Abstract
Background
Flip-flop and barefoot running are increasingly common; however, their biomechanical effects remain unclear. This study examined lower-limb kinematics and muscle activation during treadmill running under barefoot, conventional flip-flop, Y-Sandal, and shod conditions.
Methods
Sixteen volunteers were recruited to run on a treadmill under four footwear conditions at three speeds. A three-dimensional motion capture system was used to collect kinematic data, and the activities of the tibialis anterior, extensor digitorum longus, and flexor hallucis longus muscles were simultaneously recorded using a wireless electromyography system.
Results
Barefoot running was associated with higher cadence, whereas the Y-Sandal condition was associated with the shortest stance duration. The conventional flip-flop condition was associated with greater sagittal-plane hip motion compared with the Y-Sandal and shod conditions. The barefoot and conventional flip-flop conditions were associated with similar ankle and forefoot movement patterns in sagittal and frontal planes, whereas the Y-Sandal and shod conditions produced different patterns. Furthermore, significantly higher lower-leg muscle activation was observed under the conventional flip-flop and Y-Sandal conditions compared with the barefoot and shod conditions.
Conclusion
Footwear design significantly affects running biomechanics and neuromuscular strategies. Although barefoot running promotes a higher cadence, thong-style footwear increases the demand on the distal musculature. The specific design features of the Y-Sandal appear to mitigate the compensatory kinematic strategies typically observed with conventional flip-flops, resulting in movement patterns that more closely resemble those observed during shod running. These findings highlight the distinct neuromuscular demands and kinematic adaptations associated with different footwear interfaces.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s13102-026-01870-y.
Keywords: Flip-flop, Barefoot, Running, Kinematics, Electromyography
Introduction
Running is among the most popular forms of physical activity worldwide. A study reported that 59.1% of active adults consider outdoor activities such as running among the best methods to stay fit, and 39.4% of gym members have shifted toward running and other outdoor activities [1]. Although running does not require specialized equipment or facilities, recent technological advances have accelerated the development of running shoes designed to improve performance and reduce injury risk.
The incidence and prevalence of running-related musculoskeletal injuries are 40.2% and 44.6%, respectively [2]. Although softer and thicker midsoles are commonly used to reduce the impact force and loading rate [3], systematic reviews have indicated that cushioned insoles may not effectively prevent running-related injuries [4, 5]. By contrast, barefoot and minimalist running have gained popularity as alternative running approaches [6–8]. As described in the book Born to Run [9], the Tarahumara Huarache, a traditional thong-style sandal resembling a flip-flop, has resulted in increased interest in thong-style footwear for running. However, clinicians often caution against running in conventional flip-flops because these footwear types provide minimal arch support and toe protection and are considered potential contributors to various musculoskeletal complaints [10].
Studies on flip-flop footwear have mainly focused on walking [11–19]. According to our review of the literature, only one study examined running kinematics in children wearing flip-flops [17]. Because flip-flops provide a relatively loose connection between the foot and sole, users may adopt compensatory kinematic strategies, particularly in the sagittal plane, to prevent the footwear from slipping during walking and running [13–15, 17, 18]. Shroyer et al. observed participants wearing flip-flops and reported greater ankle plantarflexion throughout the gait cycle [18]; they hypothesized that increased toe flexion helps secure the footwear, which is consistent with the increased hallux pressure reported by Price et al. [16]. However, other studies comparing participants wearing flip-flops with barefoot participants have reported increased dorsiflexion and decreased plantarflexion in those wearing flip-flops [13–15, 17] without a corresponding increase in hallux pressure [19]. These inconsistent findings may reflect differences in footwear design and movement tasks and highlight a key gap in the literature: most interpretations of footwear-securing strategies have been indirectly inferred from kinematics and plantar pressure data. Without electromyography (EMG) measurements, the specific neuromuscular demands involved, particularly those associated with toe flexor and extensor muscles, remain unclear. Only a few EMG studies on gait in flip-flops have focused on the thigh and calf muscles [11, 12], with little attention given to the tibialis anterior (TA), extensor digitorum longus (EDL), and flexor hallucis longus (FHL), which may help stabilize the flip-flops during the swing and push-off phases. Understanding how footwear design features affect ankle and forefoot kinematics can provide preliminary biomechanical evidence regarding the demands placed on the lower limb. Such evidence is a necessary prerequisite for future research on footwear selection and the development of evidence-based training recommendations.
The present study investigated lower-limb kinematics and lower-leg muscle activation during running under barefoot, conventional flip-flop, Y-Sandal, and shod conditions. We hypothesized that joint kinematics would differ among footwear conditions and that toe-related muscles (e.g., the FHL and EDL) and the dorsiflexor TA would exhibit higher activation under flip-flop conditions than under barefoot and shod conditions, reflecting the additional neuromuscular effort required to secure thong-style footwear. By combining three-dimensional kinematic analysis with targeted EMG assessment of the toe flexor and extensor muscles during running, this study addresses a crucial gap in the literature by characterizing the acute neuromuscular response to different footwear designs. These findings provide preliminary insights into the biomechanics of running in flip-flops and establish a foundation for future longitudinal studies investigating footwear adaptation and injury prevention.
Materials and methods
Sample size estimation
The sample size was calculated using G*Power (Version 3.1.9.7, University of Kiel, Germany). On the basis of the effect size reported in a previous study that examined center-of-pressure parameters under different footwear conditions [15], the present study used an effect size of 0.85. With an alpha level of 0.05 and a statistical power of 0.80, a minimum sample size of 16 participants was required for this study.
Participants
A total of 16 participants (8 women and 8 men) with running experience were recruited from race participants and the university community. The inclusion criteria were as follows: (1) having the ability to run continuously for at least 5 km, (2) having previous experience running on a treadmill, and (3) not having prior experience running in flip-flops. The exclusion criteria were as follows: (1) having a lower-limb injury within the previous 3 months; (2) having a history of neurological, orthopedic, or rheumatic conditions affecting gait; or (3) having foot deformities that could interfere with wearing flip-flops (e.g., severe hallux valgus). This study was approved by the Research Ethic Committee of China Medical University Hospital (CMUH113-REC3-036). All experimental procedures were performed in accordance with relevant guidelines and the principles of the Declaration of Helsinki. All participants received written and verbal information regarding the study and provided written informed consent before testing.
Footwear conditions
This study conducted running experiments under four footwear conditions: barefoot, conventional flip-flop, Y-Sandal, and shod conditions. For the barefoot condition, the participants wore adhesive soles (NakeFit, Italy) to protect their feet from abrasions. For the conventional flip-flop condition, the participants wore conventional flip-flops purchased from the local market; they featured a thong-style design with a Y-shaped strap and a flat sole. For the Y-Sandal condition, the participants wore Y-Sandals (BN521, B&M, Lurng Furng Development Ltd., Taiwan; Fig. 1); this model is used by athletes during ultramarathon races. Similar to conventional flip-flops, Y-Sandals have a thong-style design but also include medial arch support, longer and wider straps for enhanced stability, and a thickened cushioned heel for shock absorption, resulting in a 15-mm heel-to-toe drop. Finally, for the shod condition, the participants wore their usual running shoes. These shoes were verified by the researchers as standard daily trainers in the neutral or stability categories and did not have specialized features such as carbon plates or maximalist extra-thick cushioning. Although this approach introduced some variability in stack height and foam density, it was chosen to reflect the participants’ habitual shod running mechanics.
Fig. 1.
Footwear conditions tested: barefoot adhesive soles (A), conventional flip-flops (B), and Y-Sandals (C)
Protocol
The anthropometric characteristics of the participants were measured at the beginning of the experiments. After placement of the retro-reflective markers and EMG sensors, static calibration was performed, followed by the removal of the calibration markers before the running trials began. The participants then underwent several minutes of treadmill acclimatization (HT-318, Healthpit, Tokuyo, Japan) under each footwear condition. After the participants verbally confirmed adaptation and a steady gait pattern was observed, they ran at three speeds (8, 9, and 10 km/h), with each speed maintained for 2 min. Kinematic and EMG data were collected during three randomly selected 10-second intervals at each speed. The participants were not informed of the timing of data collection. A 5-minute break was provided between footwear conditions for rest, footwear changes, and model calibration. The order of footwear conditions was randomized.
EMG
A wireless EMG system (Trigno Delsys Co., MA, USA) was used to record the activities of the TA, FHL, and EDL muscles. Before electrode placement, the skin was shaved and cleaned with 75% alcohol. The electrodes were positioned in accordance with SENIAM recommendations and previous studies [20–22]. Ultrasound imaging (Lumify, Philips, Amsterdam, The Netherlands) was used to identify the location of the FHL for electrode placement (Fig. 2) [20]. For each participant, the highest EMG signal recorded for each muscle across all running conditions and speeds was defined as the dynamic EMG peak (EMGPEAK). This dynamic normalization approach was used instead of traditional isometric maximum voluntary contraction testing to ensure that the normalization value was task-specific and representative of the maximum muscle activation observed during running [23–26]. All subsequent EMG data are expressed as a percentage of this EMGPEAK (%EMGPEAK).
Fig. 2.
Placement of the surface electrode for the FHL determined using ultrasound according to a previous study [20]
Kinematics
Three-dimensional motion data were collected using a Vicon motion capture system (Vicon, Oxford Metric Ltd., Oxford, UK). Spherical retro-reflective markers were placed on anatomical landmarks following the guidelines of the Conventional Gait Model 2.4 lower-body model. Because participants may exhibit various foot-strike patterns during running, including heel strike, midfoot strike, and forefoot strike, initial contact (IC) was identified as the first occurrence of the lowest point reached in the trajectory of either the heel marker (HEE), first metatarsophalangeal joint marker (FMH), or fifth metatarsophalangeal joint marker (VMH). Toe-off (TO) was defined as the point at which the FMH trajectory began to rise. During data processing, the interval between two consecutive IC events was normalized to 100% of the gait cycle, and late swing was defined at 85% of the gait cycle.
Data collection and processing
Kinematic and EMG data were collected simultaneously using Vicon Nexus software at sampling rates of 100 and 2,000 Hz, respectively. The collected data were processed and exported to MATLAB (MathWorks, Natick, MA, USA) for further analysis. Approximately 14 strides were analyzed for each trial, and joint angles were ensemble averaged across strides. For each trial, the minimum and maximum angles of the hip, knee, ankle, and forefoot joints in the sagittal, frontal, and transverse planes were obtained and used to calculate the range of motion (ROM) during the stance phase.
EMG data were initially processed using a third-order Butterworth high-pass filter at 6 Hz to remove movement artifacts. The root-mean-square values of the signals were then calculated and smoothed using an 8-Hz low-pass filter. For each stride, muscle activity data were resampled on the basis of the normalized gait cycle (0%–100%), which was determined from kinematic gait events and then averaged across strides. Subsequently, the signal amplitude was normalized to its corresponding EMGPEAK value. The maximum percentage of EMGPEAK (%EMGPEAK) for each muscle was calculated for each footwear condition and running speed.
Statistical analysis
Statistical analyses were performed using SPSS (IBM SPSS Statistics, IBM Corp., Armonk, NY, USA). Paired t tests were used to examine differences between left- and right-side measurements. If no significant side-to-side differences were identified, the mean of the left- and right-side values was used for subsequent analyses. Two-way repeated-measures analysis of variance was used to examine differences in gait parameters, kinematic variables, and EMG activity among the footwear conditions and running speeds. The Greenhouse–Geisser correction was applied when the assumption of sphericity was violated. Post hoc pairwise comparisons were conducted using the least significant difference method when significant differences were noted. A p value of < 0.05 was considered statistically significant.
Results
Gait parameter
The participants exhibited significantly higher cadence under the barefoot condition (174.25 ± 2.53 steps/min, p = 0.001) than they did under the other footwear conditions. Additionally, the cadence observed at a speed of 10 km/h (171.94 ± 2.37 steps/min) was significantly higher than that observed at 8 and 9 km/h (p = 0.03 and 0.02, respectively). A significant interaction effect between running speed and footwear condition was observed, with cadence increasing with speed under the barefoot and both flip-flop conditions, whereas cadence remained unchanged under the shod condition (Fig. 3A). The stance phase duration decreased significantly as the running speed increased and was significantly shorter under the Y-Sandal condition than it was under the other footwear conditions (Table 1).
Fig. 3.
Interaction effects of footwear conditions and running speeds on (A) cadence (F = 2.765, p = 0.031, 
= 0.156, power = 0.754); (B) knee ROM (F = 2.971, p = 0.042, 
= 0.165, power = 0.665); (C) forefoot ROM (F = 3.619, p = 0.016, 
= 0.194, power = 0.795); (D) right TA activation (F = 3.528, p = 0.021, 
= 0.190, power = 0.757); and (E) left EDL activation (F = 3.683, p = 0.017, 
= 0.197, power = 0.781)
Table 1.
Mean values, SDs, F values, p values, and effect sizes for differences in gait parameters among footwear conditions and running speeds
| Condition | Speed (km/hr) | Cadence (steps/min) | Stance Phase (%) |
|---|---|---|---|
| Mean (SD) | |||
| Barefoot | 8 | 170.25 (12.56) | 32.98 (3.92) |
| 9 | 174.75 (11.57) | 32.94 (3.35) | |
| 10 | 177.75 (9.0) | 31.65 (3.36) | |
| Flip-flop | 8 | 165.75 (13.3) | 33.13 (2.68) |
| 9 | 167.25 (11.14) | 32.05 (2.14) | |
| 10 | 171.75 (10.48) | 30.97 (2.36) | |
| Y-Sandal | 8 | 168.75 (11.97) | 31.89 (3.07) |
| 9 | 166.5 (10.62) | 30.73 (2.45) | |
| 10 | 171 (12.0) | 30.00 (2.37) | |
| Shoes | 8 | 167.25 (11.97) | 33.82 (4.85) |
| 9 | 167.25 (12.75) | 33.43 (4.82) | |
| 10 | 167.25 (11.14) | 33.18 (4.78) | |
| F | Condition | 8.595 | 4.248 |
| Speed | 8.167 | 13.690 | |
| Condition x Speed | 2.765 | 1.496 | |
| Sig. | Condition | < 0.001abc | 0.029bdf |
| Speed | 0.001hi | 0.001ghi | |
| Condition x Speed | 0.031 | 0.188 | |
|
Condition | 0.364 | 0.221 |
| Speed | 0.353 | 0.477 | |
| Condition x Speed | 0.156 | 0.091 | |
Data are presented as the mean (SD)
Abbreviations: SD standard deviation
A p value of < 0.05 was considered statistically significant. Statistical significance was denoted as follows
a, barefoot significantly different from flip-flops; b, barefoot significantly different from Y-Sandals; c, barefoot significantly different from shoes; d, flip-flops significantly different from Y-Sandals; e, flip-flops significantly different from shoes; f, Y-Sandals significantly different from shoes; g, 8 km/h significantly different from 9 km/h; h, 8 km/h significantly different from 10 km/h; i, 9 km/h significantly different from 10 km/h
Kinematics
No significant differences were observed between the left- and right-side measurements; therefore, the mean value for both sides was used for subsequent comparisons (Table 2). Under the barefoot and conventional flip-flop conditions, the participants exhibited significantly greater ankle ROM in the sagittal plane (26.48° ± 0.90° and 26.01° ± 0.85°, respectively) than they did under the Y-Sandal and shod conditions throughout the stance phase (all p < 0.01). Moreover, under the barefoot and conventional flip-flop conditions, the participants demonstrated greater plantarflexion throughout the stance phase than they did under the Y-Sandal and shod conditions. Under the conventional flip-flop condition, the participants exhibited a greater hip ROM (33.05° ± 1.24°) than they did under the Y-Sandal (p = 0.02) and shod (p = 0.02) conditions. The greatest knee ROM was observed under the shod condition (25.94° ± 1.63°, all p < 0.001). Furthermore, under the shod condition, the participants exhibited significantly greater knee flexion during late swing than they did under the other conditions; however, the barefoot condition was associated with the least knee flexion.
Table 2.
Mean values, SDs, F values, p values, and effect sizes for differences in joint range of motion among footwear conditions and running speeds
| Condition | Plane | Sagittal | Frontal | Transverse | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Speed (km/hr) | Hip | Knee | Ankle | Fore-foot | Hip | Knee | Ankle | Fore-foot | Hip | Knee | Ankle | Fore-foot | |
| Barefoot | 8 | 29.11 (4.91) | 19.97 (4.32) | 25.78 (3.97) | 7.12 (2.7) | 8.14 (2.2) | 3.46 (0.93) | 1.7 (0.26) | 7.62 (3.51) | 7.16 (2.28) | 11.88 (2.94) | 11.07 (1.87) | 4.89 (1.82) |
| 9 | 32.53 (5.58) | 20.75 (4.54) | 25.65 (4.24) | 7.68 (2.66) | 8.95 (2.76) | 4.21 (1.38) | 1.69 (0.3) | 8.13 (3.65) | 7.35 (2.75) | 12.62 (3.11) | 11.1 (1.73) | 5.34 (1.76) | |
| 10 | 34.4 (5.14) | 21.52 (3.44) | 28.02 (4.28) | 8.83 (3.06) | 9.62 (2.3) | 4.63 (1.56) | 1.8 (0.36) | 8.47 (3.61) | 7.62 (2.75) | 13.71 (2.8) | 11.71 (1.77) | 5.36 (1.9) | |
| Flip-flop | 8 | 30.69 (4.71) | 21.28 (5.12) | 25.11 (4.11) | 7.88 (3.47) | 9.51 (2.99) | 4.13 (0.98) | 1.43 (0.35) | 7.87 (2.73) | 6.91 (2.41) | 11.47 (2.63) | 10.14 (1.85) | 6.63 (2.25) |
| 9 | 33.97 (6.87) | 22.92 (8.18) | 26.67 (4.72) | 8.38 (3.32) | 10.44 (3.33) | 4.59 (1.48) | 1.46 (0.35) | 7.9 (2.83) | 7.43 (4.26) | 12.06 (2.99) | 10.36 (1.61) | 6.86 (2.36) | |
| 10 | 34.48 (4.28) | 21.56 (3.91) | 26.24 (3.56) | 8.44 (3.57) | 10.77 (3.34) | 4.84 (1.66) | 1.42 (0.4) | 8.05 (2.96) | 7.26 (2.72) | 12.72 (2.58) | 10.11 (2.1) | 6.42 (2.18) | |
| Y-Sandal | 8 | 29.51 (5.52) | 19.49 (3.61) | 22.59 (4.8) | 7.48 (2.68) | 8.68 (2.56) | 3.65 (1.12) | 2.1 (0.68) | 6.82 (3.16) | 6.81 (2.65) | 11.23 (2.73) | 10.18 (2.61) | 5.39 (2.61) |
| 9 | 31.6 (5.67) | 20.59 (3.55) | 22.78 (4.69) | 7.8 (3.35) | 9.32 (2.63) | 3.99 (1.29) | 2.04 (0.56) | 6.9 (2.87) | 7.23 (2.93) | 11.6 (2.84) | 9.98 (2.31) | 5.35 (2.36) | |
| 10 | 33.2 (5.9) | 21.29 (3.84) | 23.67 (5.69) | 8.42 (3.99) | 9.95 (2.69) | 4.45 (1.51) | 2.12 (0.58) | 7.05 (3.09) | 7.63 (2.38) | 12.09 (3.31) | 10.44 (2.52) | 5.56 (2.71) | |
| Shoes | 8 | 27.77 (4.71) | 24.34 (7.07) | 19.32 (4.02) | 7.35 (2.4) | 8.43 (2.3) | 4.09 (1.3) | 1.06 (0.4) | 7.17 (3.53) | 8.24 (4.13) | 12.29 (4.78) | 8.89 (3.48) | 7.53 (2.07) |
| 9 | 31.06 (5.55) | 26.5 (7.08) | 21.44 (5.02) | 8.1 (2.43) | 9.27 (2.97) | 4.88 (1.62) | 1.18 (0.46) | 8.08 (3.83) | 9.07 (4.71) | 13.88 (5.07) | 9.74 (3.44) | 8.44 (2.71) | |
| 10 | 34.21 (4.57) | 27.0 (5.66) | 24.64 (5.46) | 8.65 (2.58) | 10.09 (3.32) | 5.39 (1.97) | 1.22 (0.47) | 8.7 (3.88) | 9.77 (5.26) | 15.24 (5.63) | 9.96 (3.27) | 8.95 (2.35) | |
| F | Condition | 3.195 | 18.409 | 9.567 | 0.131 | 6.923 | 3.859 | 23.306 | 1.106 | 4.085 | 3.135 | 2.937 | 9.625 |
| Speed | 71.654 | 10.839 | 12.583 | 25.276 | 29.872 | 19.344 | 2.267 | 10.414 | 2.708 | 14.381 | 2.994 | 8.027 | |
| Condition x Speed | 2.012 | 1.720 | 2.123 | 1.551 | 0.456 | 1.151 | 1.493 | 1.547 | 0.793 | 2.971 | 1.744 | 3.619 | |
| Sig. | Condition | 0.032de | < 0.001cef | < 0.001bcde | 0.873 | 0.001ade | 0.015cdf | < 0.001abcdef | 0.356 | 0.033ef | 0.035f | 0.043ac | < 0.001acef |
| Speed | < 0.001ghi | 0.002gh | < 0.001hi | < 0.001ghi | < 0.001ghi | < 0.001ghi | 0.121 | < 0.001ghi | 0.083 | < 0.001ghi | 0.065 | 0.002gh | |
| Condition x Speed | 0.072 | 0.191 | 0.058 | 0.206 | 0.839 | 0.338 | 0.226 | 0.172 | 0.485 | 0.042 | 0.12 | 0.016 | |
|
Condition | 0.176 | 0.551 | 0.389 | 0.009 | 0.316 | 0.205 | 0.608 | 0.069 | 0.214 | 0.173 | 0.164 | 0.391 |
| Speed | 0.827 | 0.419 | 0.456 | 0.628 | 0.666 | 0.563 | 0.131 | 0.410 | 0.153 | 0.489 | 0.166 | 0.349 | |
| Condition x Speed | 0.118 | 0.103 | 0.124 | 0.094 | 0.029 | 0.071 | 0.091 | 0.093 | 0.050 | 0.165 | 0.104 | 0.194 | |
Data are reported as the mean (SD)
Abbreviations: SD standard deviation
A p value of < 0.05 was considered statistically significant. Statistical significance was denoted as follows: a, barefoot significantly different from flip-flops; b, barefoot significantly different from Y-Sandals; c, barefoot significantly different from shoes; d, flip-flops significantly different from Y-Sandals; e, flip-flops significantly different from shoes; f, Y-Sandals significantly different from shoes; g, 8 km/h significantly different from 9 km/h; h, 8 km/h significantly different from 10 km/h; i, 9 km/h significantly different from 10 km/h
In the frontal plane, the participants exhibited significantly greater hip ROM under the conventional flip-flop condition (10.24° ± 0.79°) than they did under the other three footwear conditions during the stance phase (p < 0.01). Additionally, the participants exhibited greater knee ROM under the conventional flip-flop condition (4.52° ± 0.29°) than they did under the Y-Sandal condition (4.03° ± 0.31°; p = 0.04), but they exhibited greater knee ROM under the shod condition (4.79° ± 0.39°) than they did under the barefoot (4.10° ± 0.31°, p = 0.015) and Y-Sandal (p = 0.03) conditions. Significant differences in ankle ROM were observed among the four footwear conditions (p < 0.01). The Y-Sandal condition was associated with the greatest ankle ROM (2.09° ± 0.15°), followed by the barefoot (1.73° ± 0.07°) and conventional flip-flop (1.44° ± 0.09°) conditions, whereas the shod condition was associated with the smallest ankle ROM (1.16° ± 0.11°) during the stance phase. In the transverse plane, the shod condition was associated with greater hip ROM (9.04° ± 1.15°) and forefoot ROM (8.31° ± 0.57°) during the stance phase compared with both flip-flop conditions (all p < 0.05) and tended to be associated with greater internal rotation. The barefoot condition was associated with lower forefoot ROM (5.20° ± 0.44°) compared with the shod condition (p < 0.001) but was associated with greater ankle ROM (11.29° ± 0.40°) compared with the conventional flip-flop and shod conditions (p = 0.03 and 0.004, respectively).
Joint ROM in different planes increased with speed (p < 0.02), except for hip ROM in the transverse plane and ankle ROM in both the frontal and transverse planes. Sagittal-plane ankle ROM exhibited a significant increase only at 10 km/h (p < 0.02), whereas transverse-plane forefoot ROM was significantly lower at 8 km/h (p < 0.02). A significant interaction effect was observed for knee and forefoot ROM in the transverse plane (Fig. 3B, C). As the running speed was increased, knee ROM increased to a greater extent under the shod condition than under the other footwear conditions. By contrast, although forefoot ROM increased with speed under most footwear conditions, forefoot ROM decreased under the flip-flop condition when the running speed was increased from 9 to 10 km/h. Mean differences and 95% confidence intervals (CIs) for all kinematic comparisons across running conditions and speeds are provided in Supplementary Tables S1 and S2.
EMG
TA activation exhibited a double-peak pattern, with the highest activation occurring at IC and a second, slightly lower peak occurring during mid-swing (approximately 60% of the gait cycle). FHL activation reached its peak near TO (approximately 32% of the gait cycle). By contrast, the timing of peak EDL activation varied across trials, with peaks occurring at IC, TO, and mid-swing and displaying either a single- or double-peak pattern (Fig. 4).
Fig. 4.
Representative EMG patterns from one participant during a running trial, normalized to the gait beginning at initial contact with the ground. Vertical dotted lines indicate toe-off events. (Abbreviations: L, left; R, right; TA, tibialis anterior; FHL, flexor hallucis longus; EDL, extensor digitorum longus)
For EMG outcomes, significant differences in muscle activation were observed among the footwear conditions (p < 0.001). The conventional flip-flop condition was associated with the highest TA activation level (left: 78.71% ± 4.10%, right: 83.43% ± 2.97%), followed by the Y-Sandal condition, whereas the barefoot and shod conditions were associated with the lowest activation levels. The Y-Sandal condition was associated with the highest FHL activation (left: 76.33% ± 2.95%, right: 70.41% ± 6.01%) and EDL activation (left: 69.11% ± 4.86%, right: 55.89% ± 7.53%) levels, followed by the conventional flip-flop condition. The barefoot condition was associated with the lowest FHL activation level (left: 17.72% ± 5.15%, right: 18.05% ± 3.75%), but it produced a similar EDL activation level to that observed under the shod condition (left: 34.57% ± 32.25%, right: 18.54% ± 4.15%; Table 3).
Table 3.
Mean values, SD, F values, p values, and effect sizes for differences in %EMGPEAK among footwear conditions and running speeds
| Condition | Muscle | TA | FHL | EDL | |||
|---|---|---|---|---|---|---|---|
| Speed (km/hr) | L | R | L | R | L | R | |
| Barefoot | 8 | 45.24 (12.54) | 54.93 (21.56) | 17.16 (18.63) | 16.98 (15.48) | 33.72 (20.93) | 16.96 (16.23) |
| 9 | 48.55 (14.03) | 53.64 (18.53) | 17.95 (23.19) | 18.14 (15.71) | 35.17 (24.06) | 20.06 (19.67) | |
| 10 | 57.49 (15.73) | 52.98 (15.83) | 18.04 (20.26) | 19.03 (14.35) | 34.81 (26.76) | 18.61 (16.05) | |
| Flip-flop | 8 | 69.91 (18.11) | 76.53 (14.58) | 45.89 (24.61) | 45.99 (31.39) | 35.35 (21.43) | 40.28 (31.24) |
| 9 | 76.77 (16.68) | 83.08 (14.41) | 52.91 (26.75) | 51.2 (28.76) | 55.23 (27.17) | 37.06 (20.44) | |
| 10 | 89.44 (20.82) | 90.69 (14.87) | 50.53 (20.35) | 47.89 (25.56) | 61.48 (27.96) | 51.56 (36.42) | |
| Y-Sandal | 8 | 62.04 (16.62) | 65.07 (13.93) | 73.35 (20.13) | 61.74 (29.73) | 55.08 (21.99) | 44.98 (32.08) |
| 9 | 64.92 (17.63) | 68.67 (17.41) | 82.41 (17.58) | 73.1 (28.75) | 72.86 (22.75) | 58.31 (35.17) | |
| 10 | 71.35 (21.65) | 74.24 (16.49) | 73.22 (25.31) | 76.38 (29.13) | 79.39 (22.64) | 64.38 (39.41) | |
| Shoes | 8 | 45.18 (17.24) | 53.03 (18.6) | 26.21 (22.26) | 31.58 (26.23) | 25.89 (20.19) | 15.72 (24.13) |
| 9 | 52.97 (20.62) | 54.74 (16.12) | 34.15 (29.68) | 33.51 (24.07) | 33.77 (29.66) | 16.08 (19.51) | |
| 10 | 59.16 (22.75) | 56.81 (16.99) | 33.37 (24.23) | 36.62 (28.69) | 37.09 (32.53) | 17.79 (14.38) | |
| F | Condition | 19.916 | 33.559 | 35.295 | 18.696 | 11.412 | 13.871 |
| Speed | 54.393 | 3.726 | 3.703 | 2.015 | 36.337 | 2.923 | |
| Condition x Speed | 0.845 | 3.528 | 0.544 | 1.363 | 3.683 | 1.417 | |
| Sig. | Condition | < 0.001abdef | < 0.001abdef | < 0.001abcdef | < 0.001abcdf | < 0.001abdef | < 0.001abef |
| Speed | < 0.001ghi | 0.052 | 0.037g | 0.167 | < 0.001ghi | 0.101 | |
| Condition x Speed | 0.473 | 0.021 | 0.655 | 0.264 | 0.017 | 0.253 | |
|
Condition | 0.570 | 0.691 | 0.702 | 0.555 | 0.432 | 0.480 |
| Speed | 0.784 | 0.199 | 0.198 | 0.118 | 0.708 | 0.163 | |
| Condition x Speed | 0.053 | 0.190 | 0.035 | 0.083 | 0.197 | 0.086 | |
Data are reported as the mean (SD)
Abbreviations: SD standard deviation, EDL extensor digitorum longus, FHL flexor hallucis longus, TA tibialis anterior
A p value of < 0.05 was considered statistically significant. Statistical significance was denoted as follows: a, barefoot significantly different from flip-flops; b, barefoot significantly different from Y-Sandals; c, barefoot significantly different from shoes; d, flip-flops significantly different from Y-Sandals; e, flip-flops significantly different from shoes; f, Y-Sandals significantly different from shoes; g, 8 km/h significantly different from 9 km/h; h, 8 km/h significantly different from 10 km/h; i, 9 km/h significantly different from 10 km/h
Muscle activation differed between the left and right sides across running speeds. EDL and TA activation increased with running speed on both sides; however, significant differences in muscle activation among the three speeds were observed only on the left side (p < 0.001). For FHL activation, a significant difference was observed only on the left side between 8 and 9 km/h (p = 0.005).
An interaction effect was observed for right TA and left EDL activation. Under the barefoot condition, right TA activation decreased as running speed increased; conversely, right TA activation increased with running speed under all other footwear conditions (Fig. 3D). By contrast, left EDL activation increased with speed under all footwear conditions, except for the barefoot condition (Fig. 3E). Mean differences and 95% CIs for all EMG comparisons across conditions and speeds are detailed in Supplementary Tables S3 and S4.
Discussion
Running is among the most popular forms of physical activity worldwide, and the present study examined lower-limb biomechanics during running under four conditions (i.e., barefoot, conventional flip-flop, Y-Sandal, and shod conditions). According to data from World Athletics (formerly IAAF) and the State of Running 2019 report, typical running speeds for a 5-km distance range from 5.77 to 11.11 km/h [27]. Furthermore, Strava metrics indicate that running speeds between 8 and 10 km/h correspond to approximately 30th to 60th percentiles among recreational runners [28]. Thus, running speeds of 8, 9, and 10 km/h were selected to reflect speeds commonly observed among recreational runners and to allow participants to comfortably maintain a steady-state gait during the 2-minute running trials.
A significantly higher cadence was observed during barefoot running, consistent with the findings of previous studies comparing barefoot and shod running [29, 30] and another study comparing barefoot and sandal walking [31]. Both flip-flop conditions were also associated with slightly higher cadence than the shod condition, although the difference was not statistically significant. Previous studies have also reported that during walking, barefoot conditions were associated with the shortest stance duration, followed by flip-flop and shod conditions [15, 18]. In the present study, the Y-Sandal condition was associated with the shortest stance phase duration, whereas no significant differences were observed among the other three conditions. Y-Sandals differ from conventional flip-flops primarily in their thicker sole with a 15-mm heel-to-toe drop and the presence of arch support. However, previous studies have indicated that heel-to-toe drop does not affect stance duration [32, 33]. Cranage et al. also reported that sole hardness did not significantly affect cadence, step time, or stance phase duration in children wearing sandals during walking or running [34]. Therefore, further research is required to explain the shorter stance phase duration observed under the Y-Sandal condition.
The conventional flip-flop condition was associated with similar hip sagittal-plane ROM data to those observed under the barefoot condition but produced greater hip sagittal-plane ROM compared with the Y-Sandal and shod conditions. This finding differs from that reported in a previous study, which observed no differences in hip flexion or extension angles during walking under barefoot, flip-flop, and shod conditions [15]. However, limited research has examined hip kinematics during running in flip-flops. Hence, additional studies are required to better understand these movement characteristics. Consistent with the findings of a previous study, the present study revealed that the shod condition was associated with greater knee sagittal-plane ROM compared with the other footwear conditions. One possible explanation for this finding is that the loose structure of flip-flops may reduce knee flexion to prevent the footwear from becoming loose during movement. Although a previous study also reported greater knee flexion during IC under barefoot walking conditions [15], no significant differences in knee flexion during IC were observed among the footwear conditions in the present study. In addition, the barefoot condition was associated with the lowest knee flexion during late swing, which may be attributed to the absence of a footwear sole.
A previous study involving children reported that jogging in flip-flops produced greater ankle dorsiflexion and midfoot inversion during IC and produced greater midfoot plantarflexion during mid-stance and propulsive phases compared with barefoot jogging [17]. In the present study, the barefoot and conventional flip-flop conditions were associated with different movement patterns; they were associated with greater ankle plantarflexion compared with the Y-Sandal and shod conditions. This finding may be attributed to the combination of supportive features incorporated into the Y-Sandal and running shoes. Unlike the barefoot and conventional flip-flop conditions, which provide a flat, unstructured interface, the Y-Sandal incorporates multiple design features, including medial arch support, a 15-mm heel-to-toe drop, and enhanced strap stability, which may synergistically limit arch displacement and modulate ankle kinematics during movement.
Previous studies have reported no significant differences in frontal- or transverse-plane kinematics between barefoot and flip-flop walking conditions [15, 17]. In the present study, the shod condition tended to be associated with a less everted ankle position during running. This finding may be due to the arch support provided by the footwear, which can help prevent excessive foot pronation during movement. The greatest ankle eversion was observed under the Y-Sandal condition despite the presence of arch support. One possible explanation for this finding is that the loose structure of Y-Sandals causes difficulty in maintaining consistent foot positioning over the arch support during running. Furthermore, the barefoot condition was associated with lower forefoot internal rotation but produced greater ankle internal rotation in the transverse plane compared with the shod condition, which may indicate a more pronated foot posture during movement. This finding may be attributed to the unrestricted movement experienced during barefoot running. Joint ROM generally increased with speed, with exceptions in specific planes, indicating that different movement strategies are used to achieve higher running speeds. This pattern may contribute to the observed interaction effects in knee and forefoot transverse-plane ROM (Fig. 3).
This study also examined the activities of the TA, EDL and FHL muscles. TA activation peaked at IC and during mid-swing, reflecting eccentric control during the heel rocker phase and concentric dorsiflexion during the swing phase. The two flip-flop conditions were associated with the highest TA activation levels. The conventional flip-flop condition was also associated with greater ankle ROM. This result is consistent with the findings of Price et al., who hypothesized that increased TA activity may be required to maintain dorsiflexion and stabilize the footwear during the swing phase [14]. Furthermore, the two flip-flop conditions were associated with higher EDL and FHL activation compared with the barefoot and shod conditions. The peak in FHL activation near TO suggests that this muscle contributes to propulsion and helps stabilize the flip-flop against the foot. A previous study reported that strapped sandals result in a stronger relationship between stride mechanics and muscle activation compared with strapless sandals. Our results are consistent with the possibility that the loose fit of flip-flops imposes greater biomechanical demands during the gait cycle [31]. However, the barefoot condition was associated with the lowest FHL activation, consistent with the findings of a previous study indicating that footwear can increase propulsion-related muscle activity [11]. The variability in the timing of peak EDL activation across trials suggests that EDL might play a synergistic role in fine-tuning foot movement and stabilizing footwear throughout the gait cycle. Side-to-side differences in muscle activation were observed across running speeds; however, the interaction effects were inconsistent, suggesting that additional studies are required to clarify the nature of these asymmetries. Although the present study focused on distal musculature, previous findings regarding the gastrocnemius, biceps femoris, rectus femoris, and gluteus maximus remain inconsistent [31, 35, 36]. Accordingly, future studies should include these proximal muscles to determine how footwear-mediated changes at the foot and ankle affect the broader kinetic chain and overall gait stability.
Overall, our results suggest that footwear choice affects the distribution of mechanical demands across the lower limb. The greater ankle ROM and higher activation of the TA, FHL, and EDL muscles observed under the barefoot and conventional flip-flop conditions could indicate increased neuromuscular demands on the distal musculature. However, because this study did not measure joint moments or plantar pressure, it remains unclear whether these kinematic and muscle activation changes translate into increased tissue loading or long-term injury risk. Future research incorporating kinetic analyses is warranted to determine how these movement patterns relate to mechanical loading and clinical outcomes. By contrast, the greater knee ROM and transverse-plane rotation observed under the shod condition suggest a different movement strategy at the knee. These findings provide preliminary evidence of kinematic adaptations at the proximal joints in response to different footwear interfaces. However, because joint moments and ground reaction forces were not measured, this study could not determine how the observed changes in knee rotation and ROM affect the actual mechanical loading of the knee. Consequently, these observed kinematic differences should be interpreted as evidence of altered movement coordination rather than as direct evidence of changes in joint loading.
From a practical perspective, previous studies have demonstrated that increased cadence and decreased stance time can reduce vertical loading rates [37, 38]. Although the present findings suggest that the barefoot and Y-Sandal conditions may have induced these gait modifications, further investigation is required before translating these findings into clinical practice. Future studies should incorporate kinetic and plantar pressure measurements to better determine the relationship between the kinematic and EMG differences observed in this study and actual injury-related loading patterns. Such data would be valuable for developing clinical guidelines for gait retraining and footwear prescription.
This study has several limitations that should be considered. First, the participants wore their own running shoes under the shod condition. Although we confirmed that all shoes were standard daily trainers and excluded shoes with advanced features (e.g., carbon plates or extra-thick cushioning), the inherent variability in midsole hardness and heel-to-toe drop may have affected individual kinematic responses. Future studies should consider using a standardized neutral shoe model to further reduce interparticipant variability. Second, shoe uppers may have affected foot motion tracking because retro-reflective markers were not directly placed on the foot. Shoe volume can obscure actual foot movement and alter marker trajectories. Third, treadmill running may differ from overground running, which may limit generalizability. Forth, the participants had varying running abilities; thus, the same absolute running speeds likely represented different relative exercise intensities, which may have affected gait patterns. Finally, the use of least significant difference post hoc comparisons increases the risk of Type I errors because of the large number of variables analyzed. Although we report effect sizes and CIs to aid in the interpretation of these findings, the results should be considered preliminary and interpreted with caution until they are replicated in larger cohorts by using more conservative statistical thresholds.
Future studies should use standardized, well-documented shoe models and include kinetic and plantar pressure to better characterize the relationships among movement patterns, muscle activity, and loading characteristics. Additionally, researchers should evaluate practical interventions, such as cadence cues, strengthening programs targeting the dorsiflexors and intrinsic foot muscles, and tailored footwear, to determine their effects on biomechanics and clinical outcomes.
Conclusion
The findings of this study demonstrate that footwear design significantly affects lower-limb kinematics and neuromuscular strategies during running. Barefoot running was characterized by a higher cadence, whereas both flip-flop conditions were associated with significantly greater activation of the distal musculature (TA, EDL, and FHL), indicating increased neuromuscular demands during the gait cycle. Notably, the movement patterns observed with the Y-Sandal more closely resembled those observed during shod running. These findings suggest that the specific design features of thong-style footwear can modulate ankle and forefoot biomechanics, which may mitigate compensatory movement strategies typically associated with conventional flip-flops. Although these findings provide a biomechanical foundation for understanding footwear-specific adaptations, future studies incorporating kinetic analyses are essential to determine the long-term implications for tissue loading and injury risk.
Supplementary Information
Acknowledgements
This manuscript was edited by Wallace Academic Editing.
Abbreviations
- EMG
Electromyography
- TA
Tibialis anterior
- EDL
Extensor digitorum longus
- FHL
Flexor hallucis longus
- 3D
Three-dimensional
- MVC
Maximal voluntary contraction
- IC
Initial contact
- TO
Toe-off
- HEE
Heel
- FMH
First metatarsophalangeal joint
- VMH
Fifth metatarsophalangeal joint
- ROM
Range of motion
Authors’ contributions
CW Wong and HC Lin conceived and designed the study. CW Wong, YC Tseng, and CW Chen contributed to data acquisition. CW Wong, CM Chang, YC Tseng, CW Chen, and HC Lin contributed to data analysis. CW Wong, CM Chang, and HC Lin contributed to data interpretation and drafted the manuscript.
Funding
This work was supported by the National Science and Technology Council of Taiwan under Grant NSTC 112-2813-C-039-031-H and China Medical University under Grants CMU111-S-31 and CMU112-SR-28.
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
This study was approved by the Ethics Committee of the China Medical University Hospital (approval No.: CMUH113-REC3-036). The clinical trial was registered at ClinicalTrials.gov (identifier: NCT07003087) and was initially released on February 26, 2025. All participants provided written informed consent before enrollment.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Fitness, Trends. March report. https://runrepeat.com/fitness-trends. Accessed 1 2024.
- 2.Kakouris N, Yener N, Fong DTP. A systematic review of running-related musculoskeletal injuries in runners. J Sport Health Sci. 2021;10(5):513–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sun X, Lam WK, Zhang X, Wang J, Fu W. Systematic Review of the Role of Footwear Constructions in Running Biomechanics: Implications for Running-Related Injury and Performance. J Sports Sci Med. 2020;19(1):20–37. [PMC free article] [PubMed] [Google Scholar]
- 4.Relph N, Greaves H, Armstrong R, Prior TD, Spencer S, Griffiths IB, Dey P, Langley B. Running shoes for preventing lower limb running injuries in adults. Cochrane Database Syst Rev. 2022;8(8):Cd013368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bonanno DR, Landorf KB, Munteanu SE, Murley GS, Menz HB. Effectiveness of foot orthoses and shock-absorbing insoles for the prevention of injury: a systematic review and meta-analysis. Br J Sports Med. 2017;51(2):86–96. [DOI] [PubMed] [Google Scholar]
- 6.Roth J, Neumann J, Tao M. Orthopaedic perspective on barefoot and minimalist running. JAAOS - J Am Acad Orthop Surg. 2016;24(3):180–7. [DOI] [PubMed]
- 7.Hryvniak D, Dicharry J, Wilder R. Barefoot running survey: Evidence from the field. J Sport Health Sci. 2014;3(2):131–6. [Google Scholar]
- 8.Rixe JA, Gallo RA, Silvis ML. the barefoot debate: can minimalist shoes reduce running-related injuries? Curr Sports Med Rep. 2012;11(3):160–5. [DOI] [PubMed]
- 9.McDougall C. Born to Run: A Hidden Tribe, Superathletes, and the Greatest Race the World Has Never Seen. 1 ed. NewYork: Knopf Doubleday Publishing Group; 2009. [Google Scholar]
- 10.I’m a. foot doctor — this popular summer shoe is a scary health risk. https://nypost.com/2023/06/30/im-a-foot-doctor-this-summer-shoe-is-a-scary-health-risk/. Accessed 2 March 2024.
- 11.Péter A, Arndt A, Hegyi A, Finni T, Andersson E, Alkjær T, Tarassova O, Rönquist G, Cronin N. Effect of footwear on intramuscular EMG activity of plantar flexor muscles in walking. J Electromyogr Kinesiol. 2020;55:102474. [DOI] [PubMed] [Google Scholar]
- 12.Chen TL, Wong DW, Xu Z, Tan Q, Wang Y, Luximon A, Zhang M. Lower limb muscle co-contraction and joint loading of flip-flops walking in male wearers. PLoS ONE. 2018;13(3):e0193653. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sharpe T, Malone A, French H, Kiernan D, O’Brien T. Effect of flip-flops on lower limb kinematics during walking: a cross-sectional study using three-dimensional gait analysis. Ir J Med Sci. 2016;185(2):493–501. [DOI] [PubMed] [Google Scholar]
- 14.Price C, Andrejevas V, Findlow AH, Graham-Smith P, Jones R. Does flip-flop style footwear modify ankle biomechanics and foot loading patterns? J Foot Ankle Res. 2014;7(1):40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zhang X, Paquette MR, Zhang S. A comparison of gait biomechanics of flip-flops, sandals, barefoot and shoes. J Foot Ankle Res. 2013;6(1):45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Price C, Graham-Smith P, Jones R. A comparison of plantar pressures in a standard flip-flop and a FitFlop using bespoke pressure insoles. Footwear Sci. 2013;5(2):111–9. [Google Scholar]
- 17.Chard A, Greene A, Hunt A, Vanwanseele B, Smith R. Effect of thong style flip-flops on children’s barefoot walking and jogging kinematics. J Foot Ankle Res. 2013;6(1):8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Shroyer JF, Weimar WH. Comparative analysis of human gait while wearing thong-style flip-flops versus sneakers. J Am Podiatr Med Assoc. 2010;100(4):251–7. [DOI] [PubMed] [Google Scholar]
- 19.Carl TJ, Barrett SL. Computerized analysis of plantar pressure variation in flip-flops, athletic shoes, and bare feet. J Am Podiatr Med Assoc. 2008;98(5):374–8. [DOI] [PubMed] [Google Scholar]
- 20.Oku K, Kimura D, Ito T, Matsugi A, Sugioka T, Kobayashi Y, Satake H, Kumai T. Effect of increased flexor hallucis longus muscle activity on ground reaction force during landing. Life (Basel). 2021;11(7):630. [DOI] [PMC free article] [PubMed]
- 21.Reynard F, Dériaz O, Bergeau J. Foot varus in stroke patients: muscular activity of extensor digitorum longus during the swing phase of gait. Foot (Edinb). 2009;19(2):69–74. [DOI] [PubMed] [Google Scholar]
- 22.Hermens HJ, Freriks B, Merletti R, Stegeman D, Blok J, Rau G, Disselhorst-Klug C, Hägg G. SENIAM 8: european recommendations for surface electromyography. Roessingh Research and Development. 1999;8(2):13–54
- 23.Ball N, Scurr J. Electromyography normalization methods for high-velocity muscle actions: review and recommendations. J Appl Biomech. 2013;29(5):600–8. [DOI] [PubMed] [Google Scholar]
- 24.Williams JJ, Roshinski WC, Watso JC. Upper Leg Muscular Co-Contraction During Maximal-Speed Sprinting in Male Club Ice Hockey Athletes. Sports Med Open. 2025;11(1):1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Phillips DA, Paletta S, Perlet M. Voluntary contractions underestimate peak muscle activity in drop jumps. Int Biomech. 2025;12(1):1–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Commandeur D, Klimstra M, Brodie R, Hundza S. A comparison of bioelectric and biomechanical emg normalization techniques in healthy older and young adults during walking gait. J Funct Morphol Kinesiol. 2024;9(2):90. [DOI] [PMC free article] [PubMed]
- 27.The State of Running. 2019 https://runrepeat.com/state-of-running. Accessed 31 March 2024.
- 28.What is the average running speed?. https://www.runnersworld.com/uk/training/beginners/a44826741/average-running-speed/#what-is-the-average-time-for-a-5k. Accessed 1 March 2024.
- 29.Hollander K, Liebl D, Meining S, Mattes K, Willwacher S, Zech A. Adaptation of Running Biomechanics to Repeated Barefoot Running: A Randomized Controlled Study. Am J Sports Med. 2019;47(8):1975–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.McCarthy C, Fleming N, Donne B, Blanksby B. Barefoot running and hip kinematics: good news for the knee? Med Sci Sports Exerc. 2015;47(5):1009–16. [DOI] [PubMed] [Google Scholar]
- 31.Xuan B, Sun D, Wang D, Chen D, Li F, Song Y, Cen X, Fekete G, Jemni M, Gu Y. Biomechanical Effects of Sandal Strap Design on Gait Kinematics and Electromyographic Activation Patterns: A Speed-Dependent Analysis. Appl Bionics Biomech. 2025;2025:8802614. [DOI] [PMC free article] [PubMed]
- 32.Richert FC, Stein T, Ringhof S, Stetter BJ. The effect of the heel-to-toe drop of standard running shoes on lower limb biomechanics. Footwear Sci. 2019;11(3):161–70. [Google Scholar]
- 33.Moody D, Hunter I, Ridge S, Myrer JW. Comparison of Varying Heel to Toe Differences and Cushion to Barefoot Running in Novice Minimalist Runners. Int J Exerc Sci. 2018;11(1):13–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Cranage S, Perraton L, Bowles KA, Williams C. A comparison of young children’s spatiotemporal gait measures in three common types of footwear with different sole hardness. Gait Posture. 2021;90:276–82. [DOI] [PubMed] [Google Scholar]
- 35.Burgess KE, Swinton PA. Do FitflopsTM increase lower limb muscle activity? Clin Biomech Elsevier Ltd. 2012;27(10):1078–82. [DOI] [PubMed] [Google Scholar]
- 36.Maharaj JN, Barber L, Walsh HPJ, Carty CP. Flip-flops do not alter the neuromuscular function of the gastrocnemius muscle and tendon during walking in children. Gait Posture. 2020;77:83–8. [DOI] [PubMed] [Google Scholar]
- 37.Futrell EE, Gross KD, Reisman D, Mullineaux DR, Davis IS. Transition to forefoot strike reduces load rates more effectively than altered cadence. J Sport Health Sci. 2020;9(3):248–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Heiderscheit BC, Chumanov ES, Michalski MP, Wille CM, Ryan MB. Effects of step rate manipulation on joint mechanics during running. Med Sci Sports Exerc. 2011;43(2):296–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.