Walking in high-heel shoes induces redistribution of joint power and work

ABSTRACT

Walking in high-heel shoes (HHS) decreases the push-off power and little research has examined the specific muscle groups that compensate for it. The purpose was to examine the effects of walking in HHS compared to barefoot on lower extremity net joint work and power. Fourteen young women walked in HHS and barefoot at a fixed speed of 1.3 m·s⁻¹. Marker position and ground reaction force data were synchronously measured at 100 and 1000 Hz, respectively. Peak power and joint work variables were computed over the power phases of the gait cycle using an inverse dynamic approach. When walking in HHS was compared to barefoot, participants exerted a diminished push-off characterized by lesser peak power and lesser work by the ankle plantar flexors in late stance (A2 phase; p < 0.001). To compensate for the reduced ankle plantar flexor power, greater peak power was generated and work was performed in early stance by hip extensors (H1 phase; p ≤ 0.001), in mid-stance by knee extensors (K2 phase; p < 0.001) and in late stance and early swing phase by hip flexor muscles (H3 phase; p ≤ 0.001). Walking in HHS induces biomechanical plasticity and causes distal-to-proximal redistribution of net joint power and work during walking.

KEYWORDS:

• Plasticity
• gait
• plantar flexor
• push-off

Introduction

Millions of women currently wear high-heeled shoes (HHS) daily. The structure of HHS places the ankle in a more plantar flexed position compared to normal walking (Wang et al., 2016) and further raises the center of mass of the body above the ground, decreasing stability and increasing fall risk (Tencer et al., 2004; Sun et al., 2017). During walking, the plantar flexors make the highest contribution to energy generation needed for propulsion and maintaining upright posture (DeVita and Hortobagyi 2000; Silder et al., 2008; Cofre et al., 2011; Buddhadev and Martin 2016). However, the plantar flexed position characteristic of HHS walking places the gastrocnemius, soleus and associated non-contractile elements in a shortened, less efficient state, which necessitates compensation from muscles about other lower extremity joints (Farrag and Elsayed 2016).

Recent examination of the biomechanical consequences of wearing HHS has revealed important kinematic and kinetic alterations in walking gait. Both preferred walking speed and stride length decrease with increasing heel height, while the foot flat phase and stance time durations increase (Opila-Correia 1990; de Lateur BJ et al., 1991; Titchenal et al., 2015). Ankle plantar flexion both at initial contact and at maximum increases with increasing heel height (Ebbeling et al., 1994), although the overall plantar flexion range of motion (ROM) decreases (Wang et al., 2016). Hip and knee flexion increases at initial contact and throughout stance in HHS versus barefoot. Increased hip and knee flexion motion could be occurring with increasing heel height, potentially to compensate for ankle being in a more plantar flexed position and to absorb impact loads (Opila-Correia 1990; Ebbeling et al., 1994; Esenyel et al., 2003).

Examining the lower extremity kinetic changes can lend further insight into possible mechanisms underlying the observed kinematic changes in HHS gait. Increased maximal vertical and braking ground reaction forces with increasing heel heights characterize HHS gait (Ebbeling et al., 1994; Hong et al., 2005). In addition, plantar pressure has been shown to shift toward the forefoot in HHS (Hong et al., 2005). Simonsen et al., (2012) reported a greater ankle dorsiflexor moment just after heel strike, moving toward foot flat in HHS. Conversely, ankle plantar flexor moment is reportedly lower both at its peak (Kerrigan et al., 2001) and during push-off in HHS (Esenyel et al., 2003; Simonsen et al., 2012). This decreased ankle plantar flexor moment requires an increased knee extensor moment in walking, which is supported by the finding of greater peak patellofemoral joint reaction force, knee extensor moment magnitude and duration during limb loading in HHS (Esenyel et al., 2003; Ho et al., 2012; Simonsen et al., 2012). These kinetic consequences of HHS also result in a greater duration of the peak hip flexor moment prior to toe-off in HHS.

Majority of the current literature has explored HHS walking gait with respect to lower extremity net joint moments, but limited data are available related to power generation and work performed by the muscles about the hip, knee and ankle joints across various phases of the gait cycle. These variables can help to further our understanding of contributions of specific muscle groups toward generating and absorbing energy needed for providing propulsive power and maintaining support during walking. For example, in early stance, hip extensor muscles generate positive power to cause hip extension to move the body upward and forward (H1 power phase). During mid-stance, knee extensor muscles generate power to cause knee extension to move the body over the stance foot (K2 power phase). In the late-stance phase, plantar flexors generate positive power to produce a vigorous push-off (A2 power phase). And in terminal stance and early swing phase, hip flexor muscles generate positive power to cause hip flexion and pull the thigh and leg into swing phase (H3 power phase) (Winter 1991; Eng and Winter 1995). Examination of these power phases can reveal which muscle groups are specifically compensating for the reduced push-off power observed with HHS compared to barefoot walking.

Although much research has explored the kinematic and kinetic consequences of walking in HHS, the strategies used to compensate for these changes and to accomplish the task have received limited attention. Specifically, the effect of wearing HHS on the lower extremity muscular power generation and work performed during level walking is sparsely researched. Therefore, the purpose of this study was to examine the effects of walking in HHS compared to barefoot on peak power and work performed across the various phases of the gait cycle to gain an understanding of the adaptations (i.e. biomechanical plasticity) in muscular power and work needed to walk in HHS. We hypothesize that: (1) the push-off characterized by ankle positive power and work performed in late stance (A2 power generation phase) is lesser in HHS vs. barefoot walking; (2) the proximal knee and hip muscles compensate by showing greater power generation and work performed from mid-stance to early swing (K2, H1 and H3 power generation phases) for HHS compared to barefoot walking and (3) ratios of peak power generation and positive work performed by ankle, knee and hip in specific power phases (i.e. K2/A2, H1/A2 and H3/A2) indicative of biomechanical plasticity are greater for HHS compared to barefoot walking.

Methods

Participants

Fourteen young female participants (20.5 ± 0.9 years; 62.6 ± 8.0 kg; 167.1 ± 3.1 cm) with prior experience walking in HHS were recruited for the study. Individuals with neurological disorders and those who reported injury in the lower extremity in the past year were excluded from the study. The university Institutional Review Board approved the study design and procedures. Prior to participating in the study, all participants gave written informed consent.

Data collection

Participants completed a single testing session which began with screening for the inclusion and exclusion criteria. Participants then changed into lycra shorts and tank top provided by the researchers. Anthropometric characteristics such as body mass, height, leg length, inter anterior superior iliac spine distance, knee width and ankle width were measured. Sixteen reflective markers were placed on the participants’ skin and lycra clothing according to the lower extremity Vicon Plug-in Gait Model (Centennial, CO, USA). Three-dimensional marker position data were captured at 100 Hz using a 10-camera VICON motion capture system (Centennial, CO, USA). Ground reaction force data were captured synchronously with marker position data at 1000 Hz using two AMTI in-ground force plates (Watertown, MA, USA).

A static calibration trial was completed according to the VICON PlugIn Gait model guidelines during which the participants were instructed to stand stationary in the ‘motorcycle pose’. Participants then completed three over-ground walking trials in a gait laboratory at a controlled speed of 1.3 m·s⁻¹ for the two experimental conditions, barefoot and wearing HHS, in a random order. A controlled walking speed was chosen to control for the confounding effects of speed on the lower extremity net joint power and work (Silder et al., 2008; Monaco et al., 2009; Cofre et al., 2011). Furthermore, a walking speed of 1.3 m·s⁻¹ implemented in the current study closely approximates the preferred walking speed of young adults when walking barefoot and also when walking in HHS (Barkema et al., 2012). All participants wore 10.8-cm commercially available HHS provided by the researchers. For both experimental conditions, several familiarization trials were performed prior to data collection, so the participants could walk across the walkway within ±3% of the target speed and strike the force plate without making visible alterations to their walking pattern (Barkema et al., 2012; Buddhadev and Martin 2016). Walking speed was facilitated by providing immediate post-trial feedback on speed. Immediately after each trial, walking speed was calculated as the horizontal velocity of the sacral marker over a gait cycle in the measurement zone. If the walking speed was not within ±3% of 1.3 m·s⁻¹, the trial was repeated until three acceptable trials of data were collected for each experimental condition (Buddhadev and Martin 2016).

Data analysis

Marker position and ground reaction forces were filtered using a low pass fourth-order Butterworth filter at 8 Hz. Location of the hip, knee and ankle joint centers was computed from anthropometric measurements and marker position data collected during the static calibration and walking trials. Pelvis, thigh, leg and foot segments’ mass, center of mass and moment of inertia were determined using marker position data and Dempster’s (1955) equations. For each trial, a stride (i.e. gait cycle) was identified as two consecutive heel strikes of the right foot. The first heel strike occurred on the force platform and was identified at the instant when the vertical ground reaction force exceeded a threshold of 20 N (Barkema et al., 2012; Buddhadev and Martin 2016). Second heel strike did not occur on the force platform and was predicted using an algorithm developed by Zeni et al., (2008). The second heel strike occurred at the instant when anterior–posterior distance between the sacral marker and the right heel marker became the largest (Zeni et al., 2008).

An inverse dynamic analysis was then used to compute sagittal plane net internal joint forces, moments and powers at the ankle, knee and hip for a stride using the VICON Plug-in gait post-capture pipeline. Specific power phases at the ankle (A1 and A2), knee (K0–K4) and hip joints (H1–H3) during the gait cycle were identified as described previously (Winter 1991; Eng and Winter 1995; Silder et al., 2008; Teixeira-Salmela et al., 2008; Monaco et al., 2009; Cofre et al., 2011; Kuhman et al., 2018). These power phases provide insights into which lower extremity muscle groups are generating and absorbing energy during walking. Changes in lower extremity muscular function induced by walking in HHS compared to barefoot walking were determined by examining the magnitude of peak power and work performed during these power phases (DeVita and Hortobagyi 2000; Monaco et al., 2009; Cofre et al., 2011). Furthermore, biomechanical plasticity induced by HHS walking was examined by computⁱng ratios of hip and ankle (H1/A2 and H3/A2) and knee and ankle (K2/H1) work and peak power for the experimental conditions as described previously (Kuhman et al., 2018). Increase in the magnitude of ratios indicates a distal-to-proximal redistribution of power and work and an increase in magnitude of biomechanical plasticity (Kuhman et al., 2018).

Statistical analysis

For each dependent variable, differences between barefoot and HHS walking conditions were determined using paired t-tests. The alpha level was adjusted to 0.017 to reduce the risk of Type I error related to multiple hypothesis tests (0.05/3). The effect sizes were calculated as Cohen’s d. Small, medium and large effect sizes correspond to Cohen’s d values of 0.20, 0.50 and 0.80, respectively (Cohen 1988). All statistical analysis was performed using the SPSS software version 25 (Chicago, IL, USA, 2011).

Results

Calculated walking speed results show that the participants walked within 3% of the target speed under both experimental conditions (barefoot: 1.26 ± 0.02 m·s⁻¹; HHS: 1.27 ± 0.02 m·s⁻¹; p = 0.020; Cohen’s d = 0.41). These walking speeds are not statistically different between the HHS and barefoot conditions at the adjusted alpha level of p = 0.017. The magnitude of the difference is negligible (<1%), and thus, this difference is not meaningful. Additionally, the stride lengths at experimental walking speeds were not different between the two conditions (barefoot: 1.34 ± 0.02 m; HHS: 1.34 ± 0.07 m; p = 0.999; Cohen’s d = 0.00). The step lengths were also not different between the two experimental conditions (barefoot: 0.68 ± 0.04 m; HHS: 0.65 ± 0.09 m; p = 0.156).

Figure 1 shows ensemble average profiles for ankle, knee and hip sagittal plane joint angles, net joint moments and powers for barefoot and HHS walking conditions. Tables 1 and 2 summarize the peak net joint powers and work performed during the power phases, respectively. When walking in HHS, the participants generated less peak power and performed less positive work during late-stance ankle A2 push-off phase compared to barefoot walking. Conversely, the power generation and positive work performed at the early-stance knee K0 and mid-stance knee K2, and early-stance hip H1 and early swing hip H3 power phases were greater for HHS compared to barefoot walking. Similarly, power absorption and negative work performed during mid-stance hip H2 phases was also greater for HHS compared to barefoot walking. For the early to mid-stance ankle A1 power phase, less negative work was performed for HHS compared to barefoot walking. No difference in peak power absorption was observed during the early to mid-stance ankle A1 power phase between the experimental conditions. In addition, the power absorption and negative work performed during the early swing knee K3 and late swing knee K4 phases were not affected by the experimental conditions.

Figure 1. Ensemble average sagittal plane joint angle, moment and power profiles for the ankle, knee and hip joints. * indicates significant differences (p < 0.05) in peak joint powers and work performed between the high-heel and barefoot walking conditions during the specific power phases.

Table 1. Peak net joint powers across the power phases of a gait cycle.

Power phases	Peak power (W·kg^−1)		t value (t13)	p value	Cohen’s d
	Barefoot	High-heel
A1-ankle power absorption early-mid stance	−0.851 ± 0.261	−0.734 ± 0.138	−1.41	0.183	0.54
A2-ankle power generation late stance*	4.453 ± 0.904	1.785 ± 0.700	18.83	<0.001	5.03
K0-knee power generation early stance*	0.404 ± 0.095	0.548 ± 0.134	−5.36	<0.001	1.44
K1-knee power absorption early-mid stance*	−0.794 ± 0.324	−1.242 ± 0.562	3.73	0.003	1.00
K2-knee power generation mid-stance*	0.832 ± 0.374	1.645 ± 0.424	−7.83	<0.001	2.09
K3-knee power absorption early swing	−1.934 ± 0.638	−2.118 ± 0.532	1.07	0.305	0.29
K4-knee power absorption late swing	−1.554 ± 0.348	−1.407 ± 0.376	−1.54	0.147	0.41
H1-hip power generation early stance*	0.256 ± 0.116	0.565 ± 0.246	−4.46	0.001	1.19
H2-hip power absorption mid stance*	−0.811 ± 0.405	−1.477 ± 0.721	6.21	<0.001	1.66
H3-hip power generation late stance–early swing*	1.558 ± 0.361	2.103 ± 0.433	−4.57	0.001	1.22

Data presented as mean ±1 standard deviation (SD). * denotes significant difference (p < 0.017).

Table 2. Net joint work across the power phases of a gait cycle.

Power phases	Work (J·kg^−1)		t value (t13)	p value	Cohen’s d
	Barefoot	High-heel
A1-ankle negative work early-mid stance *	−0.145 ± 0.045	−0.096 ± 0.036	−4.61	p < 0.001	1.22
A2-ankle positive work late stance*	0.377 ± 0.070	0.128 ± 0.058	30.65	<0.001	8.12
K0-knee positive work early stance*	0.014 ± 0.005	0.028 ± 0.010	−5.77	<0.001	1.51
K1-knee negative work early-mid stance	−0.050 ± 0.023	−0.073 ± 0.040	2.69	0.018	0.72
K2-knee positive work mid-stance*	0.103 ± 0.053	0.247 ± 0.057	−10.25	<0.001	2.75
K3-knee negative work early swing	−0.261 ± 0.089	−0.263 ± 0.054	0.08	0.932	0.03
K4-knee negative work late swing	−0.141 ± 0.032	−0.135 ± 0.036	−0.85	0.411	0.25
H1-hip positive work early wtance*	0.022 ± 0.016	0.064 ± 0.027	−4.39	0.001	1.84
H2-hip negative work mid-stance *	−0.065 ± 0.027	−0.045 ± 0.023	2.86	0.013	0.74
H3-hip positive work late stance–early swing*	0.242 ± 0.049	0.302 ± 0.038	−4.81	<0.001	1.29

Data presented as mean ±1SD. * denotes significant difference (p < 0.017).

Biomechanical plasticity ratios indicating a distal-to-proximal redistribution of joint power and work are presented in Table 3. All peak joint power and work ratios (K2/A2, H1/A2 and H3/A2) were greater for the HHS compared to the barefoot walking conditions.

Table 3. Biomechanical plasticity ratios for walking.

	Barefoot	High-heel	t-value (t13)	p	Cohen’s d
Peak net joint power generation ratios
Knee mid-stance/ankle late-stance power ratio (K2/A2 power)*	0.193 ± 0.093	1.068 ± 0.534	−6.58	<0.001	1.76
Hip early-stance/ankle late-stance power ratio (H1/A2 power)*	0.067 ± 0.035	0.396 ± 0.338	−3.97	0.002	1.04
Hip late stance–early swing/ankle late-stance power ratio (H3/A2 power)*	0.361 ± 0.095	1.481 ± 0.987	−4.51	0.001	1.20
Net joint positive work ratios
Knee mid-stance/ankle late-stance work ratio (K2/A2 work)*	0.281 ± 0.164	2.368 ± 1.284	−6.36	<0.001	1.70
Hip early-stance/ankle late-stance work ratio (H1/A2 work)	0.063 ± 0.060	0.771 ± 1.053	−2.64	0.020	0.71
Hip late stance–early swing/ankle late-stance work ratio (H3/A2 work)*	0.654 ± 0.136	3.083 ± 2.000	−4.70	<0.001	1.26

Data presented as mean ±1 SD. * denotes significant difference (p < 0.017).

Discussion

The purpose of the present study was to examine the effects of wearing HHS on peak power and work performed across the power phases of gait. The three working hypotheses were: (1) walking in HHS reduces push-off shown by lower positive ankle power generated and work performed in the late stance (A2 phase) compared to barefoot walking, (2) the knee and hip musculature crossing more proximal joints compensates for the reduced push-off via greater power generation and work performed from mid stance to early swing (in K2, H1 and H3 power phases) for HHS compared to barefoot walking and (3) biomechanical plasticity ratios of peak power and work performed, indicative of distal-to-proximal redistribution of power and work in specific power phases (i.e. K2/A2, H1/A2 and H3/A2), are greater for HHS compared to barefoot walking. All three hypotheses were supported by the data, as discussed further.

Walking in HHS substantially affects foot and ankle ROM, moments, powers and work during walking. Compared to barefoot walking, the ankle remains in a plantar flexed position throughout the gait cycle when walking in HHS (Figure 1). Such foot position could disrupt the ankle-foot rocker mechanics which contribute to smooth forward progression of the body over the feet during the stance phase. During early to mid-stance when walking barefoot, plantar flexor muscles absorb energy and act eccentrically during the A1 power phase to control the forward rotation of the tibia over the foot (Winter 1991; Eng and Winter 1995). And during terminal stance, plantar flexors generate positive power and act concentrically to exert a vigorous push-off (A2 power phase) (Winter 1991; Eng and Winter 1995). When walking in HHS, 34% less work was performed during the early to mid-stance A1 power absorption phase. Furthermore, the peak power was 2.5 times lower and work performed was 2.9 times less during the late-stance A2 power generation phase for HHS compared to barefoot walking. Substantial reduction in plantar flexor contribution with HHS walking is consistent with Simonsen et al., (2012), who reported significantly lower plantar flexor moments during the stance phase of walking for HHS compared to barefoot walking. The reduced contribution of plantar flexors during the stance phase of HHS walking, especially the diminished push-off in late stance, could be explained by the plantar flexed position characteristic of HHS walking which places the gastrocnemius, soleus and associated non-contractile elements in a shortened, less efficient state (Farrag and Elsayed 2016). During stance phase of walking, plantar flexors perform majority of the work needed for forward propulsion and maintaining upright support (DeVita and Hortobagyi 2000; Silder et al., 2008; Cofre et al., 2011; Buddhadev and Martin 2016). Owing to the reduced plantar flexor contribution during push-off, compensatory increases would be needed from proximal knee and hip muscles when walking in HHS (Simonsen et al., 2012).

Evidence supporting the hypothesized compensation for diminished plantar flexor push-off power was, indeed, observed at the knee in the present study. Peak knee flexor power was 36% greater and positive work performed was two times greater at the knee during initial contact (K0 power generation phase) in HHS compared to barefoot. HHS demonstrated 56% greater power and 46% greater work absorbed to control knee collapse during early stance (K1 power absorption phase) than in barefoot. In the mid-stance K2 power generation phase, peak knee extensor power and work performed were 2 times greater and 2.4 times greater in HHS than in barefoot, respectively. Our data also showed greater knee ROM in HHS during the mid-stance K2 power generation phase (Figure 1). These findings are consistent with those of Simonsen et al., (2012), who reported greater knee flexion moment in early-stance KO power absorption phase and higher knee extension moment in mid-stance K2 power generation phase during HHS walking than barefoot. The greater mid-stance knee power generation/late-stance ankle power generation (K2/A2) biomechanical plasticity ratios observed for peak power and work in HHS further illustrate the shifting contributions to the task away from the ankle and toward the more proximal joints. Presumably because the ankle is plantar flexed at initial contact, and due to the smaller surface area of the HHS heel, the knee must assume a more flexed position in order to provide greater stability and shock absorption to begin the stance phase. This ankle position also appears to disrupt the rocker mechanisms of the foot and ankle (both the heel and ankle rocker) requiring greater power generation at the knee to maintain a given walking speed (Adams and Cerny 2018).

Walking in HHS necessitated substantial increases in hip function compared to barefoot walking. Over the gait cycle, we observed increase in hip ROM and increases in hip flexors and extensor moments for HHS compared to barefoot walking (Figure 1). The peak hip extensor power in early stance was 2.2 times greater and work performed was 2.98 times greater in early stance (H1 power generation phase) for HHS than barefoot walking. This increased hip extensor activity can be attributed to more power generated by these muscles to pull the body upward and forward (Winter 1991; Eng and Winter 1995) to compensate for the less effective push-off and disruption of the foot-ankle rocker mechanisms. The compensatory role of the hip extensors is supported by greater early-stance hip power generation/late-stance ankle power generation (H1/A2 ratio) biomechanical plasticity ratio for HHS compared to barefoot walking. In addition to hip extensor activity, hip flexor activity was also substantially greater for HHS walking. Generally, during walking from mid-to-terminal stance, the hip flexor muscles absorb energy to decelerate the thigh extension (H2 power absorption phase) (Winter 1991; Eng and Winter 1995). And then, from late stance to early swing, hip flexors generate positive power to pull the leg into swing phase (H3 power generation phase) (Winter 1991; Eng and Winter 1995). For HHS walking, we observed that the greater negative power and work were needed to decelerate the thigh extension at the hip joint. Also, hip flexors generated 1.3 times greater peak power and performed 1.3 times greater work in the early swing H3 power generation phase to pull the leg into swing for HHS compared to barefoot walking. The increase in hip flexor activity during the early swing H3 power generation phase could have compensated for the reduced push-off power by the plantarflexors. Greater late stance–early swing hip power generation/late-stance ankle power generation (H3/A2) and positive work biomechanical plasticity ratios for HHS compared to barefoot walking support the compensatory role of the hip flexor muscles.

The current study has two limitations. First, we used a fixed speed rather than preferred speed for the experiment. Previous research has shown that preferred walking speed is lower when walking in HHS than barefoot walking (Opila-Correia 1990; de Lateur BJ et al., 1991; Titchenal et al., 2015). Thus, outcomes of the study may not be representative of gait of young women at their preferred speed. We chose to use a fixed walking speed because it is a confounding variable that directly affects net joint powers and work, the primary dependent variables. Previous research has shown that net joint work and power increase with walking speed (Silder et al., 2008; Teixeira-Salmela et al., 2008; Cofre et al., 2011; Buddhadev and Martin 2016). Had we conducted the study at a preferred speed, young women may have walked at a slower speed for HHS compared to barefoot walking. With different walking speeds for HHS vs. barefoot walking, we may not have been able to discern whether the changes observed in gait kinetics were a result of walking in HHS or simply because the participants walked at different speeds for the two experimental conditions. Future researchers should consider examining gait adaptations induced by HHS walking at fixed and preferred speeds to determine whether adaptations observed are consistent across the two approaches.

The second limitation is that we only assessed the gait of young women. Use of HHS is common in women of all ages. With aging, older adults show deficit in exerting a vigorous push-off power during late-stance phase. They compensate for the reduced push-off power by generating more power and performing more work with the knee and hip muscles (DeVita and Hortobagyi 2000; Silder et al., 2008; Cofre et al., 2011; Buddhadev and Martin 2016). Owing to the similarity between the nature of compensatory mechanisms for walking in HHS and with aging, it is possible that older women could demonstrate much greater biomechanical plasticity (i.e. distal-to-proximal redistribution of power and work) when walking in HHS compared to young women. Future researchers could examine if the age-associated biomechanical plasticity of gait is amplified when walking in HHS.

In summary, when walking in HHS compared to barefoot, participants exerted a diminished push-off characterized by lesser peak power and lesser work at the ankle joint in late stance. To compensate greater peak power was generated and work was performed in early stance by hip extensors, mid-stance by knee extensors and late stance and early swing phase by hip flexor muscles. Walking in HHS induces biomechanical plasticity and causes distal-to-proximal redistribution of net joint power and work during walking.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The participants in the study did not agree for their data to be shared publicly and therefore, the supporting data is not available.

Additional information

Funding

This work was supported by the manuscript grant from the Office of Research and Sponsored Programs at Western Washington University.