Journal Logo

APPLIED SCIENCES

Effect of Shoe and Surface Stiffness on Lower Limb Tendon Strain in Jumping

FIRMINGER, COLIN R.1,2,3; BRUCE, OLIVIA L.1,3; WANNOP, JOHN W.1; STEFANYSHYN, DARREN J.1,2; EDWARDS, W. BRENT1,2,3

Author Information
Medicine & Science in Sports & Exercise: September 2019 - Volume 51 - Issue 9 - p 1895-1903
doi: 10.1249/MSS.0000000000002004

Abstract

Tendinopathy is defined as an overuse injury of the tendon characterized by localized pain and reduced function. This injury, which often requires extensive rest and interruption from sport participation, is frequently observed at the Achilles (AT) and patellar tendons (PT) of athletes participating in sports associated with frequent bouts of jumping and running activity (1). Previous epidemiological studies examining lower limb injuries among volleyball, soccer, and basketball athletes have reported a prevalence of 18% and 45% for Achilles and patellar tendinopathy, respectively (1,2).

Many athletes have been forced to prematurely retire from their athletic careers because of symptoms of tendinopathy, which can last more than 15 yr after diagnosis (3). Although several conservative treatments exist, including rest, eccentric training programs, and shockwave therapy, surgical intervention may ultimately be required. Unfortunately, all treatments for tendinopathy, either conservative or surgical, have exhibited highly variable outcomes in terms of their efficacy (4,5). The inadequacy of current tendinopathy treatments highlights the need to reduce the incidence of tendinopathy through primary prevention strategies.

During jumping and running, AT and PT experience submaximal repetitive loading at magnitudes in excess of six body weights (6,7). This scenario, known as fatigue loading, has been characterized by microdamage accumulation (in the form of kinked collagen fibers and matrix disruptions) in conjunction with a biological cascade of proinflammatory cellular activity (8,9). The accumulation of microdamage within tendon, and therefore the number of loading cycles it may endure before failure (i.e., fatigue life), is highly dependent on the resulting strain from the applied mechanical load (10). In fact, the fatigue life of tendon decreases exponentially with strain magnitude (11,12), suggesting that small reductions in tendon strain may greatly lessen the likelihood of tendinopathy.

Changes in shoe midsole stiffness and surface construction represent two potential interventions to reduce the risk of AT and PT overuse injury associated with jumping. An epidemiological study reported that elite volleyball athletes training on a wooden surface had a lower prevalence of patellar tendinopathy when compared with those training on concrete (13). Significant reductions in peak ground reaction force and loading rate have also been reported when landing in shoes with a less stiff midsole, or on surfaces with greater shock absorption (14). The latter findings may be meaningful because lower peak ground reaction force and loading rates have been associated with a reduced risk of lower limb injury in runners (15–17). Although these results infer that shoe midsole stiffness and surface construction may alter injury risk in jumping, their effect on AT and PT strains during this activity has not been examined.

The purpose of this study was to investigate the effect of shoe stiffness and surface construction on AT and PT strains during the takeoff and landing phases of maximal countermovement jumps in basketball athletes. To this end, we recruited participants to perform maximal countermovement jumps in three shoes and on three surfaces, each with different stiffness properties. The resulting AT and PT strains were calculated using musculoskeletal modeling combined with participant-specific measurements of tendon geometry and material properties. In accordance with previous epidemiological and biomechanical findings, we hypothesized that less stiff shoes and surfaces would reduce AT and PT strains during jumping.

METHODS

Study population

Thirty male high school and collegiate-level basketball athletes (18.9 ± 3.4 yr, 1.90 ± 0.10 m, 77.6 ± 12.7 kg) were recruited from the greater Calgary area. Participants were injury free, defined as having no trauma or pain limiting sport participation within the previous 3 months of testing. Ethics approval was obtained from the Conjoint Health Research Ethics Board at the University of Calgary, and all participants provided written informed consent before testing.

This sample size was chosen after an a priori power analysis based on preliminary data from six university basketball players performing maximum countermovement jumps in two shoes with different midsole stiffness properties (184 vs 130 N·mm−1). Using a combined inverse dynamics and musculoskeletal modeling approach (see Methodological Overview section), peak PT strains during landing were found to be 15% greater in the stiffer shoe. This difference corresponded to a moderate effect size of d = 0.60, which would require a minimum sample size of n = 24 to detect a significant difference between footwear conditions with a criterion α = 0.05 and β = 0.2 (power = 0.8).

Methodological overview

Participants attended data collection sessions on two separate days. In the first session (day 1), images were captured with magnetic resonance imaging (MRI) to quantify select morphological features (Fig. 1A). Dynamometry, ultrasound images, and EMG were collected in the second session (day 2) as participants performed submaximal and maximal isometric contractions at the ankle and knee to quantify AT and PT stiffness (Fig. 1B). Immediately after this, participants performed maximal countermovement jumps in three shoes × three surfaces while motion capture and force platform data were recorded (Fig. 1C). A musculoskeletal model was then used to quantify AT and PT force during the countermovement jumps (Fig. 1D), which were converted to strain.

FIGURE 1
FIGURE 1:
Overview of study methodology to obtain AT and PT strain. A, Sagittal knee MRI of a representative subject with PT moment arm (d PT). B, Sagittal knee ultrasound image with the tibia, PT, and patella visible. C, Motion capture setup with a participant performing a countermovement jump. D, Musculoskeletal models were used to calculate AT force (FAT) and PT force (FPT). Data from panel A were collected in the first testing session (day 1), whereas data from panels B and C were collected in the second testing session (day 2). Data from each methodology were combined to calculate AT and PT strain.

MRI imaging

A sagittal MRI scan of the right knee was obtained for each participant using an Optima MR430s 1.5 T scanner (GE Healthcare, Chicago, IL). Participants were seated upright, and their right leg was placed in the scanner with their knee at 10° of flexion. Images were collected using a PD-weighted, fast spin-echo sequence (echo time = 30.8 ms, repetition time = 3500 ms, acquisition matrix = 320 × 256, flip angle = 90°, field of view = 140 mm, 22 slices, slice thickness = 3.5 mm, scan time = 4.5 min).

PT resting length was calculated by connecting a straight line between the patellar and the tibial insertions of the tendon. PT moment arm was calculated geometrically using the tibiofemoral contact point method (18).

Dynamometry/ultrasound/EMG testing

A combined dynamometry/ultrasound/EMG testing protocol was used to determine AT resting length, moment arm, stiffness, and PT stiffness. AT resting length was first measured as participants stood upright. The locations of calcaneal insertion and musculotendinous junction were identified using ultrasound and marked on the skin, and the distance between these two points was measured with a tape measure. Next, participants were seated upright in a Biodex System 3 dynamometer (Biodex Medical, Shirley, NY), and the ankle joint axis (defined as a line connecting the medial and lateral malleoli) was visually aligned with the dynamometer axis. The ankle was positioned at 0° (neutral position) with the knee at 60° of flexion. Next, straps were placed around each participant’s waist, thigh, and foot to secure the foot to the footplate and limit any limb motion during trials. Three pairs of EMG electrodes with a conductive adhesive hydrogel (Kendall 100 series; Medtronic, Minneapolis, MN) were placed over the medial gastrocnemius (MG), soleus, and tibialis anterior (TA) at SENIAM-suggested locations. The interelectrode distance between each pair of electrodes was 30 mm. The AT moment arm was calculated using the tendon excursion method (19) as the ankle was passively rotated from 0° to 30° of plantarflexion. EMG signals were monitored to ensure no muscles were active during the passive rotation trial.

AT stiffness was measured with participants still seated in the dynamometer as described above. Participants performed two warm-up ankle plantarflexion trials at a self-estimated 50% of maximum voluntary contraction (MVC). Participants then performed 5-s ramped isometric plantarflexion contractions at three different forces in the following order: trials 1 and 2, MVC; trials 3 and 4, 80% MVC; and trials 5 and 6, 50% MVC. Two 5-s ramped isometric 100% MVC dorsiflexion trials (trials 7 and 8) were subsequently recorded to correct for any antagonist (i.e., TA) coactivation during the plantarflexion trials (see below). After each submaximal and maximum trial, rests were given (1 and 2 min, respectively) to minimize any fatigue effects.

PT stiffness was measured while participants were seated in the dynamometer with their knee at 70° of flexion. The knee joint axis (defined as a line connecting the medial and the lateral epicondyles) was visually aligned with the dynamometer, and straps were placed around each participant’s hips, thigh, and ankle to limit any movement during the isometric contractions. Pairs of electrodes with a 30-mm interelectrode distance were placed over the vastus lateralis, vastus medialis, and biceps femoris (BF) muscles as per the SENIAM-suggested sensor locations. After two self-estimated 50% MVC extension warm-up trials, participants performed 2 × MVC (trials 1 and 2), 2 × 80% MVC (trials 3 and 4), and 2 × 50% MVC (trials 5 and 6) trials in extension followed by 2 × MVC flexion (trials 7 and 8) to assess antagonist coactivation.

Ultrasound video was recorded during each trial at 78 Hz using a Logiq E9 ultrasound (GE Healthcare). The ultrasound parameters were as follows: gain = 50, depth = 3.0 cm, frequency = 13 MHz. EMG and torque data were recorded simultaneously at 2000 Hz. The baseline offset was removed from the EMG data, which was then full-wave rectified and filtered with a zero-lag low-pass fourth-order Butterworth filter with a cutoff frequency of 3 Hz (20). Processed EMG signals were then normalized to the maximum EMG activation from all trials within each muscle. The net maximum torque for each trial and the corresponding normalized EMG signal for the three muscles were then determined for each force condition. The normalized EMG data of the two extensor muscles (MG and soleus for the ankle, vastus lateralis and vastus medialis for the knee) were summed to create a single EMG value for the plantarflexor muscle group. Coefficients relating torque to normalized EMG signals were calculated using least squares optimization according to (21):

where Ti is the torque for trial i, EMGext,i is the normalized maximum EMG signal for the extensor muscle group for a given trial, EMGflex,i is the normalized maximum EMG signal for the flexor muscle (TA for the ankle and BF for the knee) for a given trial, ρextensors is the EMG coefficient for the extensor muscle group, and ρflexor is the EMG coefficient for the flexor muscle. The torque during the trials corrected for antagonistic co-contraction was calculated using equation 2:

where EMGant is the normalized EMG activation for the antagonist muscle (i.e., TA for ankle plantarflexion, BF for knee extension), and ρant is the EMG–torque coefficient of the antagonist muscle. Tendon force during the maximum isometric extension trial was calculated by dividing the corrected torque with the moment arm.

AT elongation was measured by manually tracking the myotendinous junction between the AT and the MG using ImageJ software (National Institutes of Health, Bethesda, MD). PT elongation was calculated using modified MATLAB region-tracking code downloaded from MathWorks File Exchange (MathWorks, Natick, MA). This code automatically tracked user-defined regions of interest on the patella and tibia using a Kanade–Lucas–Tomasi algorithm from the MATLAB Computer Vision System Toolbox (22). PT elongation was calculated between each frame as the difference between the current and the initial patellar/tibial position. A second- or third-order polynomial was respectively fit to the AT and PT force versus elongation data. Tendon stiffness was calculated from the slope of the fitted curve from 90% to 100% of tendon maximum force (23). A reliability study was performed to test the intra- and interday reliability of tendon stiffness measurements (see Text, Supplemental Digital Content 1, which outlines the reliability study, http://links.lww.com/MSS/B581).

Motion capture session

Participants performed maximal countermovement jumps in each of the nine shoe–surface combinations. The prototype shoes had identical uppers, outsoles, and dimensions but were fabricated using a different midsole material giving each shoe a different compressive and bending stiffness (adidas AG, Herzogenaurach, DE) (see Table 1). The shoe with a 70C ethylene vinyl acetate (EVA) midsole was the stiffest, followed by the shoes with 55C EVA and thermoplastic polyurethane (TPU) midsoles, respectively. The three surfaces varied in subfloor construction and spanned the range of stiffness values that would be typically used in a professional arena (Robbins Sports Surfaces, Cincinnati, OH) (see Table 1, Fig. 2). The surface with laminated foam (LF) subfloor was the stiffest, followed by the metal sleeper (MS) subfloor surface and the wooden sleeper (WS) subfloor surfaces, respectively.

TABLE 1
TABLE 1:
Shoe and surface material properties.
FIGURE 2
FIGURE 2:
Three-dimensional representation of each surface type. All surfaces had identical top layers with different subfloor construction. (Images courtesy of Robbins Sports Surfaces).

Each surface was bolted to a pair of force plates (Kistler Instrument Corp., Winterthur, CH), and kinetic data were collected at 2400 Hz. An identical surface was positioned adjacently to make the total surface area 2.16 m2. All jumps were performed with the participant’s right foot on the surface positioned over the force plates and their left foot over the adjacent surface.

Twenty-three retroreflective markers were placed on the right lower limb of each participant, and kinematic data were recorded at 240 Hz using eight high-speed video cameras (Eagle cameras; Motion Analysis Corporation, Santa Rosa, CA). A static motion capture trial was collected for each shoe–surface condition as the participants stood upright to establish a neutral anatomical coordinate system for each segment. Participants were instructed to perform maximal countermovement jumps and land in a controlled fashion. Vertical jump height was measured using a Vertec jump height measurement tool (Sports Imports, Hilliard, OH). Participants were given a 15-min rest between the end of the dynamometry/ultrasound/EMG session and the beginning of the motion capture session. Before recording, participants performed a minimum of five warm-up jumps to account for any preconditioning effect, as shown by Maganaris (24). Next, each participant performed five jumps per shoe on each of the three sample surfaces, resulting in a total of 45 jumps. Participants were also given a minimum of 5 min of rest in between each condition. The order of each shoe and surface condition was counterbalanced to reduce bias.

Data reduction

Raw data from each individual force plate were combined to calculate a single center of pressure (COP) using custom MATLAB code (https://isbweb.org/software/movanal/vaughan/kistler.pdf). Force platform data were downsampled from 2400 to 240 Hz, and both motion capture and force platform data were filtered using a low-pass zero-lag fourth-order Butterworth filter with a cutoff frequency of 22 Hz, which retained 95% of the raw signal power. Each countermovement jump was split into takeoff and landing phases. The takeoff phase started at the instant the sacrum marker illustrated a negative (downward) velocity and ended when the ground reaction force was less than 30 N. Similarly, the landing phase began when the ground reaction force reached 30 N and ended when the landing phase impulse was equal to the takeoff impulse. Cardan angles were used to describe segment and joint motion, which were calculated using a flexion–extension, abduction–adduction, internal–external rotation sequence (25). The equations of Vaughan et al. (26) were used to obtain segment masses, center of mass locations, and moments of inertia for the thigh, shank, and whole foot. Intersegmental joint moments for the ankle and knee joint were calculated using an inverse dynamics approach (25).

A musculoskeletal model of the lower limb was used to calculate AT and PT force during jumping (27). The resulting stress in the plantarflexor muscles (Sp) was first calculated from the sagittal ankle joint moment (Ma) from inverse dynamics analysis:

where PCSAMG and PCSAsoleus are the physiological cross-sectional areas of the MG and soleus muscles obtained from Arnold and Delp (28), and dAT is the AT moment arm corrected for ankle angle (29) obtained from the tendon excursion method. Force in the AT (FAT) was then computed directly from Sp:

Force in the PT (FPT) was calculated from sagittal knee joint moment (Mk) after accounting for antagonist co-contraction of the biarticular gastrocnemius and hamstring muscles. First, stress in the hip extensor muscles (Se) was estimated by dividing the extensor hip joint moment (Mh) by the sum of the product of each muscle’s physiological cross-sectional area and moment arm:

where dham,hip and dGM are the moment arms of the hamstrings and gluteus maximus about the hip joint as a function of hip joint angle, respectively (30). PCSAham and PCSAGM are the physiological cross-sectional areas of the hamstring muscles (i.e., BF long head, BF short head, semitendinosus, and semimembranosus) and gluteus maximus muscles, respectively (31). Next, PT force was calculated as follows:

where dMG and dham,knee are the moment arms of the MG and hamstring muscles about the knee joint center. Values for dMG and dham,knee were a function of sagittal knee joint angle (32,33). The PT moment arm (dPT) was obtained from MRI data and was adjusted as a function of knee angle throughout the jumping movement using previously published data (34). The AT strain (εAT) and the PT strain (εPT) were directly estimated from FAT and FPT, respectively:

where kAT and kPT are AT and PT stiffness, respectively, and lAT and lPT are the AT and PT resting lengths, respectively.

Statistical analysis

The primary outcomes for this study were peak AT and PT strains during the takeoff and landing phases of each of the nine shoe–surface conditions. Strains were trial averaged, and statistical analysis was performed using SPSS software (SPSS Inc., Chicago, IL). A 3 × 3 repeated-measures ANOVA was used to test for the main effects of shoe stiffness and surface construction on peak AT and PT strains with the criterion alpha level set to α = 0.05. Paired t-tests (α = 0.05) with a Bonferroni correction were used to investigate pairwise differences if main effects of shoe or surface were found. Differences in jump height were also tested to investigate whether the shoes or surfaces affected jump performance. Cohen’s d effect sizes were also calculated for any significant pairwise differences. To aid in the interpretation of any significant findings in peak tendon strains, additional 3 × 3 ANOVA was performed to examine the trial-averaged COP location, peak vertical ground reaction force magnitude, and sagittal plane foot and shank angles at the instant of peak strain.

RESULTS

AT strain

No significant interactions were observed between shoe and surface conditions for peak AT strain during takeoff (P = 0.343) or landing (P = 0.961). There was also no significant effect of shoe stiffness (P = 0.840) or surface construction (P = 0.884) on peak AT strain during takeoff.

A significant effect of shoe stiffness was observed for peak AT strain during landing (P = 0.028, see Table 2, Figs. 3 and 4). Pairwise comparisons between shoes indicated that the 70C EVA shoe had significantly greater peak AT strain than the TPU shoe (P = 0.021, d = 0.14). Post hoc analysis performed at the instant of peak AT strain revealed an 8-mm anterior shift in COP location in the 70C EVA shoe compared with the TPU shoe (P < 0.001, d = 0.29) and a 0.6° increase in foot angle (i.e., more plantarflexed) in the 70C EVA shoe compared with the TPU shoe (P = 0.004, d = 0.15). Shoe stiffness had no effect on peak vertical ground reaction force at the instant of peak AT strain (P = 0.340).

TABLE 2
TABLE 2:
Estimated marginal means for peak tendon strains during the takeoff and landing phases of a maximum countermovement jump.
FIGURE 3
FIGURE 3:
Time series graphs of PT (top) and AT (bottom) strain throughout the landing phase of a representative participant. Left graphs show shoe effects, whereas right graphs show surface effects. Shaded areas represent ±1 SD.
FIGURE 4
FIGURE 4:
Participant-averaged percent differences in peak AT strain. Shoe condition differences are shown to the right, whereas surface condition differences are shown to the left. For each bar, a positive value represents a larger strain in the first labeled condition compared with the second. (*P < 0.05).

Surface construction also had a significant effect on peak AT strains during landing (P = 0.021, see Table 2, Figs. 3 and 4). Pairwise comparisons indicated that the WS and the MS surfaces were associated with significantly greater peak AT strains compared with the LF surface (WS surface: P = 0.047, d = 0.16; MS surface: P = 0.029, d = 0.22). Post hoc analysis of the COP location at the instant of peak AT strain approached significance (P = 0.064), with the COP location tending to shift anteriorly on the MS surface compared with the LF surface (P = 0.110). A significant effect of surface existed for peak vertical ground reaction force at the instant of peak AT strain (P = 0.017), whereas no differences in foot angle were observed (P ≥ 0.580).

PT strain

No significant interactions were observed between shoe and surface conditions for peak PT strain during takeoff (P = 0.303) or landing (P = 0.326). There were also no main effects of shoe stiffness (takeoff: P = 0.114; landing: P = 0.295) or surface construction (takeoff: P = 0.501; landing: P = 0.170, see Table 2, Fig. 3) on peak PT strain.

Jump height

No significant interactions were observed between shoe stiffness and surface construction for jump height (P = 0.366). Furthermore, no significant effects of shoe stiffness or surface construction on jump height were observed (P > 0.243).

DISCUSSION

The purpose of this study was to examine the influence of shoe stiffness and surface construction on peak AT and PT strains during maximal countermovement jumps. Peak AT strains were significantly smaller when landing in the least stiff shoe (TPU) compared with the stiffest shoe (70C) regardless of surface, and peak AT strains were significantly smaller when landing on the stiffest surface (LF) compared with the other two surfaces (WS and MS) regardless of shoe. No effects of shoe or surface were observed for AT strains during takeoff, PT strains during takeoff and landing, or jump height. These results indicate that shoe stiffness and surface construction represent two promising and practical methods to reduce AT strain and the risk of Achilles tendinopathy during jumping activity.

The peak AT strain values calculated in this study were comparable with the AT strains of 6.2%–10.3% measured during single leg hopping (35). The peak PT strains in our study were slightly lower than the 6.6% average PT strain measured using cine phase contrast sagittal MR images at submaximal effort levels (36). This difference may be due to our population being competitive athletes with relatively high PT stiffness properties (2716 ± 795 N·mm−1) when compared with untrained individuals (2187 ± 713 N·mm−1) (37).

No effects of shoe stiffness or surface construction were observed at the knee. One potential reason for this finding is because the differences in shoe stiffness and surface construction were not large enough to necessitate kinetic and/or kinematic changes at the knee joint. However, one study examining changes in knee kinematics and kinetics during shod versus barefoot single leg drop landings only found small changes in peak flexion angle and peak moment (38). The use of shod versus barefoot conditions represents a much larger change in “shoe stiffness” compared with our study; therefore, we would speculate that even with a larger range of shoe stiffness and/or surface properties, differences at the knee joint still may not be observed. Another potential reason could be the two-legged movement that was analyzed in our study. Several basketball movements (such as a lay-up or a block) involve taking off and/or landing on one leg. These single leg movements may cause the participant to use the shoe and surface to a larger extent and may produce kinematic or kinetic differences at the knee between surface or shoe conditions.

Estimates of AT strain obtained from the musculoskeletal model are a function of the ankle joint moment (Ma) from inverse dynamics. For this reason, additional analyses were conducted to examine changes to the main components of Ma, namely, the magnitude of the vertical ground reaction force, the location of the ground reaction force (i.e., COP), and the foot segment angle. At the instant of peak AT strain, the COP was 8 mm further from the heel, and the foot angle was 0.6° more plantarflexed in the stiffest (70C EVA) shoe when compared with the least stiff (TPU) shoe. These results would counteract each other, as a more anterior COP would augment Ma by increasing the moment arm of the vertical ground reaction force in relation to the ankle joint center, whereas a more plantarflexed foot would reduce Ma by decreasing the distance between the vertical ground reaction force and the ankle joint. However, using a participant-averaged foot length of 29.1 cm, we calculated that an increase in foot plantarflexion of 0.6° corresponded to a decrease in horizontal distance of 0.2 mm, much smaller than the observed 8 mm increase in COP location. This result is supported by Farris et al. (2016), who showed that effective mechanical advantage of the triceps surae–AT muscle–tendon unit was highly dependent on the COP location (39). In terms of our study, a more anterior COP location observed in the 70C EVA shoe would reduce the effective mechanical advantage of the AT, resulting in increased plantarflexor demands (and greater AT strain) for a given ground reaction force magnitude.

The observed change in COP location may be due to an increased bending stiffness in the 70C EVA shoe (see Table 1), which has been shown by Stefanyshyn and Nigg (40) to decrease the amount of metatarsophalangeal (MTP) joint dorsiflexion during takeoff. Assuming decreased MTP joint dorsiflexion occurs during landing as well, the COP location would shift forward as the toes are able to dorsiflex less. Increased footwear bending stiffness tends to improve performance during jumping by reducing neuromuscular fatigue (41) and negative work at the MTP joint (40); however, the results of this study suggest that increased midsole bending stiffness may not be beneficial from an AT injury perspective.

Contrary to the shoe stiffness results, we observed that landing on the stiffer (WS and MS) surfaces produced larger AT strains than landing on the less stiff (LF) surface. For the LF surface, the COP location was slightly closer to the ankle joint compared with the other two conditions. In addition, the peak vertical ground reaction force at the instant of peak strain was slightly reduced. The combination of these two factors would generate a smaller ankle plantarflexion moment, resulting in smaller peak AT strain. These results are counterintuitive to our hypothesis; however, this leads us to believe that the observed differences in AT strain may be related to surface properties other than compressive stiffness. Previous research has reported that surfaces with greater shock absorption properties lowered vertical instantaneous loading rate during jumping; however, no effect was observed for peak vertical ground reaction force (14). The LF surface had a greater energy loss compared with the other two surfaces (see Table 1), which may have allowed more energy to be absorbed by the surface and less by the participant. This observation suggests that surface properties other than compressive stiffness may be more relevant when investigating lower extremity loading during jumping.

The results of this study have elucidated several significant differences in peak AT strain between shoe and surface conditions; however, the practical relevance of these differences may not be readily understood. It is therefore beneficial to relate AT strains to a more relevant metric for an athlete or clinician such as fatigue life, or the number of cycles to failure. Mechanical fatigue test data of cadaveric AT (12) suggest that the 5.3% decrease in AT strain between the TPU and the 70C EVA shoes would correspond to a 64% increase in fatigue life; the 5.7% decrease in AT strain between the WS and the LF surfaces would correspond to a 67% increase in fatigue life, and the 8.1% decrease in strain between the MS and the LF surfaces would correspond to a 92% increase in fatigue life. Assuming an average of 46 jumps per basketball game (42), these increases in fatigue life would correspond to 17, 23, and 29 extra games before injury for each of the aforementioned comparisons. Therefore, it is important to note that although the decreases in strain found in this study may seem relatively trivial, they may produce meaningful changes in tendinopathy risk.

When investigating the outcomes of this study, the relative importance of injury prevention versus performance improvement needs to be weighed. Previous research has shown that increased shoe bending stiffness can significantly improve jump height (40) and running economy (43); however, our results illustrate that these performance gains may come at the expense of increased Achilles tendinopathy risk. In our study, however, we did not observe any difference in jump performance between shoe conditions. This difference in results may be due to larger range of bending stiffness among shoes in the study by Stefanyshyn and Nigg (40) (0.04–0.38 N·m·deg−1; 160% difference) compared with our study (14.0–20.3 N·mm−1; 36% difference). Although a trade-off may exist between AT strain reduction and jump height improvement, we have shown that shoe bending stiffness may influence AT strain without affecting jump performance. On the basis of these results, future work should investigate the effect of jumping in a range of forefoot bending stiffness on AT strain and whether a performance trade-off exists.

CONCLUSIONS

AT strains were significantly influenced by shoe stiffness and surface construction. In particular, landing in a less stiff shoe or on a stiffer surface increased fatigue life between 64% and 92%, allowing athletes to play ~17–29 more games before injury. No effects of shoe or surface stiffness were observed for PT strain or jump height. These results indicate that shoe stiffness and surface construction may be effective interventions to decrease the likelihood of Achilles tendinopathy without affecting jump performance in athletes.

This work was funded by a research grant from the NBA and GE Healthcare Orthopedics and Sports Medicine Collaboration. Shoes and surfaces were provided by adidas AG and Robbins Sport Surfaces, respectively, and neither company was involved in the collection, analysis, or interpretation of any data. The authors report no conflicts of interest. The results of this study do not constitute endorsement by the American College of Sports Medicine. The authors declare that the results of this study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation.

REFERENCES

1. Lian OB, Engebretsen L, Bahr R. Prevalence of jumper’s knee among elite athletes from different sports: a cross-sectional study. Am J Sports Med. 2005;33(4):561–7.
2. Kujala UM, Sarna S, Kaprio J. Cumulative incidence of Achilles tendon rupture and tendinopathy in male former elite athletes. Clin J Sport Med. 2005;15(3):133–5.
3. Kettunen JA, Kvist M, Alanen E, Kujala UM. Long-term prognosis for jumper’s knee in male athletes. A prospective follow-up study. Am J Sports Med. 2002;30(5):689–92.
4. Visnes H, Bahr R. The evolution of eccentric training as treatment for patellar tendinopathy (jumper’s knee): a critical review of exercise programmes. Br J Sports Med. 2007;41(4):217–23.
5. Maffulli N, Frcs M, Pankaj O, Mrcs S, Luscombe KL. Achilles tendinopathy: aetiology and management. J R Soc Med. 2004;97(8):472–6.
6. Scattone Silva R, Purdam CR, Fearon AM, et al. Effects of altering trunk position during landings on patellar tendon force and pain. Med Sci Sports Exerc. 2017;49(12):2517–27.
7. Fukashiro S, Komi PV, Järvinen M, Miyashita M. In vivo Achilles tendon loading during jumping in humans. Eur J Appl Physiol Occup Physiol. 1995;71(5):453–8.
8. Fung DT, Wang VM, Andarawis-Puri N, et al. Early response to tendon fatigue damage accumulation in a novel in vivo model. J Biomech. 2010;43(2):274–9.
9. Millar NL, Murrell GAC, Mcinnes IB. Inflammatory mechanisms in tendinopathy—towards translation. Nat Rev Rheumatol. 2017;13(2):110–22.
10. Edwards WB. Modeling overuse injuries in sport as a mechanical fatigue phenomenon. Exerc Sport Sci Rev. 2018;46(4):224–31.
11. Schechtman H, Bader DL. In vitro fatigue of human tendons. J Biomech. 1997;30(8):829–35.
12. Wren TAL, Lindsey DP, Beaupre GS, Carter DR. Effects of creep and cyclic loading on the mechanical properties and failure of human Achilles tendons. Ann Biomed Eng. 2003;31(6):710–7.
13. Ferretti A, Puddu G, Mariani PP, Neri M. Jumper’s knee: an epidemiological study of volleyball players. Phys Sportsmed. 1984;12:97–106.
14. Malisoux L, Gette P, Urhausen A, Bomfim J, Theisen D. Influence of sports flooring and shoes on impact forces and performance during jump tasks. PLoS One. 2017;12(10):e0186297.
15. Van Der Worp H, Vrielink JW, Bredeweg SW. Do runners who suffer injuries have higher vertical ground reaction forces than those who remain injury-free? A systematic review and meta-analysis. Br J Sports Med. 2016;50(8):450–7.
16. Milner CE, Ferber R, Pollard CD, Hamill J, Davis IS. Biomechanical factors associated with tibial stress fracture in female runners. Med Sci Sports Exerc. 2006;38(2):323–8.
17. Davis IS, Bowser BJ, Mullineaux DR. Greater vertical impact loading in female runners with medically diagnosed injuries: a prospective investigation. Br J Sports Med. 2016;50(14):887–92.
18. Tsaopoulos DE, Baltzopoulos V, Maganaris CN. Human patellar tendon moment arm length: measurement considerations and clinical implications for joint loading assessment. Clin Biomech (Bristol, Avon). 2006;21(7):657–67.
19. Maganaris CN, Baltzopoulos V, Sargeant AJ. In vivo measurement-based estimations of the human Achilles tendon moment arm. Eur J Appl Physiol. 2000;83(4–5):363–9.
20. Winter DA. Biomechanics and Motor Control of Human Movement. 4th ed. Waterloo: John Wiley & Sons, Inc.; 2009.
21. Buchanan TS, Moniz MJ, Dewald JPA, Rymer WZ. Estimation of muscle forces about the wrist joint during isometric tasks using an EMG coefficient method. J Biomech. 1993;26(4–5):547–60.
22. Tomasi C, Kanade T. Detection and Tracking of Point Features. Rochester (NY): Image; 1991.
23. Maganaris CN, Paul JP. Tensile properties of the in vivo human gastrocnemius tendon. J Biomech. 2002;35(12):1639–46.
24. Maganaris CN. Tendon conditioning: artefact or property? Proc Biol Sci. 2003;270(1 Suppl):S39–42.
25. Robertson DGE, Caldwell GE, Hamill J, Kamen G, Whittlesey SN. Research Methods in Biomechanics. Windsor: Human Kinetics; 2004.
26. Vaughan CL, Davis BL, O’Conner JC. Dynamics of Human Gait. Cape Town, South Africa: Kiboho Publishers; 1999.
27. Scott SH, Winter DA. Internal forces of chronic running injury sites. Med Sci Sports Exerc. 1990;22(3):357–69.
28. Arnold EM, Ward SR, Lieber RL, Delp SL. A model of the lower limb for analysis of human movement. Ann Biomed Eng. 2010;38(2):269–79.
29. Fath F, Blazevich AJ, Waugh CM, Miller SC, Korff T. Direct comparison of in vivo Achilles tendon moment arms obtained from ultrasound and MR scans. J Appl Physiol. 2010;109(6):1644–52.
30. Németh G, Ohlsén H. In vivo moment arm lengths for hip extensor muscles at different angles of hip flexion. J Biomech. 1985;18(2):129–40.
31. Ward SR, Eng CM, Smallwood LH, Lieber RL. Are current measurements of lower extremity muscle architecture accurate? Clin Orthop Relat Res. 2009;467(4):1074–82.
32. Visser JJ, Hoogkamer JE, Bobbert MF, Huijing PA. Length and moment arm of human leg muscles as a function of knee and hip-joint angles. Eur J Appl Physiol Occup Physiol. 1990;61(5-6):453–60.
33. Herzog W, Read LJ. Lines of action and moment arms of the major force-carrying structures crossing the human knee joint. J Anat. 1993;182(Pt 2):213–30.
34. Tsaopoulos DE, Baltzopoulos V, Richards PJ, Maganaris CN. In vivo changes in the human patellar tendon moment arm length with different modes and intensities of muscle contraction. J Biomech. 2007;40(15):3325–32.
35. Lichtwark GA, Wilson AM. In vivo mechanical properties of the human Achilles tendon during one-legged hopping. J Exp Biol. 2005;208(Pt 24):4715–25.
36. Sheehan FT, Drace JE. Human patellar tendon strain. Clin Orthop Relat Res. 2000;370:201–7.
37. Reeves ND, Maganaris CN, Narici MV. Effect of strength training on human patella tendon mechanical properties of older individuals. J Physiol. 2003;548(Pt 3):971–81.
38. Webster KE, Kinmont CJ, Payne R, Feller JA. Biomechanical differences in landing with and without shoe wear after anterior cruciate ligament reconstruction. Clin Biomech (Bristol, Avon). 2004;19(9):978–81.
39. Farris DJ, Lichtwark GA, Brown NA, Cresswell AG. The role of human ankle plantar flexor muscle–tendon interaction and architecture in maximal vertical jumping examined in vivo. J Exp Biol. 2016;219(Pt 4):528–34.
40. Stefanyshyn DJ, Nigg BM. Influence of midsole bending stiffness on joint energy and jump height performance. Med Sci Sports Exerc. 2000;32(2):471–6.
41. Tinoco N, Bourgit D, Morin JB. Influence of midsole metatarsophalangeal stiffness on jumping and cutting movement abilities. Proc Inst Mech Eng P J Sport Eng Technol. 2010;224(3):209–17.
42. McInnes SE, Carlson JS, Jones CJ, McKenna MJ. The physiological load imposed on basketball players during competition. J Sports Sci. 1995;13(5):387–97.
43. Hoogkamer W, Kipp S, Frank JH, Farina EM, Luo G, Kram R. A comparison of the energetic cost of running in marathon racing shoes. Sports Med. 2018;48:1009–19.
Keywords:

ACHILLES TENDON; PATELLAR TENDON; MUSCULOSKELETAL MODELING; TENDINOPATHY; INJURY PREVENTION

Supplemental Digital Content

Copyright © 2019 by the American College of Sports Medicine