during walking, J. Biomech. 34 (2001) 1387–1398. [56] S. Pajala, P. Era, M. Koskenvuo, J. Kaprio, T. Tormakangas, T. Rantanen, Force platform balance measures as predictors of indoor and outdoor falls in communitydwelling women aged 63-76 years, J. Gerontol. A Biol. Sci. Med. Sci. 63 (2008) 171–178. [57] N. Pinsault, N. Vuillerme, Test–retest reliability of centre of foot pressure measures to assess postural control during unperturbed stance, Med. Eng. Phys. 31 (2009) 276–286.

[58] R.J. Prescott, Editorial: avoid being tripped up by statistics: statistical guidance for a successful research paper, Gait Posture 72 (2019) 240–249. [59] J.A. Raymakers, M.M. Samson, H.J. Verhaar, The assessment of body sway and the choice of the stability parameter(s), Gait Posture 21 (2005) 48–58. [60] A.H. Razak, A. Zayegh, R.K. Begg, Y. Wahab, Foot plantar pressure measurement system: a review, Sensors (Basel, Switzerland) 12 (2012) 9884–9912. [61] P.O. Riley, B.J. Benda, K.M. Gill-Body, D.E. Krebs, Phase plane analysis of stability in quiet standing, J. Rehabil. Res. Dev. 32 (1995) 227–235. [62] B. Rosner, Fundamentals of Biostatistics, 8th ed., (2015) Boston, MA. [63] A. Ruhe, R. Fejer, B. Walker, The test–retest reliability of centre of pressure measures in bipedal static task conditions–a systematic review of the literature, Gait Posture 32 (2010) 436–445. [64] B.R. Santos, A. Delisle, C. Larivière, A. Plamondon, D. Imbeau, Reliability of centre of pressure summary measures of postural steadiness in healthy young adults, Gait Posture 27 (2008) 408–415. [65] M. Schmid, S. Conforto, V. Camomilla, A. Cappozzo, T. D’Alessio, The sensitivity of posturographic parameters to acquisition settings, Med. Eng. Phys. 24 (2002) 623–631. [66] K.M. Slattery, L.K. Wallace, D.J. Bentley, A.J. Coutts, Effect of training load on simulated team sport match performance, Appl. Physiol. Nutr. Metab. 37 (2012) 315–322. [67] G.G. Sleivert, H.A. Wenger, Reliability of measuring isometric and isokinetic peak torque, rate of torque development, integrated electromyography, and tibial nerve conduction velocity, Arch. Phys. Med. Rehabil. 75 (1994) 1315–1321. [68] M.J. Spink, M.R. Fotoohabadi, H.B. Menz, Foot and ankle strength assessment using hand-held dynamometry: reliability and age-related differences, Gerontology 56 (2010) 525–532. [69] J. Swanenburg, E. Bruin, K. Kathrin Favero, D. Uebelhart, T. Mulder, The reliability of postural balance measures in single and dualtasking in elderly fallers and nonfallers, BMC Musculoskelet. Disord. 9 (2008) 1–10. [70] E.-P. Takala, I. Korhonen, E. Viikari-Juntura, Postural sway and stepping response among working population: reproducibility, long-term stability, and associations with symptoms of the low back, Clin. Biomech. (Bristol, Avon) 12 (1997) 429–437. [71] K.L. Taylor, J. Cronin, N.D. Gill, D.W. Chapman, J. Sheppard, Sources of variability in iso-inertial jump assessments, Int. J. Sports Physiol. Perform. 5 (2010) 546–558. [72] R. Topp, A. Mikesky, Reliability of isometric and isokinetic evaluations of ankle dorsi/plantar strength among older adults, Isokinet. Exerc. Sci. 4 (1994) 157–163. [73] S.C. Webber, M.M. Porter, Reliability of ankle isometric, isotonic, and isokinetic strength and power testing in older women, Phys. Ther. 90 (2010) 1165–1175. [74] N. Yalcin, E. Esen, U. Kanatli, H. Yetkin, Evaluation of the medial longitudinal arch: A comparison between the dynamic plantar pressure measurement system and radiographic analysis, Acta Orthop. Traumatol. Turc. 44 (2010) 241–245.

62