Introduction

Jump rope exercises have been shown to improve cardiovascular endurance, enhance limb coordination, and promote overall physical health1,2,3,4. Since becoming a competitive sport in 1973, jump rope has gained global recognition, with more than 60 countries establishing jump rope clubs and an increasing number of international competitions5,6. Among various techniques, the “speed step”, known as single-leg alternated foot jumps, involves rapidly alternating foot landings to allow the rope to pass underfoot. This technique is frequently used in speed competitions because it can achieve a higher number of jumps within a short period. Indeed, current world records for the most jumps in 30 s and 3 min were achieved using the speed step7.

Speed step requires repetitive force generation and impact absorption in the lower limbs. Current research on the speed step has primarily focused on ground reaction forces and joint power8,9. However, plantar loading during the speed step remains insufficiently characterized, limiting the development of footwear tailored to task-specific loading demands. Shek et al. reported preliminary evidence that peak plantar pressure during rope skipping is concentrated at the metatarsal heads and hallux9. However, their findings were based on a single elite participant and did not include measures of plantar force or propulsive-relevant indices. Therefore, it remains unclear whether this region-specific pressure pattern is consistent and robust across athletes during the speed step.

Footwear selection is a critical factor influencing athletic performance in addition to training. Embedding carbon fiber plates in the midsole increases the longitudinal bending stiffness of shoes, which has been shown to improve performance in activities such as sprinting, long-distance running, and vertical jumping10,11,12,13,14,15,16,17. Increased stiffness limits dorsiflexion at the metatarsophalangeal (MTP) joint, thereby reducing negative work loss during movement. Curved carbon fiber plates, in combination with other advanced footwear technologies, have been reported to further enhance performance18,19. Longman J suggested that the curved carbon fiber plates act like a slingshot, storing elastic energy by deforming under multi-directional forces and then releasing this energy during toe-off20. Additionally, Nigg et al. proposed the “teeter-totter effect” (class one lever mechanism), where the curved carbon plate’s geometry induces an upward reaction force at the heel when the forefoot contacts the ground, as evidenced by plantar pressure21,22,23. This may potentially reduce the muscular effort required, ultimately enhancing performance.

The purpose of this study was to investigate the plantar pressure characteristics of the speed step when wearing a regular jump rope shoe (control) and to evaluate the effects of curved carbon fiber plates with varying stiffness (stiff and stiffest) on plantar pressure and performance. We hypothesized that: (1) the forefoot region would exhibit significantly higher plantar loads (both pressure and force) than other regions during the speed step; (2) compared with the control, the stiff and stiffest conditions would show significantly higher heel load during ground contact; and (3) athletes would achieve better performance in the stiff and stiffest conditions.

Methods

Participants

A priori power analysis was performed using G*Power (version 3.1.9.7, Univ. Kiel, Germany), which indicated that a sample size of 28 participants was required for a one-way repeated measures ANOVA (medium effect size, f = 0.25; α = 0.05; power = 0.80). Inclusion criteria were: (1) elite jump rope athletes (≥ 120 repetitions in 30 s); (2) ≥ 3 h of regular jump rope training per week; (3) male athletes with European shoe sizes 42 or 43 to ensure proper shoe fit; and (4) no musculoskeletal, cardiovascular, or neurological disorders in the past 3 months. During the screening session, participants were required to submit a 30-second speed step video to confirm eligibility. All participants provided written informed consent, and the study was approved by the Ethics Committee of Shanghai University of Sport (Approval No.: 102772023RT029). 40 male professional jump rope athletes were initially recruited; four were excluded due to data quality issues, leaving 26 participants for the final analysis (Table 1).

Table 1 Basic information of participants.
Full size table

Experimental shoes

Three types of jump rope shoes with nearly identical appearances but different midsole constructions were used: (1) control, EVA foam midsole without a carbon plate; (2) stiff, EVA midsole embedded with a curved carbon plate of medium stiffness; and (3) stiffest, EVA midsole embedded with a high-stiffness curved carbon plate (Fig. 1). All shoes had identical uppers and midsole materials (except for the carbon plate) and were similar in weight (within 30 g).

Forefoot longitudinal stiffness was measured according to the Chinese national standard GB/T 32,023 − 2015 (“Footwear – Test methods for whole shoe – Rigidity of flexing area”). Briefly, each shoe was fixed at the heel on a testing platform with the forefoot placed on a flexion device. A bending moment was applied at the metatarsophalangeal joint until a 45° angular deflection was achieved, and the torque (N·m) required at 45° was recorded as forefoot rigidity. Toe-to-heel drop was measured following GB/T 3903.5–2011 (“Footwear – General test methods – Appearance quality”) (Table 2). All tests were conducted at 23 °C and 50% relative humidity, with three repeated measurements per shoe.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Experimental shoes. (a: control; b: stiff; c: stiffest).

Table 2 Specifications of experimental shoes.
Full size table

Instrumentation

The pedar®-X insole plantar pressure measurement system (Novel, Germany) was used to collect plantar pressure data. This system has been validated as a reliable tool for in-shoe plantar pressure assessment24,25. The pressure insoles are 1.9 mm-thick and contain 99 pressure sensors per insole, with a sampling rate set at 100 Hz. Before the experiment, the pedar®-X insoles were calibrated using the trublu® calibration device (Novel, Munich, Germany) along with a pressure pump, following the system manual. Calibration was verified by reapplying pressure to the insoles and comparing the applied pressure values with the actual displayed mean pressure, ensuring accurate calibration.

Procedure

Before the test began, participants changed into experimental vests and shorts. They performed the trials in a randomized order of the three shoe conditions. The plantar pressure signal acquisition system secured to the participant’s chest; flexible pressure insoles, matching the participant’s foot size, were placed inside the experimental shoes. The insoles were connected to the signal acquisition box via cables, which were secured to prevent interference during the test. Participants then completed a 5-minute warm-up routine, including walking, running, and rope-skipping at a self-selected pace. Following the warm-up, the signal acquisition box was paired with the PC via Bluetooth. Participants were instructed to lift their feet off the ground to zero the pressure insoles, eliminating any pre-load effect. For each condition, participants performed a 30‑second speed step using their personal ropes in a designated area while plantar pressure data were continuously recorded. An additional experimenter counted the number of jumps. Trials were repeated for all three shoe conditions, with a 5‑minute rest interval between conditions. If a participant tripped on the rope, the trial was stopped and repeated after recovery.

Data processing

Preliminary data extraction was performed using plantar pressure software. Ten consecutive rope-skipping ground contact phases from 20 to 30 s under each shoe condition exported in ASCII format for further analysis. The stance phase for each speed step cycle was identified from the plantar pressure data using a dual-threshold method: (1) initial contact was defined as the first frame where total plantar force reached ≥ 50 N with a distinct change in multiple sensors’ peak pressures; (2) toe-off was defined as the last frame before force fell below 50 N, accompanied by a local peak pressure change. All automatically detected stance phases were visually inspected and further confirmed to match the step analysis results from the pedar®-X recorder software. The plantar regions were then masked using a custom Python script (Python 3.10.6). Given that the heel plays a limited role in rope-skipping activities9, the plantar was divided into eight regions26,27,28(Fig. 2) as follows: forefoot–hallux (H), lesser toes (LT), medial forefoot (MedFF), middle forefoot (MidFF), and lateral forefoot (LatFF); midfoot–medial midfoot (MedMF) and lateral midfoot (LatMF); and heel–rearfoot (RF). The plantar pressure data from ten phases of the participant’s right foot were averaged, encompassing contact time, force, peak force, peak pressure, maximum peak pressure, and force–time integral across the eight regions and the entire foot. Force was calculated as the sum of pressure × sensor area for all sensors in a region at each time point. Peak force was defined as the maximum force within a region during stance, and the force–time integral as the total accumulated force over the stance phase. Peak pressure was the highest sensor pressure in a region at each time point, while maximum peak pressure referred to the highest of these peak pressure values across stance phase. The force data from the ten cycles were interpolated to 101 points and averaged for subsequent analysis.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Schematic of plantar pressure mask. H: hallux; LT: lesser toes; MedFF: medial forefoot; MidFF: middle forefoot; LatFF: lateral forefoot; MedMF: medial midfoot; LatMF: lateral midfoot; RF: rearfoot.

Statistics

Results are presented as mean ± standard deviation. The Shapiro–Wilk test was applied to assess normality. For normally distributed data, one-way ANOVA was used, whereas Friedman’s test was applied to non-normally distributed data (α = 0.05). For plantar pressure in the control condition during speed step, normality was not met across the eight plantar regions, so the Friedman test was used for comparison. In cases of significant differences, post-hoc analysis was conducted using the Wilcoxon test with Bonferroni–Holm correction29. For plantar pressure in the LT, MedMF, and RF regions during speed step under three conditions, normality was not satisfied, and the Friedman test was applied. Parameters from other regions were analyzed using one-way repeated ANOVA, with significant differences further examined using paired t-tests with Bonferroni–Holm correction29. Statistical analyses were conducted in Python (version 3.10.6). Repeated-measures ANOVA was performed using the AnovaRM function in statsmodels (v0.14.0), while Wilcoxon signed-rank tests were conducted with SciPy (v1.11.2). Time-series ground reaction force data were analyzed using one-dimensional statistical parametric mapping (SPM1d) provided by Todd Pataky’s open-source package spm1D (www.spm1d.org). SPM1d, based on random field theory, allows for statistical analysis of one-dimensional datasets after normalization. If the computed F or t values for the one-dimensional dataset exceed the critical threshold α for a smooth one-dimensional Gaussian distribution, the null hypothesis is rejected30. One-way repeated ANOVA in SPM1d was used to compare results across the three shoe conditions (α = 0.05). If a significant interval was identified, post-hoc analysis was performed using paired t-tests with Bonferroni correction.

Results

Performance

No significant differences were observed among the three shoe conditions in either the number of rope turns within 30 s (control: 120 ± 9.6; stiff: 123 ± 10.3; stiffest: 120 ± 13.8, p = 0.12, partial η² = 0.076) or contact time (control: 158 ± 19 ms; stiff: 157 ± 19 ms; stiffest: 158 ± 17 ms, p = 0.94, partial η² = 0.002).

Plantar distribution of speed step

Significant differences in peak force were observed across the eight plantar regions (p < 0.001, Kendall’s W = 0.799, Fig. 3). Mean peak force values were as follows: H (0.23 ± 0.09 BW), LT (0.27 ± 0.11 BW), MedFF (0.45 ± 0.11 BW), MidFF (0.56 ± 0.08 BW), LatFF (0.22 ± 0.06 BW), MedMF (0.12 ± 0.07 BW), LatMF (0.22 ± 0.12 BW), and RF (0.11 ± 0.13 BW). In pairwise comparisons, peak force in H and LT was significantly greater than in MedMF and RF (all p < 0.001). Peak force in MedFF was significantly greater than in H, LT, LatFF, MedMF, LatMF, and RF (all p < 0.001). Peak force in MidFF was significantly greater than all regions(all p < 0.001). Peak force in LatFF was significantly greater than in MedMF and RF (all p < 0.001). Peak force in MedMF was not significantly different from RF. Finally, peak force in LatMF was significantly greater than in MedMF and RF (all p < 0.001).

Significant differences in maximum peak pressure were observed across the eight plantar regions (p < 0.001, Kendall’s W = 0.861, Fig. 3). Mean maximum peak pressure values were as follows: H (290.27 ± 98.39 kPa), LT (177.78 ± 57.69 kPa), MedFF (338.24 ± 107.81 kPa), MidFF (321.21 ± 87.53 kPa), LatFF (178.74 ± 54.95 kPa), MedMF (86.24 ± 50.05 kPa), LatMF (97.76 ± 38.90 kPa), and RF (41.59 ± 31.53 kPa). In pairwise comparisons, maximum peak pressure in H was significantly greater than in LT, LatFF, MedMF, LatMF, and RF (all p < 0.001). Maximum peak pressure in LT was significantly greater than in MedMF, LatMF, and RF (all p < 0.001). Maximum peak pressure in MedFF and MidFF was significantly greater than in LT, LatFF, MedMF, LatMF, and RF (all p < 0.001). Maximum peak pressure in LatFF was significantly greater than in MedMF, LatMF, and RF (all p < 0.001). Maximum peak pressure in MedMF was significantly greater than in RF (p < 0.001). Finally, maximum peak pressure in LatMF was significantly greater than in MedMF and RF (all p < 0.001).

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Comparison of peak force and maximum peak pressure across plantar regions of speed step H: hallux; LT: lesser toes; MedFF: medial forefoot; MidFF: middle forefoot; LatFF: lateral forefoot; MedMF: medial midfoot; LatMF: lateral midfoot; RF: rearfoot; BW: body weight. ***: highly significant difference between regions (p < 0.001).

Plantar pressure across shoe conditions

Significant differences were found in peak force in total plantar (p < 0.05, partial η² = 0.133), LT (p < 0.05, Kendall’s W = 0.138) and MedMF (p < 0.01, Kendall’s W = 0.217) across shoe conditions (p < 0.05, Fig. 4). Post-hoc analysis indicated that the control condition had a significantly higher plantar peak force than the stiffest condition (Control: 2.16 ± 0.28 BW vs. Stiffest: 2.08 ± 0.31 BW, p < 0.05, Fig. 4). In the LT region, the control condition also showed significantly higher peak force than the stiffest condition (Control: 0.27 ± 0.11 BW vs. Stiffest: 0.25 ± 0.10 BW, p < 0.05, Fig. 4). In the MedMF region, the control condition had a significantly higher peak force than both the stiff and stiffest conditions (Control: 0.12 ± 0.07 BW vs. Stiff: 0.10 ± 0.07 BW, p < 0.01; vs. Stiffest: 0.10 ± 0.06 BW, p < 0.05, Fig. 4).

Significant differences in maximum peak pressure were observed in MedFF across shoe conditions (p < 0.05, Kendall’s W = 0.031, Fig. 5). Post-hoc analysis revealed that the control condition had significantly lower maximum peak pressure than the stiff condition (Control: 338.24 ± 107.81 kPa vs. Stiff: 362.38 ± 114.44 kPa, p < 0.05).

For force-time integral, significant differences were found in LT (p < 0.01, Kendall’s W = 0.188) and MedMF (p < 0.05, Kendall’s W = 0.160) across shoe conditions (Fig. 6). Post-hoc analysis showed that the control condition had a significantly higher force–time integral in the LT region compared with the stiffest condition (Control: 21.24 ± 10.12 N·s vs. Stiffest: 17.97 ± 8.37 N·s, p < 0.001, Fig. 6). In the MedMF region, the control condition exhibited a significantly higher force–time integral than the stiff condition (Control: 5.08 ± 4.04 N·s vs. Stiff: 4.16 ± 3.93 N·s, p < 0.01, Fig. 6).

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Comparison of peak force across plantar regions under shoe conditions. H: hallux; LT: lesser toes; MedFF: medial forefoot; MidFF: middle forefoot; LatFF: lateral forefoot; MedMF: medial midfoot; LatMF: lateral midfoot; RF: rearfoot; BW: body weight. *: significant difference between shoe conditions (p < 0.05); **: significant difference between shoe conditions (p < 0.01).

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Comparison of maximum peak pressure across plantar regions under shoe conditions. H: hallux; LT: lesser toes; MedFF: medial forefoot; MidFF: middle forefoot; LatFF: lateral forefoot; MedMF: medial midfoot; LatMF: lateral midfoot; RF: rearfoot. *: significant difference between shoe conditions (p < 0.05).

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

Comparison of force-time integral across plantar regions under shoe conditions. H: hallux; LT: lesser toes; MedFF: medial forefoot; MidFF: middle forefoot; LatFF: lateral forefoot; MedMF: medial midfoot; LatMF: lateral midfoot; RF: rearfoot. **: significant difference between shoe conditions (p < 0.01); ***: significant difference between shoe condition (p < 0.001).

Plantar force SPM analysis across shoe conditions

Significant differences in the force SPM results were found across shoe conditions in the LT, MedFF, and overall plantar (Fig. 7). Post-hoc tests indicated that, in the LT region, the control condition was significantly higher than the stiffest condition during 1%–71% and 82%–93% of the ground contact phase (p < 0.001). In the MedMF region, the control condition was significantly higher than the stiff condition during 36%–49% of the ground contact phase (p < 0.01) and significantly higher than the stiffest condition during 32%–52% of the ground contact phase (p < 0.001). For the overall plantar, the control condition was significantly higher than the stiffest condition during 29%–32% of the ground contact phase (p < 0.05). No significant differences were found in other regions.

Fig. 7
Fig. 7The alternative text for this image may have been generated using AI.
Full size image

Comparison of force time series across plantar regions under shoe conditions. Red rectangle: Significant difference interval between control and stiff; Blue rectangle: Significant difference interval between control and stiffest.

Discussion

Plantar pressure analysis showed that in the control condition, peak force was the highest in MidFF, followed by MedFF. The maximum peak pressure was highest in MidFF, MedFF, and H, followed by LT and LatFF. There were significant differences in the plantar pressure among shoe conditions. Compared with the control, the peak force and force–time integral in the stiff condition were significantly lower in MedMF, and the maximum peak pressure was significantly higher in MedFF. The stiffest condition had significantly lower peak force and force–time integral in the LT and significantly lower MedMF and plantar peak force than the control. The results of SPM of force showed that, compared with the control, the stiff showed a significantly lower force interval in MedMF during the ground contact period, and the stiffest condition had a significantly lower force interval in LT, MedMF, and overall plantar during the ground contact phase. We found no significant difference in speed step performance among the three shoe conditions. These results partially supported the hypotheses of this study.

Hypothesis 1 proposed that the forefoot would exhibit significantly greater plantar loads than the midfoot and heel regions during the speed step. This hypothesis was fully supported. Both maximum peak pressure and peak force differed significantly across plantar regions (maximum peak pressure: p < 0.001, Kendall’s W = 0.861; peak force: p < 0.001, Kendall’s W = 0.799). Maximum peak pressure was highest in the forefoot regions, particularly in the MedFF, MidFF, and H, was significantly greater than values in the midfoot (MedMF, LatMF) and heel (RF) (all p < 0.001). Similarly, peak force followed the order MidFF > MedFF > H ≈ LT ≈ LatFF ≈ LatMF > RF ≈ MedMF, with forefoot regions demonstrating greater values than the midfoot and heel. These results are consistent with previous observations of elevated forefoot loading during rope-skipping9. By quantifying both pressure and force across eight anatomically defined regions, the present study provides a more detailed characterization of regional loading patterns during the speed step. Regions with greater peak force, such as the MidFF and MedFF, are likely to contribute more to propulsion during the speed step. Moreover, although peak force was greater in the MidFF and MedFF, maximum peak pressure in the H was comparable to these regions, suggesting that the hallux should be considered in footwear design for the speed step.

Hypothesis 2 proposed that, compared with the control condition, the stiff and stiffest conditions would exhibit greater heel loading during ground contact. This hypothesis was not supported. No significant differences were observed in the rearfoot (RF) region across shoe conditions for peak force (p > 0.05, Kendall’s W = 0.028), maximum peak pressure (p > 0.05, Kendall’s W = 0.075), or force–time integral (p > 0.05, Kendall’s W = 0.045). According to the “teeter-totter effect” proposed by Nigg et al., increased forefoot stiffness may create a pivot point under the metatarsal region, theoretically producing an upward counterforce at the heel21. However, such an effect was not observed in the present study. A likely explanation is that the stiffness and toe spring angle of our experimental shoes were lower than those reported in commercially available curved-plated running shoes. Previous studies have shown higher bending stiffness values (e.g., > 20 N·m/rad at 30° deflection), whereas our experimental models did not reach comparable bending stiffness levels31,32. Similarly, preliminary toe spring measurements indicated smaller curvature angles than those reported for high-performance racing shoes33. Insufficient stiffness and reduced toe spring may therefore have limited the mechanical conditions necessary to elicit a measurable teeter-totter response.

Hypothesis 3 proposed that athletes would achieve better performance in the stiff and stiffest conditions. This hypothesis was also not supported. No significant differences in speed step performance were observed among shoe conditions, as reflected by jump count (p = 0.12, partial η² = 0.076,) and ground contact time (p = 0.94, partial η² = 0.002). The absence of performance enhancement may partly relate to the lack of a pronounced stiffness. Additionally, the speed step is a technically demanding task requiring high-frequency coordination between lower-limb force production and upper-limb rope turning, and may also be influenced by neuromuscular control34,35,36. As Bruce et al. reported, jump rope performance is jointly constrained by these two subsystems, making inter-limb coordination a critical determinant of maximal speed8,37.

Importantly, although performance did not improve, the stiffest condition reduced overall plantar peak force (p < 0.05, partial η² = 0.133), primarily driven by decreases in the LT (p < 0.05, Kendall’s W = 0.138) and MedMF (p < 0.01, Kendall’s W = 0.217). Force–time integral was likewise reduced in these regions under stiffer plate conditions (all p < 0.05). These findings suggest that comparable performance can be maintained under lower localized mechanical loading. While injury outcomes were not assessed, reduced localized loading and cumulative mechanical exposure may have implications for mechanical stress regulation during repetitive speed step. Thus, in this context, the effects of curved carbon plates may relate more to altered load distribution than to direct performance enhancement.

Limitations

This study examined plantar pressure characteristics during the speed step technique and evaluated the effects of curved carbon-fiber plates with varying stiffness on plantar loading and performance. Several limitations should be noted:

  • External load device (< 1 kg): All participants wore the same device in every trial, which may have slightly increased the mechanical demand and thus plantar loading (e.g., forces/impulses) compared with skipping without the device. However, as the device mass was ~ 1% of body mass, its impact on the absolute magnitude is expected to be small, and the reported values should be interpreted as a close approximation of plantar loading under typical speed step conditions. Moreover, because the device was identical across all shoe conditions within each participant, it is unlikely to have systematically biased the between-condition comparisons of plate stiffness.

  • Rope variability (length/weight): Participants used their habitual ropes, with rope length self-optimized to their anthropometrics and technique, an essential prerequisite for stable elite performance. Enforcing a standardized rope could artificially disrupt timing and landing strategy, introducing non-footwear-related changes in plantar loading. Because rope settings were held constant within each participant across all footwear trials, non-standardized rope is unlikely to have influenced the within-subject comparisons between shoe conditions. Accordingly, the reported plantar loading should be representative of elite speed step conditions.

  • Sex-specific sample: Only male elite athletes were recruited, as the performance criterion (≥ 120 repetitions in 30 s) made female participation extremely difficult; thus, findings may not generalize to females.

  • Performance level: Results may also not extend to non-elite or recreational athletes, whose skipping strategies and plantar load patterns likely differ.

Conclusion

During the speed step, peak force was highest in the medial and middle forefoot, whereas forces at the hallux, lesser toes, lateral forefoot, and lateral midfoot were similar. Maximum peak pressure was significantly higher at the hallux and 1st–3rd metatarsophalangeal joints, indicating that these areas experience the greatest mechanical stress and likely play a key role in propulsion. Although curved carbon fiber plates with varying stiffness did not improve athletic performance, the stiffest plates notably reduced force in the lesser toes and medial midfoot regions, and decreased overall plantar force. Therefore, stiffest curved carbon fiber plates may be beneficial for jump rope athletes by helping to redistribute plantar loads.