Sage Journals: Discover world-class research

Abstract

Background:

External devices are used to manage musculoskeletal pathologies by altering loading of the foot, which could result in altered muscle activity that could have therapeutic benefits.

Objectives:

To establish if evidence exists that footwear, foot orthoses and taping alter lower limb muscle activity during walking and running.

Study design:

Systematic literature review.

Methods:

CINAHL, MEDLINE, ScienceDirect, SPORTDiscus and Web of Science databases were searched. Quality assessment was performed using guidelines for assessing healthcare interventions and electromyography methodology.

Results:

Thirty-one studies were included: 22 related to footwear, eight foot orthoses and one taping. In walking, (1) rocker footwear apparently decreases tibialis anterior activity and increases triceps surae activity, (2) orthoses could decrease activity of tibialis posterior and increase activity of peroneus longus and (3) other footwear and taping effects are unclear.

Conclusion:

Modifications in shoe or orthosis design in the sagittal or frontal plane can alter activation in walking of muscles acting primarily in these planes. Adequately powered research with kinematic and kinetic data is needed to explain the presence/absence of changes in muscle activation with external devices.

Clinical relevance

This review provides some evidence that foot orthoses can reduce tibialis posterior activity, potentially benefitting specific musculoskeletal pathologies.

Keywords

Rocker footwear foot orthoses tibialis posterior electromyography

Background

Musculoskeletal pathologies occur when structures experience more load than they can withstand.¹ If external loads are altered by a therapeutic device, there should be a corresponding change in internal muscle-tendon forces, joint loading, the potential for injury, and the rate and likelihood of healing. For example, foot orthoses (FOs) that decrease loading at the rearfoot could decrease activity of the tibialis posterior (TP) muscle and subsequently reduce strain of the TP tendon, a structure vulnerable to tendinopathy.² Clinicians can influence the forces applied to feet and muscles/tendons using footwear, FOs and taping.

Footwear that may have therapeutic benefits by altering loading of the foot include ‘motion control’ shoes (typically running shoes), and rocker/rollover shoes. While motion control shoes with dual density midsoles reduce calcaneal eversion by 2.77° (p < 0.001, 95% confidence interval (CI) = 1.74° to 3.81°),³ whether these changes impact on muscle function and injury risk is unknown. Under the ‘preferred movement pathway’ theory,⁴ footwear or FOs reduce muscle activity and metabolic demand^4,5 by promoting the path of ‘least resistance’ and reduce injury risk.^4–6 However, muscle activity could also increase to keep foot kinematics within the preferred pathway.⁶ Rocker or rollover shoes have outsoles curved in the sagittal plane and alter the contact area between the shoe and floor, plantar load, external sagittal plane joint moments, and thereafter muscular responses and joint motion.⁷ A recent review, however, found few statistically significant effects of the Masai Barefoot Technology (MBT) shoes on lower limb muscles.⁸ The effect of other rocker and motion control footwear on electromyography (EMG) data has not been reviewed, so the use of these specialised shoes for treatment and injury prevention is unclear.

FOs redistribute plantar pressure, altering external joint moments, internal joint moments (from muscles and connective tissue) and foot motion. Examples include insoles with rearfoot wedges and arch supports,⁹ also known as ‘anti-pronation’ FOs or medial posted FOs. Although FOs reduce peak rearfoot eversion by 2.08°−2.35° (p ⩽ 0.004) depending on their design,³ such small changes may not be clinically meaningful.¹⁰ FOs can change ankle moments^11–13 with peak and mean ankle eversion moments reduced by 1.1% ± 1.1% (p = 0.003) and 2.3% ± 2.1% (p < 0.001) per 2° of medial posting, respectively.¹³ Such changes would alter the requirements of tissues acting antagonistically to the external moments, including muscles. The evidence for changes in muscle function with FOs that alter joint kinematics and kinetics is important in understanding injury risk and tissue repair.

Low-Dye taping is a temporary intervention for conditions supposedly associated with foot pronation or flat-arched feet,^14,15 theoretically applying tension to the skin using tape offloads structures in the medial arch.^3,14,16 Taping has been reported to only reduce foot pronation by a non-significant 1.50° (p = 0.19, 95% CI = −0.73° to 3.73°).³ Plantar sensory stimulation is considered an important difference between FOs and taping since changes in afferent feedback due to tape might alter muscle activation.¹⁴ However, there is no evidence to support this theory.

Prior reviews investigating the effects of footwear, FOs and taping did not compare device effects.^8,16,17 Also, approaches to searching and appraisal of literature were variable and underpin the need for a more comprehensive review. Indeed, limitations of prior studies include low power, inadequate reporting of EMG procedures and low external validity.^16,17 A review of foot posture, FOs and footwear by Murley et al.¹⁷ allowed comparisons to a barefoot control, which is less clinically generalisable than a shod control and the review was broad, including all types of FOs and inserts and all footwear, not just that intended to alter foot biomechanics. The present review includes only FOs with a medial arch profile and/or medial heel/foot wedge in order to improve our understanding of the relationship between medial FOs and muscle activity. Furthermore, the recent review of MBT footwear did not assess the quality of EMG data reporting, limiting our understanding of the strength of the evidence identified.⁸ Several studies on the effect of external devices on EMG have been published since these reviews, some of which have reported detailed EMG methods,¹⁸ thus further justifying an update on the literature consensus. The aim of this systematic review was to investigate the level of evidence from any study design that investigated whether footwear, FOs and taping alter lower limb EMG during walking and running, irrespective of health status.

Methods

Search strategy

A systematic, electronic database search was performed by reviewer J.R. using CINAHL (1982–2017), MEDLINE (1950–2017), ScienceDirect, SPORTDiscus (1985–2017) and Web of Science (1900–2017) in October 2015 and updated in March 2019. The review conformed to the PRISMA guidelines for systematic reviews; however, we were unable to account for biases like publication bias.¹⁹ Searched words are included Table 1. Lines (1)–(3) were combined using ‘AND’ with lines (4) and (5). Additional sources were identified from published reviews and the reference lists of studies that passed the quality screening.

Table 1.

Inputs to the electronic databases (all databases).

Search words
(1) orthot* OR insert OR wedge OR orthos$ OR insole OR skive
(2) foot* OR feet OR shoe$ OR footwear OR motion control shoe OR Nike free OR pronation control OR heel*
(3) tape OR taping OR augmented Dye OR low-dye OR low dye
(4) electromyograph$ OR EMG OR IEMG OR muscle function
(5) Walk* OR run* OR gait OR locomotion or jog*

EMG: electromyography; iEMG: integrated electromyography.

Lines (1)–(3) were combined using ‘AND’ with lines (4) and (5). * denotes a wildcard

Inclusion criteria

The search results were assessed for eligibility based on titles and abstracts of original, full text articles using the following inclusion criteria:

A clearly defined amplitude, timing or frequency EMG outcome measure from muscles of the lower limb.

A fully specified independent variable of any footwear designed with modifications in the shape or material of the sole (including a negative heel, but excluding high heels, ankle braces and ankle destabilisation devices), FOs/insoles (orthosis had a medial arch profile and or medial heel/foot wedge, excluding lateral wedges and ankle-FOs) and taping about the foot/ankle intended to reduce foot pronation (excluding Kinesio taping).

Measures were made during level walking or running.

The footwear, FOs or taping experimental conditions were compared with a shod control condition.

For FOs and taping experimental conditions trials were performed in shoes, not sandals, with all the standard components of a shoe that brace the FOs.

Participants were free from conditions affecting the neurological systems.

Data were analysed from a minimum of three trials per condition.

Full text was published in English, French or German (due to available expertise).

Sample size of n > 1.

Only studies on locomotion were included since major theories on mechanisms of therapeutic effect of external devices relate to gait not standing.²⁰ Studies that only compared the device to barefoot were excluded because EMG amplitude can increase due to shoes alone and FOs versus barefoot (+30% and +30% to 38% respectively in tibialis anterior (TA)).²¹ Articles were excluded if there were less than three trials per condition because without contradictory evidence, this was considered the minimum required for quality data. We did not restrict studies to a specific population as we took a mechanistic approach to understanding potential effects of external devices on muscle activity.

Quality assessment

To maintain quality standards in this systematic review, the articles that met the inclusion criteria were subject to two levels of quality assessment (Table 2), performed independently by reviewers J.R. and E.P. performed independently. After studies were assessed, the two reviewers met to discuss discrepancies. When discrepancies persisted these were discussed with a third reviewer (L.B.) and a final score obtained.

Table 2.

Quality assessment criteria checklists.

Criteria
First stage: EMG reporting
(1)	Surface sensors (shape, material, size, inter-electrode distance)
(2)	Adequate skin preparation
(3)	Fixation of the sensors was described or reference was made to guidelines for sensor placement
(4)	Sensor location was based on SENIAM guidelines or a justified alternative (orientation over the muscle belly was made with consideration of fibre direction and with respect to tendons and the motor point, and inter-electrode distance was reported)
(5)	Appropriate signal processing (including specification where applicable of full or half-wave rectification or window size of RMS)²⁴
(6)	Walking or running velocity was controlled (not just reported)
(7)	Adequate description of normalisation procedure if applicable
Fine-wire EMG studies were excluded if insufficient details of the intramuscular wire electrodes (type and material) and procedures (insertion approach and method of establishing a correct insertion) were reported.²⁴
Second stage
Immediate effects studies
(1)	Statement of an aim/hypothesis
(2)	Participant characteristics reported
(3)	Outcomes described in the introduction or methods
(4)	Device material reported
(5)	Clear description of the main findings
(6)	Variability reported (within written results or figures)
(7)	Actual p values stated (or <0.001)
(8)	Participant blinding (e.g. sham FO)
(9)	Assessor blinding
(10)	Appropriate statistics (including checking data for normality and sphericity where appropriate)
(11)	Control condition a true control
(12)	Randomisation of the order conditions were tested
(13)	Power calculation performed
Additional criteria for studies on the effects of devices over time
(14)	A description of the intervention (including duration)
(15)	Comparable participant characteristics across groups
(16)	Compliance

EMG: electromyography; RMS: root mean square; FO: foot orthoses; SENIAM: surface EMG for non-invasive assessment of muscles.

The first stage focussed on the quality of the EMG methodology based on external standards of reporting,²⁴ plus controlling locomotion velocity (since velocity can affect EMG).²³ Studies scored a 1 or 0 depending on whether the criteria were fulfilled or not, and the results were summated and expressed as a percentage. Studies achieving less than 50% were excluded.

The second stage of assessment was based on a modified subset of a checklist for rating clinical interventions.²⁵ Studies were given 1 or 0 depending on whether each criteria was fulfilled, with the total score expressed as a percentage and studies that scored less than 50% were excluded.

Results

Search results

A flow chart of the selection process from identification to screening, and eligibility and inclusion¹⁹ is presented in Figure 1; six studies were excluded based on EMG quality and four studies were excluded based on study design quality. A total of 31 studies were included, of these, 22 related to footwear and 8 to FOs. Key themes of footwear studies were running shoes, rocker footwear, APOS-Therapy shoes^26,27 and the Reebok EasyTone® shoe. Only one taping study (low-Dye) was identified which passed quality assessment. No study from additional sources met the inclusion criteria. Two studies were translated from German, but did not meet the inclusion criteria. Most studies included healthy, often recreationally active participants, except two studies involving participants with knee pain/knee osteoarthritis,^28,29 one with running related overuse injuries³⁰ and one with Achilles tendinopathy.³¹ Summaries of included studies are in Tables 3 and 4 and excluded studies in the Appendix 1.

Figure 1.

Study selection flow chart.

Table 3.

Summary of studies on the effect of footwear (n = 22) and taping (n = 1) on lower limb muscle activity.

Authors	Participant characteristics	Device	Muscles	Walking or running	Variables	Main findings	QA scores, 1^st and 2^nd stage (%)
Bucheker et al. (2012)³⁸	10 overweight males: 32.0 ± 7.9 years, 1.792 ± 0.058 m, 91.3 ± 7.0 kg	Masai Barefoot Technology, participant’s own shoes as control	BF, MG, VL	Walking (15 m walkway)	Intensity (amplitude) and co-activation indices	• In midstance mean intensities of VL (p < 0.05) and VL and MG co-activation (p < 0.05) increased with MBT • In terminal stance mean intensities of VL (p < 0.05), MG (p < 0.05), and VL and MG co-activation (p < 0.05) increased in MBT	71, 62
Burgess and Swinton (2012)³⁹	23 healthy, recreationally active females: 20.8 ± 1.3 years, 1.654 ± 0.056 m, 62.9 ± 11.9 kg	Barefoot, Fitflop™ and flip flop treadmill walking, stair climbing and zigzag walk around cones	BF, Glut Med, MG, RF	Treadmill walking (1.34 m/s)	Normalised mean RMS	• No significant differences between footwear conditions	86, 62
Chen et al. (2018)⁴⁰	Ten healthy males: 25.58 ± 3.64 years, 1.737 ± 0.02 m, 59.86 ± 3.80 kg	Barefoot, sports shoes (Roshe Run, Nike Inc., Oregon, United States) and flip-flops (flat rubber sole, Flipper, Adidas, Germany)	BF, GM, PL, TA, VL	Walking (10 m walkway)	Co-contraction index	• No significant differences between conditions in co-contraction for any muscle pairs	71, 62
Cheung et al. (2009)³⁶	20 novice F runners, rearfoot pronation >6°, 25.8 ± 3.7 years, BMI: 20.54 ± 1.27 kg·m/2	“Supernova control”, (Adidas), designed to check excessive pronation; “Supernova cushion” (Adidas, control), designed to reduce impact rate	VL, VM	Treadmill running (10 km)	EMG onset timing and median frequency	• VM activated ~5.3% (95% CI 4.5 to 6.1) of a duty cycle earlier than VL with motion control shoe • Neutral shoe: delay in VM activation by ~4.6% (95% CI 3.9 to 5.3) of a duty cycle compared with VL	83, 54
Cheung et al. (2010)³³	20 novice F runners, rearfoot pronation >6°, 25.8 ± 3.7 years, BMI: 20.54 ± 1.27 kg·m/2	“Supernova control”, (Adidas, motion control), designed to check excessive pronation; “Supernova cushion” (Adidas, neutral), designed to reduce impact rate	PL, TA	Treadmill running (10 km)	Normalised RMS and median frequency	• Positive correlations between RMS EMG and running mileage in TA and PL in neutral shoe condition (p < 0.001) • Median frequency dropped in both shoe conditions with mileage, but significantly larger drop in neutral shoe than motion control shoe (p < 0.001 for PL, p = 0.074 for TA)	86, 54
Elkjaer et al. (2011)⁴¹	10 healthy males: 24.5 ± 3.8 years, BMI = 24.03 ± 1.09 kg·m/2	Reebook EasyTone® ET Calibrator; neutral Nike Lunarglide +2 (control)	BF, Glut Max, LG, TA, VL	Treadmill walking	Peaks and integrated	• No significant differences between footwear conditions	57, 62
Forghany et al. (2014)⁷	20 healthy subjects (12 M): 33.1 ± 8.4 years, 1.71 m ± 0.04 m, 68.9 kg ± 12.1, BMI 23.6 ± 4.1 kg·m/2	Rollover shoe, flat control shoe (same leather upper and last as rollover shoe), flat control footwear weighted to equivalent of the rollover shoe and MBT shoe. All insoles were removed and replaced with a 1.2 mm poron insole.	Lateral BF, ES (right), Glut Max, MG, RF, SOL, TA	Walking (10 m course)	Peak EMG and integral of the signal	• Maximum at initial contact for TA: -29% for MBT, -22% for rollover shoe vs. control • iEMG: TA -17% both MBT and rollover vs. control; SOL +13% MBT, +8% rollover, MG +8% for MBT	50, 62
Franklin et al. (2018)⁴²	70 healthy males (age range 20–87 years). YOUNG < 40 years (n = 20),MID > 40 years and < 70 years (n = 30) and OLD>70 years (n = 20)	Minimalist shoe (Product ID: 2169, Two Barefeet Boarding Co.), control shoe (Style Code: 10001, Hobos Womens, Style Code: 50109, Hobos Mens), barefoot and own shoes	MG, PL, TA	Walking	Mean amplitude acros gait cycle and at separate phases of gait cycle (EMG only normalised in graphs)	• Higher amplitude of GM in minimalist shoe and own shoe vs. control in YOUNG and MID group but not old. • Lower amplitude of PL in minimalist shoe vs. own shoe and control in YOUNG and MID group but not old. • Lower amplitude of TA in minimalist shoe vs. own shoe and control across gait cycle and at initial double support • Slower walking speed in minimalist shoe vs. own shoe and control, but less < 5% difference	67, 69
Goryachev et al. (2011)²⁷	14 females with symptomatic bilateral medial compartment knee OA for ⩾ 6 months, 59.9 ± 6.2 years, 1.607 ± 0.06 m, 77.4 ± 8.9 kg	APOS shoes in “functional neutral configuration”, without elements, lateral 1.2 cm, medial 0.8 cm (both elements).	BF, LG, MG, ST, TA, VL, VM	Walking (10 m walkway)	ARV, normalized activity duration and peak	• In less symptomatic leg, almost all muscles varied significantly with COP in at least one phase of stance • In more symptomatic leg, significant differences in ARV across COP configurations for LG in terminal stance, pre-swing and terminal contact, for TA in pre-swing and for VL at contact • Training element of the study did not meet inclusion criteria	50, 69
Goto and Abe (2017)⁴³	17 females (19.3 ± 0.9 years, 1.577 ± 0.04 m, 20.2 ± 1.8 kg/m2)	Ladies leather safety footwear (670 g; longitudinal stiffness 35.8 N; MIDORI ANZEN Co., Ltd., Tokyo, Japan, hard sole and hard-resin toe cap). Control sports shoes: (470 g; longitudinal stiffness 14.7 N; Bridgestone Corporation, Tokyo, Japan, soft sole, no toe cap).	BF, LG, TA, VL	Treadmill walking	Mean EMG amplitude of safety shoes normalised to amplitude of control shoes	• Significantly higher amplitude of safety shoes for BF (114.3% ± 20.7%, p = 0.01), TA (105.8% ± 10.8%, p = 0.04) and VL (129.5% ± 47.1%, p = 0.02) vs. control (100%). No significant difference in LG amplitude in safety shoes (103.3% ± 7.7%, p = 0.09) vs. control (100%)	83, 69
Horsak and Baca (2013)⁴⁴	7 M, 5 F: 25 ± 4 years; 1.72 ± 0.11 m; 67 ± 11 kg	Reebok Easy Tone® (Reenew model), 2 weeks familiarisation. Participant’s own shoes as control	VL, VM	Walking	Mean amplitude	• No significant differences between footwear conditions	67, 54
Horsak et al. (2015)⁴⁵	Reanalysed data: 7 M, 5 F; 25 ± 6 years; 1.74 ± 0.07 m; 68 ± 10 kg and 7 M, 5 F: 25 ± 4 years; 1.72 ± 0.11 m; 67 ± 11 kg	Reebok Easy Tone® (Reenew model), 2 weeks familiarisation and MBT shoe. Participant’s own shoes as control	BF, Glut Med, MG, PL, TA, VM, VL	Walking (10 m walkway)	Mean amplitude, co-contraction indices	• No significant difference in mean muscle activity between unstable shoes and control • Increased co-contraction of vastii and gastrocnemius muscle in MBT, (Cohen’s d 0.5-0.9)	71, 54
Kelly et al. (2010)⁴⁶	13 male, recreational runners 31.7 ± 4.9 years, 1.817 ± 0.046 m, 81.6 ± 5.9 kg	Augmented low Dye taping, control taping and Adidas Response Cushion running shoes	Glut Med, VL, VM	Treadmill running (6 mins)	Peak and average EMG signal amplitude, onset time, and burst duration	• Delayed onset of the EMG signal of all muscles with taping, moderate to large effect size	83, 77
Koyama et al. (2012)⁴⁷	6 healthy males: 26.3 ± 5.3 years; 1.72 ± 0.05 m; 68.0 ± 6.1 kg	Shape-ups (SKECHERS, USA) vs. normal walking shoe	RF, VL, BF, TA, SOL, MG	Treadmill walking at 3, 4, 5, 6, and 7 km/h (3 mins)	Integrated EMG (iEMG) calculated relative to control shoe	• Significantly higher iEMG of MG (6–16%, p < 0.05) and SOL (8–23%, p < 0.01) in Shape-ups across speeds. • Tendency towards higher iEMG in Shape-ups vs. control in RF, VL, BF and TA	86, 54
Nigg et al. (2006)⁴⁸	5 M and 3 F: 28.0 ± 3.6 years, 1.695 m ± 0.064 m, 70.1 ± 7.5 kg. Free of LE pain/ injury 6+ months prior to testing and never used MBT shoe before	Control: Adidas SuperNova running shoe (mass: 358 g). Experimental: MBT (mass: 650 g), rounded shoe-sole design in AP direction.	BF, Glut Med, MG, TA, VM	Walking (lab) and quiet standing	Intensity, wavelet analysis	• No significant differences changes in EMG intensity • Trend for reduced intensity of TA (-26% ± 24%) and BF of (-55% ± 60%), • Trend for increased intensity of MG (+52% ± 82%), VM (+4% ± 13%) and Glut Med (+16% ± 25%)	86, 62
O’Connor and Hamill (2004)³⁵	10 healthy, recreationally active males, 27 ± 5 years, 1.72 ± 0.07 m, 72.6 ± 5.3 kg. Only 4 subjects with full sets of data for TP	EVA with a durometer of 45 (Shore A). Neutral shoes constructed with heel height 2.5 cm. For 8° varus configuration, the medial aspect of the midsole at the heel was 3-cm thick, and lateral aspect 2 cm thick. Dimensions reversed for valgus shoe.	LG, MG, PL, SOL, TA, TP	Running (30 m walkway)	Integrated EMG (iEMG), mean amplitude, onset and offset	• No significant differences between footwear conditions	71, 54
O’Connor et al. (2006)³⁴	10 healthy, recreationally active males, 27 ± 5 years, 1.72 ± 0.07 m, 72.6 ± 5.3 kg. Only 4 subjects with full sets of data for TP. Same subjects and materials as O’Connor and Hamill (2004)	EVA with a durometer of 45 (Shore A). Neutral shoes constructed with heel height 2.5 cm. For 8° varus configuration, the medial aspect of the midsole at the heel was 3-cm thick, and lateral aspect 2 cm thick. Dimensions reversed for valgus shoe.	LG, MG, PL, SOL, TA, TP	Treadmill running (5 mins)	Mean amplitude, onset and offset	• Significantly less mean EMG activity in the TA and SOL in the neutral shoe vs. either wedged shoe • TA amplitude increased 16% in varus (medial wedge) shoe vs. neutral shoe	71, 69
Price et al. (2013)⁵⁰	15 healthy females: 29 ± 6.7 years, 1.671 ± 0.042 m, 62.6 ± 6.9 kg,	Earth sandal (control), FitFlop, Masai Barefoot Technology, Reebok Easy-Tone and Skechers Tone-Ups	BF, MG, PL, RF, SOL, TA	Walking (lab)	Median RMS for phase	• Fitflop, Reebok and Skechers increased PL activity during pre-swing, whereas MBT increased MG and decreased TA activity in loading response and mid-stance • Increased PL activity in loading response in MBT vs. control • SOL activation during midstance was lower in Fitflop and Skechers than MBT and control	86, 77
Price et al. (2014)⁴⁹	15 M: 30 ± 8 years, BMI: 25.9 ± 4.5 kg·m/2; 13 F: 37.8 ± 12.4 years, BMI: 23.0 ± 4.7 kg·m/2	Barefoot. Flip-flop (Havaiana Brazil), EVA midsole. Fit-flop: Walkstar I. for females, Dass for males, multi-density EVA in heel, midfoot and toe. Rubber outsole	PL, TA	Walking (lab)	Amplitude	• No significant differences between footwear conditions	83, 62
Sacco (2012)⁵¹	25 healthy females with no experience of MBT (21.8 ± 3.0 years, 1.610 ± 00.4 m, 52.6 ± 5.3 kg)	Barefoot, MBT (501 g), and standard tennis shoe (Rainha System, Alpargatas, Brazil, 171 g, neutral strike).	BF, Glut Med, LG, TA, VL	Walking (10 m walkway)	Peak, time of peak and integral of the envelope	• Less peak TA amplitude in MBT vs. standard shoe and barefoot (p < 0.01)• Walking with the MBT shoe did not increase muscle activity when compared to walking with the standard shoe	83, 62
Scott et al. (2012)³²	28 adults with flat feet (14 M/F), 21.2 ± 3.8 years, 1.71 ± 0.1 m, 73.3 ± 16.0 kg	Standard flexible shoe (Dunlop Volley), stability running shoe (Nike Air Structure Triax +10, range of features aimed at controlling moderate pronation) and barefoot	MG, TA (surface), PL, TP (fine-wire)	Walking (9 m walkway)	Time of peak amplitude and peak amplitude	• Both styles of footwear increased TA peak amplitude and decreased PL peak amplitude vs. barefoot • Little difference between footwear conditions	83, 69
Sobhani et al. (2013)⁵²	16 healthy runners (8 M/F), 29 ± 9 years, 1.771 ± 0.093 m, 69.8 ± 11 kg	Standard shoe (apex/rolling point 53% of shoe length, proximal to metatarsal region, 467 ± 87 g). Modified rocker shoe (rolling point 65%, 805 ± 157 g)	LG, MG, SOL, TA	Slow running and walking (10 m lab)	Peak and time of peak (%)	• Significant delay of EMG peak, ~2% (p < 0.001) in triceps surae walking with rocker shoes • No change in peak amplitude of triceps surae in running/walking • Peak amplitude of TA increased 20%, 64.7 mV, p < 0.001) walking with rocker shoes	67, 69
Sobhani et al. (2015)³⁰	13 Achilles tendinopathy patients (11 F), 48 ± 14.5 years, 1.72 ± 0.07 m, 77 ± 14 kg. Achilles tendinopathy 4 months to 9 years (mean 22.5 months, median 11.5 months)	Standard shoe (apex/rolling point 53% of shoe length, proximal to metatarsal region, 467 ± 87 g). Modified rocker shoe (rolling point 65%, 805 ± 157 g)	LG, MG, SOL, TA	Slow running and walking (10 m lab)	Peak and time of peak (%)	• Peak activity of TA increased (61.77 µV, 35%) for walking with rocker shoes (p = 0.015) • Delay of ~4% of the gait cycle in time of peak activity of LG (p = 0.001) in running	67, 62

ARV = average rectified value, LE = lower extremity, M = male, F = female, RMS = root mean square, TP = tibialis posterior, TA = tibialis anterior, SOL = soleus, PL = peroneus longus, MG = medial gastrocnemius, LG = lateral gastrocnemius, AT = Achilles tendinopathy, VL = vastus lateralis, VM = vastus medialis, BF = biceps femoris, Glut Med = gluteus medius, Glut Max = gluteus maximus, ES = erector spinae, RF = rectus femoris, ST = semitendinosus, MBT = Masai Barefoot Technology, OA = osteoarthritis. QA = quality assessment, for first and second stage ((number of satisfied criteria/number of applicable criteria)*100).

Table 4.

Summary of studies on the effect of foot orthoses (n = 8) on lower limb muscle activity.

Authors	Participant characteristics	Device	Muscles	Walking or running	Variables	Main findings	QA scores, 1^st and 2^nd stage (%)
Akuzawa et al. (2016)⁵³	10 healthy males: 25 ± 5.0 years, 1.68 ± 0.06 m, 61.5 ± 7.8 kg	Shoe (Calcetto Le3, Asics, Japan), shoe + prefabricated orthosis (Athlete grip7, Winning One Inc., Japan), barefoot	TP, FDL, PL	Walking	Amplitude as %MVC in contact, midstance and propulsion phase	• Significant reduction (p < 0.036) in TP activity in propulsion phase with orthoses relative to barefoot but not relative to shoe • No significant difference in FDL and PL EMG between conditions	57, 62
Baur et al. (2011)²⁹	99 runners with running-related overuse symptoms. 50 M, 49 F. CO: 37.1 ± 8.3 years, 1.74 ± 0.09 m, 68.8 ± 13.6 kg. Ort: 37.3 ± 8.2 years, 1.73 ± 0.09 m, 66.8 ± 11.6 kg	Custom, MLA support (25 mm), a detorsion wedge in the forefoot (lateral post, 3 mm), and a bowl-shaped heel. 8 week intervention	PL	Treadmill running	Activation time and mean amplitude	• Sig (p = 0.001) increase in preactivation amplitude of 22% ± 48% (95% CI = 9%–32%) in OR compared with CO	71, 75
Kelly et al. (2011)⁵⁴	12 male recreational athletes (31.2 ± 3.8 years, 76 ± 3.9 kg, 1.808 ± 0.04 m)	Prefabricated Formthotics (Foot Science International)	MG, PL, TA, VM	Treadmill running	Burst duration and average RMS amplitude	• Lower RMS signal amplitude VM (-13.3%, p < 0.02) and MG (-10.7%, p < 0.05), increased PL burst duration (+14.7%, p < 0.05), running with orthoses	86, 54
Maharaj et al. (2018)²	18 adults with flat feet recruited: 5 F, 13 M (14 included in analysis) 26 ± 5 years, 1.70 ± 0.11 m, 71.3 ± 12.6 kg	Shoe:( Gel Lyte 33, Ascics, Japan), shoe + Custom FO: ¾ length semi-rigid 4 mm polypropylene thermoplastic shell with vinyl covering, 4 mm medial skive at 15° and a 5° extrinsic rear foot post, barefoot	TP	Treadmill walking	Amplitude as % of max at preferred walking speed	• Reduced TP activity with shoe and shoe + FO vs. barefoot in early stance (1-12%) and late stance (19-22%), main effect of condition (p ⩽ 0.01), but no significant difference between shoe and shoe + FO.	100, 54
Mills et al. (2012)²⁸	40 patients with knee pain. 27 mobile (foot): 28.67 ± 6.13 years, 1.696 ± 0.149 m, 71.03 ± 11.97 kg. 13 less mobile: 31.15 ± 4.41 years, 1.71 ± 0.0841 m, 71.15 ± 11.22 kg	Prefabricated EVA FOs with varying hardnesses	BF, Glut Med, MG, RF, SOL, TA, VM, VL	Treadmill jogging (3 min intervals)	Peaks and temporal (offset only reported?)	• Orthoses, regardless of comfort, had no immediate effect on lower limb EMG or kinematics compared with baseline shoe conditions • Moderate difference in VL peak amplitude (p = 0.007) between most and least comfortable orthosis, greatest increase in peak amplitude in least comfortable	71, 69
Murley et al. (2010)⁹	30 adults with flat feet, 21.8 ± 4.3 years, 1.71 ± 0.1 m, 73.3 ± 15.5 kg	Modified prefabricated FO: ¾ length, medial heel wedge under heel, arch support heat-moulded to individual. Custom FO: ¾ length, posted at 20° inverted, heel supported by EVA wedge, plaster cast modifications to contour shell to arch	MG, PL, TA, TP	Walking (9 m walkway)	Time of peak amplitude; (RMS); peak amplitude	• In contact phase TP amplitude decreased with prefabricated orthosis (peak amplitude -19%, p = 0.007; RMS amplitude -22%, p = 0.002) and custom orthosis (peak amplitude -12%, p = 0.001, RMS amplitude -13%, p = 0.001), vs. shoe-only • During midstance/ propulsive phase PL EMG amplitude increased with prefabricated orthosis, vs. shoe-only (peak amplitude +21%, p = 0.024; RMS amplitude +24%, p = 0.019) and custom orthosis (peak amplitude +16%, p = 0.028)	57, 85
Murley and Bird (2006)²¹	Pronated foot type: 10 F, 5 M, 23 ± 5 years, 1.702 ± 0.09 m and 69.9 ± 14.4 kg	3 pairs of rigid custom-made foot orthoses (posted at 0°, 15° and 30° inverted)	MG, PL, SOL, TA	Walking (walkway)	Maximum amplitude as % of MVC, onset	• Increased maximum TA amplitude using shoe only (+30%), 0° (+33%), 15° (+38%) and 30° (+30%) inverted orthoses conditions vs. barefoot (p < 0.01) • PL maximum amplitude increased using the 15° inverted orthosis condition vs. barefoot (+21%, p = 0.04), trend for an increase vs. shoe only	57, 62
Telfer et al. (2013)¹⁸	12 pronated and 12 gender matched controls 29.9 ± 8.7years, 1.71 m ± 0.08, 71.6 ± 10.7 kg	9 variations: level of external rearfoot posting modified from 6° lateral to 10° medial in 2° increments.	BF, LG, MG, PL, SOL, TA, VL, VM	Walking (indoor walkway, length?)	Peaks and means	• No main effects due to posting level • Group effects customised FOs reducing above knee muscle activity in pronated foot types compared to normal foot types (BF mean p = 0.022; VL peak p < 0.001; VM peak p = 0.009; VM mean p = 0.001) • Interaction effect peak MG (p = 0.034) and peak SOL p = 0.015)	100, 85

LE = lower extremity, M = male, F = female, RMS = root mean square, TP = tibialis posterior, TA = tibialis anterior, SOL = soleus, PL = peroneus longus, MG = medial gastrocnemius, LG = lateral gastrocnemius, AT = Achilles tendinopathy, VL = vastus lateralis, VM = vastus medialis, BF = biceps femoris, Glut Med = gluteus medius, ES = erector spinae, RF = rectus femoris, CO = control, FO = foot orthosis, MLA = medial longitudinal arch, EVA = Ethyl Vinyl Acetate. QA = quality assessment, for first and second stage ((number of satisfied criteria/number of applicable criteria)*100).

Quality assessment

The included studies scored 50–100 on EMG quality (mean ± standard deviation (SD): 75 ± 13). Unfulfilled criteria for EMG quality spanned categories 1–6. Almost all studies were deemed to have provided adequate details on normalisation where appropriate, although none as detailed as published recommendations, for example, training to produce a maximum voluntary contraction (MVC).²⁴ Of included studies, 14/31 did not specify EMG sensor fixation and 13/31 did not control velocity.

Based on the second stage of quality assessment, the included studies were generally of moderate quality (scored 54–85, mean ± SD: 64 ± 9). Excluded studies scored 31–46 (mean ± SD: 40 ± 7). Many studies did not report device material, participant and assessor blinding, statistical analysis or power analysis. Variability was not reported in 3/4 of excluded studies. In addition, condition randomisation was absent in 8/31 of the included studies and 2/4 of the excluded studies.

Overview of included studies

Footwear

Running shoes

A stability running shoe with a dual-density medial post and foot bridge had no effect on EMG activity during walking versus a standard flexible shoe for the peroneus longus (PL), medial gastrocnemius (MG), soleus, TA and TP.⁵² However, a motion control running shoe with a dual-density midsole (firmer material on the medial versus lateral side) reduced mean TA and PL activity and delayed fatigue during running⁵⁰ versus a cushioned running shoe. Delayed fatigue was demonstrated by maintained median frequency of TA and PL during a 10-km run.

A shoe with a medial wedge (3 cm thick on the medial side, 2 cm thick on the lateral side) increased mean TA amplitude by 16% during treadmill running versus a neutral shoe,³⁶ but not during overground running.⁴⁶ The activity of the gastrocnemii, soleus, PL and TP were not affected by medial or lateral wedges.^36,46 However, TP EMG data were only available for four participants due to measurement difficulties.^36,46

The relative timing of vastus medialis (VM) and vastus lateralis (VL) activation during running was compared between a motion control running shoe with a dual-density midsole and a cushioning running shoe.⁴⁵ This comparison was made based on the premise that delayed onset of VM with respect to VL is associated with patellofemoral pain.⁴⁵ The authors normalised EMG signals to a ‘duty cycle’ (defined in the animal literature as stance expressed as a percentage of step cycle, that is, stance + swing).³⁵ In the motion control shoe, activation of VM occurred ~5.3% (95% CI = 4.5% to 6.1%) of a duty cycle earlier than VL (during a 10-km run). In contrast, VM activation occurred ~4.6% (95% CI = 3.9% to 5.3%) later than VL in the neutral shoe.⁴⁵ The implications of the findings are limited by the ambiguity of the reporting of the methods.

Rocker footwear

TA amplitude in early stance in walking reduced by ~30%–40% (p < 0.05) with rocker footwear like MBT, which is curved under the heel in the sagittal plane, versus flat heeled conventional footwear.^7,47,55 There was also a trend towards reduced TA EMG intensity when walking in MBT versus a running shoe.⁴⁹ A modified shoe with a forefoot rocker only increased peak TA activity in walking (by 20%–35%, p < 0.001, p = 0.015 respectively), but not in running.^31,44 The shoe did not alter triceps surae (TS) activity in late stance during either walking or running.^31,44

Increased PL activity throughout stance (e.g. 50% at loading response, p = 0.02) has been shown with MBT.⁵⁵ Other footwear in that study, including FitFlops^TM, designed to be unstable in the sagittal plane, increased peroneal activity during pre-swing. However, later work found no difference in PL activity with FitFlops⁵¹ or difference in co-contraction with regular flip-flops.⁴⁸ A minimalist shoe reduced TA activity and increased plantar flexion in early stance relative to control footwear, however walking speed was also slower.³⁴

Changes in TS activity during loading response with rocker footwear are opposite to that of TA.^7,55 The integral of the EMG profile in rocker footwear was 8%–13% (p < 0.05) greater than the control shoe for the soleus and 5.5%–8% for the MG (significant for MBT, p < 0.05, but not other rocker shoe, p > 0.05).⁷ Similarly, integrated EMG of MG was 6%–16% (p < 0.05) higher and 8%–23% (p < 0.01) higher in soleus in a rocker shoe compared with a regular walking shoe in treadmill walking.³⁸ There was also a trend towards increased MG EMG intensity walking in MBT compared with a running shoe in another study.⁴⁹ Activation of MG was unaffected when wearing a FitFlop sandal with a variable density sole.^43,55

Studies recording quadriceps activation during walking have generally found no effect of rocker footwear.^7,47,49,55 However, increased activation of VL and greater co-contraction of vastii and MG across stance was found in MBT.^32,33 Activation of biceps femoris or rectus femoris was unaltered by a FitFlop sandal.⁴³ Stiff soled safety shoes significantly increased VL, biceps femoris and TA activity relative to a soft-soled trainer.⁴¹

APOS-Therapy shoes

APOS-Therapy shoes have adjustable domes on the sole allowing manipulation of centre of pressure (COP) position and external joint moments.^26,27 In females with knee osteoarthritis, a lateral shift in the sole domes and COP reduced averaged TA EMG amplitude in pre-swing versus a neutral dome configuration.²⁸ The EMG amplitude of the lateral gastrocnemius increased with a medial shift in COP and decreased with a lateral shift in COP due to APOS-Therapy shoes (compared with neutral).²⁸

Reebok EasyTone shoe

The Reebok EasyTone shoe, designed to be unstable with balance pods, did not alter muscle activation in walking of any thigh, shank or gluteal muscles.^32,37,39

Foot orthoses

There is some limited evidence that FOs decrease activity of TP in early stance and increase activity of PL in mid-late stance.^9,21 Peak amplitude and root mean square (RMS) amplitude of TP during loading response was shown to reduce by 19% (p = 0.007) and 22% (p = 0.002), respectively, with custom FOs, and 12% (p < 0.001) and 13% (p = 0.001), respectively, with prefabricated FOs.⁹ Whereas PL activity increased in midstance with a prefabricated FOs (peak amplitude + 21%, p = 0.024; RMS amplitude + 24%, p = 0.019) and a custom FOs (peak amplitude + 16%, p = 0.028) compared with a shoe only. Maximum PL amplitude has also been shown to increase in walking by 19% for pronated individuals when wearing 15° inverted FOs versus shoes alone (p < 0.05).²¹ However, PL amplitude does not appear to increase linearly with wedging magnitude.^18,21 Another two studies found TP activity was not significantly different between the footwear and FOs conditions (p > 0.05), although there was a decrease of around 10% (p < 0.05) from barefoot to shod and shod with either a prefabricated or custom FOs, which was not considered clinically generalisable in this review.^2,40 However, in the study that recorded kinematics and kinetics, there was also no effect of FOs on subtalar joint displacement or supination moment relative to the shoe condition.² There was no difference in flexor digitorum longus or PL activity between conditions.⁴⁰

As for TA and TS, most evidence indicates magnitude and timing of activation is unchanged by wearing FOs during walking and running.^18,21,29,53 One study found there was a tendency for FOs to decrease TA activation during walking versus a shoe only (effect size 0.18–0.29, for custom and prefabricated FOs respectively), although the result was not statistically significant.⁹

Activity of PL may also increase with FOs during running. In one study, 99 runners with an overuse injury were assigned to customised FOs or no FOs.³⁰ In treadmill running, there was a significant increase (p = 0.003) in PL pre-activation amplitude (EMG activity prior to foot contact) after 2 months of wearing FOs, but not in the control patient group.³⁰ It is unclear whether change was in barefoot running or the running shoes or both. Another study reported a 14.7% (p < 0.05) greater duration of PL activity (the muscle was active for longer) during running with prefabricated FOs compared with no FOs as well as lower average MG and VM RMS amplitude with FOs versus no FOs.⁵³

Low-dye taping

In the only study included involving taping a significant delay (5%–7%, p = 0.001) in onset times of VM, VL and gluteus medius was found during shod running with low-Dye tape compared with control taping.⁵⁴

Discussion

The aim of this review was to establish if there is evidence that footwear, FOs and taping alter muscle activity of the lower limb during walking and running. The effect of running shoe design, FitFlops sandals and low-Dye taping on muscle activity is unclear, while rocker/rollover shoes appear to affect muscle activity of MG and TS.^7,38,55 There is evidence, albeit limited, that FOs decrease activity of TP in early stance,⁹ which could be beneficial in treating posterior tibial tendon dysfunction (PTTD). Activity of PL may increase in mid-late stance,^9,21 otherwise FOs do not appear to alter EMG of lower limb muscles.^9,18,21,29

Footwear

Running shoes

The effect of running shoe design on muscle activity remains unclear due to uncertainty regarding whether the shoes tested were effective in changing loading. No study investigating the effects of running shoe design on EMG during walking or running collected simultaneous kinematics or kinetics.^45,50,52 Without kinetic data, we cannot determine if the footwear changed loading of the foot, which might sometimes explain the absence of change in EMG. Concurrent collection of kinetic and EMG data would also be useful to establish if the difference between a nil effect of motion control shoes in walking and a reduction in fatigue during running are due to the greater forces in running, foot strike patterns or different shoe properties.^50,52

Similarly, sagittal plane kinematics were not reported in studies involving a medial wedged shoe.^36,46 Wedging could increase ankle plantar flexion and increase demand on TA, potentially explaining the 16% increase in TA amplitude during treadmill running compared with a neutral shoe.³⁶ Perhaps a medial wedged shoe is substantially different to a motion control running shoe with a dual-density midsole if the effects of the wedge are not isolated to the frontal plane.

Rocker footwear

Footwear that shifts the COP anteriorly at heel contact reduces TA amplitude between initial contact and into midstance.^7,47,55 An anterior shift in the COP increased the external dorsiflexion moment, resulting in a more dorsiflexed ankle and less work required from TA to control plantar flexion after initial contact.⁷ Increased external dorsiflexion moment/increased internal plantar flexion moment in early stance would also account for the increases in TS activity with some rocker footwear.^7,38,55 Potentially, the increased PL activity in MBT shoes⁵⁵ is due to the need for PL to contribute to sagittal plane moments. In contrast, the increase in TA activity in walking with the shoe with the modified forefoot rocker^31,44 might be explained by the greater mass and sole thickness of the modified shoe versus the control. As TA is active in swing, greater activity during early stance could result from the greater moment of inertia not the sole curvature.

Rocker footwear have been shown to reduce internal plantar flexion moment in late stance, which could be beneficial for offloading the Achilles tendon when treating Achilles tendinopathy.^8,31,44 However, reduced internal plantar flexion moment in late stance is not necessarily coupled to reduced TS activity in the same phase.^31,44,55 This could be because peak of MG activity is earlier in stance than the peak of the internal plantar flexion moment,⁷ and the energy recoil of the Achilles tendon is in terminal stance. Thus, reduced loading of the Achilles tendon suggested by a reduced internal plantar flexion moment may still be beneficial in treating Achilles tendinopathy.

The curved sole of rocker footwear purportedly reduces contact area with the ground and thus reduces stability. Increased co-activation from TA and TS in early stance with MBT may increase ankle stability to compensate.⁷ The induced instability is assumed to increase movement variability and activate muscles required to maintain balance and control movement.⁴² Greater movement variability could be beneficial in managing chronic injury if it reduced the repetitive loading of injured structures.⁵⁶ Conversely, increased co-activation increases joint loading.⁵⁷ Consequently, the clinical implications of altered muscle activation and reduced internal plantar flexion with rocker footwear remains unestablished. In addition, there was no effect of the modified rocker shoe on pain in individuals with chronic Achilles tendinopathy, and randomised clinical trials are necessary to evaluate the therapeutic effect of rocker footwear.³¹

APOS-Therapy shoes

Reduced TA and increased TS activation with APOS-Therapy shoes could be relevant to those with TS and Achilles injury, anterior compartment syndrome and intermittent claudication, but any implications remain speculative.

Foot orthoses

There is some limited evidence that FOs decrease activity of TP in early stance and increase activity of PL in mid-late stance,^9,21 but otherwise there appears to be a lack of effect of FOs on lower limb muscle activity during walking.^{9,18,21,29,53}

As the primary invertor of the foot TP acts eccentrically during early stance to generate an inversion moment that opposes the external eversion moment, and helps control rearfoot eversion. It also acts concentrically to support foot supination later in stance.⁵⁸ If FOs increase the external inversion moment, they might reduce required internal inversion moments, reducing TP activity. Reduced TP activation could mean less force through the TP tendon which could facilitate healing in pathologies like PTTD. The limited amount of evidence on the effect of FOs on TP is likely because indwelling EMG is required to measure TP activity. Further research with adequate power and concurrent collection of kinematic and kinetic data is needed to relate kinetic and kinematic changes to muscle activation.

A linear dose-response to extrinsic rearfoot posting during walking has been demonstrated in kinematic, kinetic and plantar pressure variables, but without a corresponding effect on any EMG-related muscle activity in the calf muscles (including PL), quadriceps or hamstrings.^13,18 Maximum PL amplitude did increase in walking by 19% for pronated individuals when wearing 15° inverted FOs versus shoes alone, but again without a linear dose-response to magnitude of wedging.²¹ The lack of a dose response to medial rearfoot wedging could infer that the FOs exert their effect on PL due to changes in load under the medial longitudinal arch rather than the rearfoot.²¹ The midfoot is in contact with the ground during midstance and the heel is unloading.⁵⁹ Similarly, in later work by Murley and colleagues, flat-footed participants increased PL activity in midstance with prefabricated FOs (peak amplitude + 21%, p = 0.024; RMS amplitude + 24%, p = 0.019) and custom FOs (peak amplitude + 16%, p = 0.028) compared with a shoe only.⁹ The original authors speculated that increased PL EMG amplitude resulted from the foot being more laterally unstable. If FOs increased the external inversion moment, greater PL EMG activation may be needed to maintain equilibrium. As PL is the antagonist of TP, if FOs reduced TP EMG activity this would possibly be accompanied by increased PL EMG activity. However, TP and PL activity do not necessarily represent equal opposing inversion and eversion moments, respectively, due to additional muscle tendon parameters⁶⁰ and different moment arms.

Although FOs may increase amplitude and duration of PL activity during running,^30,53 the literature is limited by low between-session reliability of PL EMG.^61–63 Reported poor inter-session reliability of EMG data from PL reduces confidence in EMG results collected weeks apart.^61–63 Amplitude of an EMG signal varies not only due to the detection of different motor units, but because of variable skin-electrode impedance between sessions.⁶⁴ Variability in amplitude between sessions could affect the ability to detect changes in duration of muscle activity due to FOs using threshold methods. As measurements were taken in separate sessions without mention of normalisation^30,53 comparing EMG measures could be beyond this technique. In addition, electrode placement in one study³⁰ followed the methods of Winter and Yack⁶⁵ (50% of the distance between the fibular head and lateral malleolus, rather than 25% of the distance recommended by surface EMG for non-invasive assessment of muscles (SENIAM)).²² A distal shift in surface electrode placement over PL of 2 cm increases the presence of crosstalk, likely from TA.⁶⁶ Given that PL is most active in mid-late stance, while TA is active prior to foot contact to enable a dorsiflexed ankle position at initial contact, potentially muscle activity reported as pre-activation of PL was actually crosstalk from TA.

Foot orthoses designed to reduce the external eversion moment at the subtalar joint would theoretically decrease the internal inversion moment required from the invertor muscles limiting eversion. The TA has an inversion moment arm when the foot is inverted, as at initial contact,⁶⁷ therefore FOs that reduce the eversion moment might also reduce TA activity. The conclusion of the review by Murley et al.¹⁷ preceding the work of Telfer et al.^13,18 that FOs may increase activation of the TA should be reconsidered. Studies that found FOs increase TA activity had notable limitations. As Murley et al.¹⁷ identified, a significant increase in EMG activation was not always supported by confidence intervals. Many studies did not simultaneously collect kinematic and kinetic data, so we cannot relate any change (or lack of) in EMG to other changes in biomechanics, or evaluate the intervention in the context of the ‘preferred movement pathway’ theory. For example, an extrinsic medial rearfoot wedge could place the foot in a more plantar flexed position since the heel is lifted in the shoe, perhaps increasing demand on the TA in the sagittal plane after initial contact.³⁶ Without kinematic and kinetic data and with variable changes in EMG, the implications of this finding are limited. As TA is not the principal invertor of the foot and its main role is dorsiflexion, perhaps any effect is too small to detect, or too variable depending upon the action of the other invertor muscles (i.e. posterior leg muscles passing medial to the ankle) and foot position. Also, as the only ankle dorsiflexor, TA function is unlikely to be compromised with more alternative invertor muscles available. Overall, the majority of studies have found FOs do not change TA activity significantly, in some cases FOs may decrease TA activity, but any effect is subtle.

Literature limitations

TP is the largest invertor, but given fine-wire EMG can be challenging few studies have investigated its function, or intrinsic muscle activity. Furthermore, the magnitude of change in muscle activity that is clinically meaningful is unknown, thus significant effects of external devices on EMG does not reveal clinically beneficial effects. EMG is only a measure of electrical activity not force production, nor mechanical work in the muscle-tendon unit. In addition, differences in electrode types, signal processing, normalisation and outcome variables make establishing a consensus regarding the meaning of changes in EMG difficult. Guidelines describe methods of EMG processing, but are not a universal best practice.²⁴

Research investigating FOs used various materials and designs, and the descriptions of FOs were limited (no excluded study provided detail on this criteria). Studies used a mixture of customised and prefabricated FOs and both FOs with modifications only in the rearfoot and FOs with additional modifications in the arch and forefoot. Whether isolated modifications in specific FOs geometry could lead to specific changes in EMG is unclear. In addition several studies may have also been inadequately powered. A final observation is that studies generally focus on the immediate effect of external devices, yet muscle function could change over time. Longitudinal EMG studies are difficult, but other approaches such as muscle morphology have proven sensitive to footwear.^68,69

Review limitations

While the quality assessment allowed the review to be based on studies of at least moderate quality, failing criteria could reflect inadequacy in reporting and not whether appropriate procedures were followed.¹⁷ In addition, database searches from one reviewer and the score of 50% as a threshold for inclusion could be considered subjective. Furthermore, each criteria was given equal weighting when some could be more influential than others. For instance blinding might be unrealistic, as so-called sham FOs can exert mechanical effects⁷⁰ and potentially influence EMG. Conversely, lack of randomisation could invalidate results due to an order effect and be grounds for exclusion alone. Nonetheless, outcomes of the excluded studies were largely in agreement with those included, except one which found a significantly longer duration (p < 0.05) of TA activity following foot contact with FOs versus control.⁷¹

Footwear outside the inclusion criteria could alter loading of the foot and subsequent muscle activity. However a general review of footwear would be far broader and by restricting our search to footwear that aims to alter foot/ankle motion with modifications in sole construction, findings can be more directly related to the other devices reviewed. This review focused on muscle activation, however devices could have other effects on soft tissue, like the capacity of the series elastic element of TP to absorb energy in early stance.⁵⁸

The review included studies with heterogeneous injury status and foot postures. The response to an intervention may vary with pathology. However the evidence that Achilles tendinopathy, for example, alters muscle activation is conflicting.^72,73 Few of the studies included patient populations and those that did not provide healthy controls and foot posture was often not reported. Consequently sub-group comparisons were not possible.

Conclusion

Modifications in shoe or FOs design in the sagittal or frontal plane can alter activation in walking of muscles acting primarily in these planes. Adequately powered research with kinematic and kinetic data is needed to explain the presence/absence of changes in muscle activation with external devices.

Supplemental Material

Quality_assessment_supplementary_data – Supplemental material for A systematic review of the effect of footwear, foot orthoses and taping on lower limb muscle activity during walking and running

Supplemental material, Quality_assessment_supplementary_data for A systematic review of the effect of footwear, foot orthoses and taping on lower limb muscle activity during walking and running by Joanna Reeves, Richard Jones, Anmin Liu, Leah Bent, Emma Plater and Christopher Nester in Prosthetics and Orthotics International

Footnotes

Appendix

Appendix 1.

Excluded studies.

Authors	Participant characteristics	Device	Muscles	Walking or running	Variables	Main findings	QA score (%), First stage; second stage
Excluded at first stage
Bourgit et al.¹	12 healthy females: 24 ± 4 years, 1.67 ± 0.06 m, 60 ± 7 kg	Standard fitness shoe (Reebok Revent Mid DMX, around 4° plantar flexion) and dorsiflexion shoes with a curvature plate in the midsole: Springboost B-Fit 2°dorsiflexion, Springboost B-Fit 4° dorsiflexion and Meridian 10° dorsiflexion	BF, Glut Max, MG, LG, TA, RF, SOL, VL, VM	Treadmill walking and running	RMS as %MVC and integrated EMG (iEMG)	• Significant increase in plantar flexor RMS EMG (average of MG, LG and SOL) with dorsiflexion shoes vs standard shoe in walking and running	43; 54
Branthwaite et al.²	Males and females, n = 15 (age = 25.2 ± 5.2 years; height = 1.68 ± 0.07 m; weight = 66 ± 10.6 kg)	Masai Barefoot Technology, participant’s own shoes as control	SOL, MG, LG, TA, PL, RF, BF and Glut Med	Walking (10 m walkway)	Mean, maximum EMG (mV) and percentage values (tip toe exercise presumably barefoot)	• No significant difference between conditions. Large inter-individual differences in muscle activity	43; 54
Demura and Demura³	10 healthy males: 24.1 ± 4.1 years, 1.72 ± 0.05 m, 69.2 ± 8.4 kg	Stretch Walker (Nosaka, Ltd, Japan), Masai Barefoot Technology and flat-bottomed shoe	Glut Med, MG, LG, ST, SOL, TA, vastus tibialis?	Walking (laps of ~40 m walkway)	Integrated EMG	• No significant difference between conditions	33; 54
Gu et al.⁴	22 healthy males: 23.5 ± 1.3 years, 1.73 ± 0.01 m, 66 ± 2.4 kg	Flat-soled control shoe, experimental shoe with adjustable hemispheres of 1.5 cm and diameter of 5.5 cm in the heel and forefoot (neutral, medial and lateral configuration)	BF, MG, LG, RF, TA, PL, VL, VM	Walking (10 m walkway)	RMS	• Increase in LG, PL and TA activity in experimental shoes vs control	33; 54
Ritchie et al.⁵	21 males: 21 ± 4 years, 1.768 ± 0.05 m, 73.3 ± 6.5 kg	Prefabricated 3/4 length Formthotic (Foot Science International)	MG, PL, TA	Walking (15 m walkway)	Activation ratio with respect to control	• No significant differences	43; 62
Tomaro and Burdett⁶	10 adults with history of leg pathology, treated with orthotics in last 6 months (3 M, 7 F, 25–30 years)	Prefabricated Sporthotic® (Langer Biomechanics, Deer Park, NY), own athletic shoes	LG, PL, TA	Treadmill walking	RMS amplitude, duration of activity	• Significantly (p < 0.05) longer duration of TA activity following heel strike with FO vs control • No significant difference in EMG amplitude between conditions for all muscles	17; 46
Excluded at second stage
Boergers⁷	University track team athletes (n = 8; 5 F, 3 M, 20 ± 1. 1 years) with self-reported anterior shin pain.	Traditional arch tape Perrin (1995), experimental arch tape (over dorsal rather than plantar surface) and presumably barefoot	TA	Walking and running	Peaks and means	• No significant difference between conditions	71; 46
Burke and Papuga⁸	6 endurance-trained recreational runners (3 M, 3 F; 32.3 ± 10.07 years)	Flexible, custom-made UltraStep orthotics (Foot Levellers, Inc., Roanoke, VA)	BF, MG, TA, VL	Treadmill running	Peaks	• No significant difference between conditions	57; 31
Romkes et al.⁹	12 healthy adults: 6 M, 6 F, 38.6 ± 13.2 years, 1.73 ± 0.06 m, 77.4 ± 12.3 kg	Masai Barefoot Technology and own shoe	MG, LG, RF, ST, TA, VL, VM	Walking	Amplitude as % of peak in barefoot	• TA activity decreased in early stance and increased in swing MBT vs control (p < 0.05) • MG and LG increased activity from late swing to mid-stance MBT vs control (p < 0.05) • VM and VL increased activity mid-stance to toe-off MBT vs control (p < 0.05) • RF increased activity in mid-stance and decreased activity in early swing MBT vs control (p < 0.05)	86; 38
Stöggl et al.¹⁰	12 healthy students: 6 M, 6 F, 25 ± 2 years, 1.72 ± 0.07 m, 65 ± 9 kg	Masai Barefoot Technology and running shoe (Adidas SuperNova. Herzogenaurach, Germany), before and after 10-week training intervention	BF, MG, TA, PL, VL, VM	Treadmill walking	Variability (SD over 15 consecutive gait cycles) for: RMS, integrated EMG (iEMG) and median power frequency	• No difference between shoes and no change by training in EMG variability	100; 44

BF: biceps femoris; MG: medial gastrocnemius; LG: lateral gastrocnemius; TA: tibialis anterior; RF: rectus femoris; SOL: soleus; VL: vastus lateralis; VM: vastus medialis; RMS: root mean square; MVS: maximum voluntary contraction; EMG: electromyography; PL: peroneus longus; Glut Med: gluteus medius; FO: foot orthosis; M: male; F: female; ST: semitendinosus; SD: standard deviation; MBT: Masai Barefoot Technology.

QA: quality assessment, second stage ((number of satisfied criteria/number of applicable criteria) × 100).

Acknowledgements

We wish to thank Dr. Henrike Greaves (Liverpool John Moores University) for her assistance in translating articles to English from German.

Associate Editor: Nerrolyn Ramstrand

Author contributions

The review was conducted by JR. The preparation of the manuscript was supervised by CN. All authors were involved in the drafting and approving of the manuscript.

Declaration of conflicting interests

CN owns equity in Salfordinsole Healthcare Ltd. (Nuneaton, UK) that manufactures foot orthoses. Other authors have no conflicts of interest to declare.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iDs

Joanna Reeves

Emma Plater

Supplemental material

Supplemental Material is available online for this article.

References

Payne

. Forces, motion and outcomes with foot orthoses and running shoes. In: Frederick

Yang

(eds) 8th footwear biomechanics symposium. Taipei, Taiwan, 2007, pp. 23–24. Taipei, Taiwan: The Footwear Biomechanics Group.

Maharaj

Cresswell

Lichtwark

. The immediate effect of foot orthoses on subtalar joint mechanics and energetics. Med Sci Sports Exerc 2018; 50(7): 1449–1456.

Cheung

RTH

Chung

RCK

. Efficacies of different external controls for excessive foot pronation: a meta-analysis. Br J Sports Med 2011; 45(9): 743–751.

Nigg

. The role of impact forces and foot pronation: a new paradigm. Clin J Sport Med 2001; 11(1): 2–9.

Nigg

Vienneau

Smith

, et al. The preferred movement path paradigm: influence of running shoes on joint movement. Med Sci Sports Exerc 2017; 49(8): 1641–1648.

Nigg

Mohr

Nigg

. Muscle tuning and preferred movement path-a paradigm shift. CISS, 2017; 2: 007.

Forghany

Nester

Richards

, et al. Rollover footwear affects lower limb biomechanics during walking. Gait Post 2014; 39(1): 205–212.

Tan

Auhl

Menz

, et al. The effect of Masai Barefoot Technology (MBT) footwear on lower limb biomechanics: a systematic review. Gait Post 2016; 43: 76–86.

Murley

Landorf

Menz

. Do foot orthoses change lower limb muscle activity in flat-arched feet towards a pattern observed in normal-arched feet. Clin Biomech (Bristol, Avon) 2010; 25(7): 728–736.

10.

Chevalier

Chockalingam

. Foot orthoses: a review focusing on kinematics. J Am Podiatr Med Assoc 2011; 101(4): 341–348.

11.

McMillan

Payne

. Effect of foot orthoses on lower extremity kinetics during running: a systematic literature review. J Foot Ankle Res 2008; 1(1): 13.

12.

Nester

van der Linden

Bowker

. Effect of foot orthoses on the kinematics and kinetics of normal walking gait. Gait Post 2003; 17(2): 180–187.

13.

Telfer

Abbott

Steultjens

MPM

, et al. Dose response effects of customised foot orthoses on lower limb kinematics and kinetics in pronated foot type. J Biomech 2013; 46(9): 1489–1495.

14.

Bishop

Arnold

May

. Effects of taping and orthoses on foot biomechanics in adults with flat-arched feet. Med Sci Sports Exerc 2016; 48(4): 689–696.

15.

Radford

Burns

Buchbinder

, et al. The effect of low-dye taping on kinematic, kinetic, and electromyographic variables: a systematic review. J Orthop Sports Phys Ther 2006; 36(4): 232–241.

16.

Franettovich

Chapman

Blanch

, et al. A physiological and psychological basis for anti-pronation taping from a critical review of the literature. Sports Med 2008; 38(8): 617–631.

17.

Murley

Landorf

Menz

, et al. Effect of foot posture, foot orthoses and footwear on lower limb muscle activity during walking and running: a systematic review. Gait Post 2009; 29(2): 172–187.

18.

Telfer

Abbott

Steultjens

, et al. Dose–response effects of customised foot orthoses on lower limb muscle activity and plantar pressures in pronated foot type. Gait Post 2013; 38(3): 443–449.

19.

Liberati

Altman

Tetzlaff

, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009; 151(4): W65.

20.

Harradine

Bevan

. A review of the theoretical unified approach to podiatric biomechanics in relation to foot orthoses therapy. J Am Podiatr Med Assoc 2009; 99(4): 317–325.

21.

Murley

Bird

. The effect of three levels of foot orthotic wedging on the surface electromyographic activity of selected lower limb muscles during gait. Clin Biomech (Bristol, Avon) 2006; 21(10): 1074–1080.

22.

Hermens

Freriks

Disselhorst-Klug

, et al. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 2000; 10(5): 361–374.

23.

Murley

Menz

Landorf

. Electromyographic patterns of tibialis posterior and related muscles when walking at different speeds. Gait Post 2014; 39(4): 1080–1085.

24.

Merletti

Di Torino

. Standards for reporting EMG data. J Electromyogr Kinesiol 1999; 9(1): 3–4.

25.

Downs

Black

. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health 1998; 52(6): 377–384.

26.

Haim

Rozen

Dekel

, et al. Control of knee coronal plane moment via modulation of center of pressure: a prospective gait analysis study. J Biomech 2008; 41(14): 3010–3016.

27.

Haim

Rozen

Wolf

. The influence of sagittal center of pressure offset on gait kinematics and kinetics. J Biomech 2010; 43(5): 969–977.

28.

Goryachev

Debbi

Haim

, et al. Foot center of pressure manipulation and gait therapy influence lower limb muscle activation in patients with osteoarthritis of the knee. J Electromyogr Kinesiol 2011; 21(5): 704–711.

29.

Mills

Blanch

Vicenzino

. Comfort and midfoot mobility rather than orthosis hardness or contouring influence their immediate effects on lower limb function in patients with anterior knee pain. Clin Biomech (Bristol, Avon) 2012; 27(2): 202–208.

30.

Baur

Hirschmuller

Muller

, et al. Neuromuscular activity of the peroneal muscle after foot orthoses therapy in runners. Med Sci Sports Exerc 2011; 43(8): 1500–1506.

31.

Sobhani

Zwerver

van den Heuvel

, et al. Rocker shoes reduce Achilles tendon load in running and walking in patients with chronic Achilles tendinopathy. J Sci Med Sport 2015; 18(2): 133–138.

32.

Horsak

Heller

Baca

. Muscle co-contraction around the knee when walking with unstable shoes. J Electromyogr Kinesiol 2015; 25(1): 175–181.

33.

Buchecker

Wagner

Pfusterschmied

, et al. Lower extremity joint loading during level walking with Masai barefoot technology shoes in overweight males. Scand J Med Sci Sports 2012; 22(3): 372–380.

34.

Franklin

F-X

Grey

. Modifications in lower leg muscle activation when walking barefoot or in minimalist shoes across different age-groups. Gait Post 2018; 60: 1–5.

35.

Liu

, et al. Quantitative gait analysis of long-term locomotion deficits in classical unilateral striatal intracerebral hemorrhage rat model. Behav Brain Res 2013; 257: 166–177.

36.

O’Connor

Price

Hamill

. Examination of extrinsic foot muscles during running using mfMRI and EMG. J Electromyogr Kinesiol 2006; 16(5): 522–530.

37.

Horsak

Baca

. Effects of toning shoes on lower extremity gait biomechanics. Clin Biomech (Bristol, Avon) 2013; 28(3): 344–349.

38.

Koyama

Naito

Ozaki

, et al. Effects of unstable shoes on energy cost during treadmill walking at various speeds. J Sports Sci Med 2012; 11(4): 632–637.

39.

Elkjaer

Kromann

Larsen

, et al. EMG analysis of level and incline walking in Reebok EasyTone ET Calibrator. In: Dremstrup

Rees

Jensen

(eds) 15th Nordic-Baltic conference on biomedical engineering and medical physics. New York: Springer, 2011, pp. 109–112.

40.

Akuzawa

Imai

Iizuka

, et al. Calf muscle activity alteration with foot orthoses insertion during walking measured by fine-wire electromyography. J Phys Ther Sci 2016; 28(12): 3458–3462.

41.

Goto

Abe

. Gait characteristics in women’s safety shoes. Appl Ergon 2017; 65: 163–167.

42.

Nigg

Federolf

von Tscharner

, et al. Unstable shoes: functional concepts and scientific evidence. Footwear Sci 2012; 4(2): 73–82.

43.

Burgess

Swinton

. Do Fitflops™ increase lower limb muscle activity? Clin Biomech 2012; 27(10): 1078–1082.

44.

Sobhani

Hijmans

van den Heuvel

, et al. Biomechanics of slow running and walking with a rocker shoe. Gait Post 2013; 38(4): 998–1004.

45.

Cheung

RTH

. Motion control shoe affects temporal activity of quadriceps in runners. Br J Sports Med 2009; 43(12): 943–947.

46.

O’Connor

Hamill

. The role of selected extrinsic foot muscles during running. Clin Biomech (Bristol, Avon) 2004; 19(1): 71–77.

47.

Sacco

ICN

Sartor

Cacciari

, et al. Effect of a rocker non-heeled shoe on EMG and ground reaction forces during gait without previous training. Gait Post 2012; 36(2): 312–315.

48.

Chen

Wong

, et al. Lower limb muscle co-contraction and joint loading of flip-flops walking in male wearers. PLoS ONE 2018; 13(3): e0193653.

49.

Nigg

Hintzen

Ferber

. Effect of an unstable shoe construction on lower extremity gait characteristics. Clin Biomech (Bristol, Avon) 2006; 21(1): 82–88.

50.

Cheung

RTH

. Motion control shoe delays fatigue of shank muscles in runners with overpronating feet. Am J Sports Med 2010; 38(3): 486–491.

51.

Price

Andrejevas

Findlow

, et al. Does flip-flop style footwear modify ankle biomechanics and foot loading patterns. J Foot Ankle Res 2014; 7(1): 40.

52.

Scott

Murley

Wickham

. The influence of footwear on the electromyographic activity of selected lower limb muscles during walking. J Electromyogr Kinesiol 2012; 22(6): 1010–1016.

53.

Kelly

Girard

Racinais

. Effect of orthoses on changes in neuromuscular control and aerobic cost of a 1-h run. Med Sci Sports Exerc 2011; 43(12): 2335–2343.

54.

Kelly

Racinais

Tanner

, et al. Augmented low dye taping changes muscle activation patterns and plantar pressure during treadmill running. J Orthop Sports Phys Ther 2010; 40(10): 648–655.

55.

Price

Smith

Graham-Smith

, et al. The effect of unstable sandals on instability in gait in healthy female subjects. Gait Post 2013; 38(3): 410–415.

56.

Hamill

van Emmerik

Heiderscheit

, et al. A dynamical systems approach to lower extremity running injuries. Clin Biomech (Bristol, Avon) 1999; 14(5): 297–308.

57.

Potthast

Lersch

Segesser

, et al. Intraarticular pressure distribution in the talocrural joint is related to lower leg muscle forces. Clin Biomech (Bristol, Avon) 2008; 23(5): 632–639.

58.

Maharaj

Cresswell

Lichtwark

. The mechanical function of the tibialis posterior muscle and its tendon during locomotion. J Biomech 2016; 49(14): 3238–3243.

59.

Williams

Nester

. CHAPTER 1 – principles of foot biomechanics and gait. In: Williams

Nester

(eds) Pocket podiatry: footwear and foot orthoses. Edinburgh: Churchill Livingstone, 2010, pp. 1–14.

60.

Murley

Buldt

Trump

, et al. Tibialis posterior EMG activity during barefoot walking in people with neutral foot posture. J Electromyogr Kinesiol 2009; 19(2): e69–e77.

61.

Barn

Rafferty

Turner

, et al. Reliability study of tibialis posterior and selected leg muscle EMG and multi-segment foot kinematics in rheumatoid arthritis associated pes planovalgus. Gait Post 2012; 36(3): 567–571.

62.

Moisan

Cantin

. Effects of two types of foot orthoses on lower limb muscle activity before and after a one-month period of wear. Gait Post 2016; 46: 75–80.

63.

Murley

Menz

Landorf

, et al. Reliability of lower limb electromyography during overground walking: a comparison of maximal- and sub-maximal normalisation techniques. J Biomech 2010; 43(4): 749–756.

64.

Burden

. Surface electromyography. In: Payton

Bartlett

(eds) Biomechanical evaluation of movement in sport and exercise: the British Association of Sport and Exercise Sciences guide. London: Routledge, 2007, pp. 77–102.

65.

Winter

Yack

. EMG profiles during normal human walking: stride-to-stride and inter-subject variability. Electroencephalogr Clin Neurophysiol 1987; 67(5): 402–411.

66.

Campanini

Merlo

Degola

, et al. Effect of electrode location on EMG signal envelope in leg muscles during gait. J Electromyogr Kinesiol 2007; 17(4): 515–526.

67.

Lee

SSM

Piazza

. Inversion-eversion moment arms of gastrocnemius and tibialis anterior measured in vivo. J Biomech 2008; 41(16): 3366–3370.

68.

Bruggemann

G-P

Potthast

Braunstein

, et al. Effect of increased mechanical stimuli on foot muscles functional capacity. In: Proceedings of the ISB XXTH Congress-ASB 29th annual meeting, Cleveland, OH, 31 July–5 August 2005.

69.

Johnson

Myrer

Mitchell

, et al. The effects of a transition to minimalist shoe running on intrinsic foot muscle size. Int J Sports Med 2016; 37(2): 154–158.

70.

McCormick

Bonanno

Landorf

. The effect of customised and sham foot orthoses on plantar pressures. J Foot Ankle Res 2013; 6: 19.

71.

Tomaro

Burdett

. The effects of foot orthotics on the EMG activity of selected leg muscles during gait. J Orthop Sports Phys Ther 1993; 18(4): 532–536.

72.

Baur

Muller

Hirschmuller

, et al. Comparison in lower leg neuromuscular activity between runners with unilateral mid-portion Achilles tendinopathy and healthy individuals. J Electromyogr Kinesiol 2011; 21(3): 499–505.

73.

Wyndow

Cowan

Wrigley

, et al. Triceps surae activation is altered in male runners with Achilles tendinopathy. J Electromyogr Kinesiol 2013; 23(1): 166–172.

74.

Bourgit

Millet

Fuchslocher

. Influence of shoes increasing dorsiflexion and decreasing metatarsus flexion on lower limb muscular activity during fitness exercises, walking, and running. J Strength Condition Res 2008; 22(3): 966–973.

75.

Branthwaite

Chockalingam

Pandyan

, et al. Evaluation of lower limb electromyographic activity when using unstable shoes for the first time: a pilot quasi control trial. Prosthet Orthot Int 2012; 37(4): 275–281.

76.

Demura

S-I

. The effects of shoes with a rounded soft sole in the anterior–posterior direction on leg joint angle and muscle activity. The Foot 2012; 22(3): 150–155.

77.

Mei

, et al. Effects of different unstable sole construction on kinematics and muscle activity of lower limb. Hum Mov Sci 2014; 36: 46–57.

78.

Ritchie

Paterson

Bryant

, et al. The effects of enhanced plantar sensory feedback and foot orthoses on midfoot kinematics and lower leg neuromuscular activation. Gait Post 2011; 33(4): 576–581.

79.

Tomaro

Burdett

. The effects of foot orthotics on the EMG activity of selected leg muscles during gait. J Orthop Sports Phys Ther 1993; 18(4): 532–536.

80.

Boergers

. Effects of arch taping on peak force, surface contact area, and neuromuscular activity at midstance, Master’s Thesis, University of Oregon, Eugene, OR, 2000.

81.

Burke

Papuga

. Effects of foot orthotics on running economy: methodological considerations. J Manipu-lative Physiol Ther 2012; 35(4): 327–336.

82.

Romkes

Rudmann

Brunner

. Changes in gait and EMG when walking with the Masai Barefoot Technique. Clin Biomech 2006; 21(1): 75–81.

83.

Stöggl

Haudum

Birklbauer

, et al. Short and long term adaptation of variability during walking using unstable (Mbt) shoes. Clin Biomech 2010; 25(8): 816–822.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.16 MB

A systematic review of the effect of footwear,foot orthoses and taping on lower limb muscle activity during walking and running

Abstract

Background:

Objectives:

Study design:

Methods:

Results:

Conclusion:

Clinical relevance

Keywords

Background

Methods

Search strategy

Inclusion criteria

Quality assessment

Results

Search results

Quality assessment

Overview of included studies

Footwear

Running shoes

Rocker footwear

APOS-Therapy shoes

Reebok EasyTone shoe

Foot orthoses

Low-dye taping

Discussion

Footwear

Running shoes

Rocker footwear

APOS-Therapy shoes

Foot orthoses

Literature limitations

Review limitations

Conclusion

Supplemental Material

Quality_assessment_supplementary_data – Supplemental material for A systematic review of the effect of footwear, foot orthoses and taping on lower limb muscle activity during walking and running

Footnotes

Appendix

Acknowledgements

Author contributions

Declaration of conflicting interests

Funding

ORCID iDs

Supplemental material

References

Supplementary Material